". J. ": " R EESE LIBRARY CALIFORNIA UNIVERSITY OF CALIFORNIA. /EARTH SCIENCES LIBRARY Received Accessions No.2--*/--3-2-- Shelf No Wi 3 3 r. PHYSICAL GEOLOGY In Preparation Part II. STRATIGRAPHICAL GEOLOGY AND PALAEONTOLOGY GEOLOGY BY A. H. GREEN, M.A., F.G.S. PROFESSOR OF GEOLOGY IN THE YORKSHIRE COLLEGE, LEEDS; LECTURER ON GEOLOGY AT THE SCHOOL OF MILITARY ENGINEERING, CHATHAM ; FORMERLY OF H.M. GEOLOGICAL SURVEY; SOMETIME A SENIOR FELLOW OF GONVILLE AND CAIUS COLLEGE, CAMBRIDGE NJVERSITY PART 1. PHYSICAL GEOLOGY " That delightful labour of the imagination which is not mere arbitrariness, but the exercise of disciplined power combining and constructing with the clearest eye for probabilities and the fullest obedience to knowledge ; and then, in yet more energetic alliance with impartial Nature, standing aloof to invent tests by which to try its own work." GEORGE ELIOT. WITH ILLUSTRATIONS RIVINGTONS WATERLOO PLACE, LONDON MDCCCLXXXII [ Third and Enlarged Edition} EARTH SCIENCES LIBRARY TO THE of THE REV. W. H. COLEMAN MY FIRST TEACHER IN GEOLOGY PREFACE. THE Science of Geology, though barely yet a century old, covers already so wide a field, and takes in such a diversity of subjects, that few, if any, men can hope thoroughly to master the whole of it. Mineralogy, Petrology, Stratigraphical Geology, Terrestrial Cosmogony, Palaeontology, and other lines of research, though they may fairly be looked upon as subdivisions of Geology, are fast becoming separate Sciences. But while it has become almost an absolute necessity for most Geologists to concentrate their attention on some one department of the science and be content with a less perfect grasp of the rest, there is yet a certain basis or groundwork, with which every one who meddles with Geology, whatever be the branch to which he specially devotes himself, must be acquainted if his work is to be sound. For want of a knowledge of this groundwork the Petrologist, looking merely to chemical and mineralogical composition, classes together rocks which differ totally in their origin or manner of occurrence ; the Palaeontologist pure and simple is apt to force into an unnatural connection, on account of similarity in fossils, formations which physical evidence shows ought to be kept widely apart ; the Field Geologist is content with tracing boundaries on his map, and forgets to ask himself how his lines were produced and what they mean. It has been my aim in the present volume to attempt to give an outline of this fundamental groundwork, which in default of a better name I have called Physical Geology. Possibly under viii Preface. the heads of Crystallography, Mineralogy, and Petrography it may be thought that I have gone rather further than was needed for this end ; but if this be so, it has been done of set purpose. I found in my own case, and I have noticed in the course of my teaching, that it is at the outset that these subjects are most puzzling to the student ; but let him once get over his initiatory troubles, and his progress becomes smooth enough. Any attempt to treat these branches of Geology exhaustively in a general treatise would be either out of place or a failure, but I hoped that enough might be given to enable the reader to master their rudiments and so to put him in a position for taking up works specially devoted to them. But I would not have the student suppose that it will be necessary for him to be thoroughly up even in the elements of these subjects before he can go on to other parts of the book. Crystallography and Mineralogy can be learned only bit by bit ; it is only by long practice that one gains the geometrical habit of thought required by the one, and only by repeated examina- tion of specimens that the familiarity with minerals requisite for the other is obtained. While the student is picking up by degrees a knowledge of these branches of his subject, he may usefully vary his work by pushing on with other portions far more important and fortunately far easier to master. Pages 16 to 1G2 are the parts which can be assimilated only very slowly : he should run through these on first reading, and when he comes across anything that he does not readily understand, should make a note of it and keep it for more leisurely consideration. From page 162 onwards he will find all comparatively plain sailing; and while he works his way through the remainder of the book, he will from time to time turn back to the points that posed him in Crystallography, take one by one the various forms, draw and model them, and make each the companion of his thoughts during leisure moments till he becomes perfectly familiar with it ; or handle, turn over, test, and examine under his microscope the minerals mentioned one after the other, till each becomes to him like a well-known face. In this way he Preface. ix will acquire early in his studies a knowledge of the great broad truths of Geology, many of which .require little more than common-sense to grasp them, and at the same time will be gradually familiarizing himself with the more difficult and refined branches of the science. And in this way he will avoid the risk of becoming so captivated by the special study of minute details as to become blind or indifferent to the broad generalizations which it is the aim of all science to attain to. The last two chapters have caused me many misgivings. To have passed over subjects of such interest on account of their difficulty would have been a piece of intellectual cowardice ; to have given without comment a bare list of the opinions of others would have been a confession of weakness that perhaps I had not the moral courage to make. But I have many a time regretted that I ever meddled with these branches of my subject ; it was in the highest degree unsatisfactory to have to deal with investigations of which one knew only the results and where one could not follow every step in the processes by which those results were arrived at : attempts to sift thoroughly the speculations of others were attended with a constant distrust of one's right to attempt the task, especially when it was recollected from whom the speculations came ; and on many other grounds the compilation of this part of the book was an undertaking that brought with it little but anxiety. A longing to get as near the truth as in me lay guided my efforts ; it furnishes no excuse for the imperfections of the result, but I hope it may be accepted as an apology for my having been rash enough to make the attempt. A work like this affords little scope for originality, and I doubt whether there is from the beginning to the end of the book a single thing of importance that can be said to be new. I have borrowed right and left \ in many cases my obligations are so obvious that it would have been unnecessarily burdening the pages with references to have acknowledged the sources of my information. But whether I have recognised my debt or not, I beg to offer my best thanks to those numerous brethren x Preface. of the hammer of whose labours I have availed myself without scruple and without stint. I must also content myself with a general acknowledgment of the not inconsiderable help I have received from private sources. Many of the additions and corrections in the second and in this edition have been suggested by friends or by readers who, though they are personally unknown to me, I venture to reckon among my friends on account of the kindly interest they have shown in my work. I must renew my thanks to Professor L. C. Miall for the important services he rendered me during the preparation of the first edition of this book, and add my acknowledgments of aid received from him subsequently and from my other colleagues Professors Riicker and Thorpe, and Mr. C. H. Bothamley. LEEDS, August 1882. CONTENTS. CHAPTER I. THE AIM AND SCOPE OF GEOLOGY, WITH A SKETCH OF ITS RISE AND PROGRESS. Pp. 1-7. CHAPTER II. DESCRIPTIVE GEOLOGY. Pp. 8-169. SECTION I. GENERAL RESULTS ARRIVED AT BY A LITHOLOGICAL EXAMINATION OF ROCKS. PAGE Descriptive Geology or Petrography ..... Historical Geology or Geogenie ...... 8 Subdivisions of Descriptive Geology ..... Lithology ......... 9 Petrology ......... 9 Instances of the Lithological Examination of Rocks ... 9 Definition of a Mineral . . . . . . .10 Definition of a Rock . . . . . . .11 SECTION II. MINERALOGY. Number of Rock-forming Minerals . . . . . .11 Chemical Composition of Rock-forming Minerals . . . .12 Means of recognising Minerals . . . . . .14 Definition of a Crystal . . . . . . .15 Simple Crystalline Forms . . . . . . .16 How the Crystalline Shape of a Mineral helps us to recognise it . . 18 Methods of denoting shortly the Faces of a Crystal . . .20 Zones ......... 25 Symmetry of Crystals ....... 25 Planes of Symmetry are of two Kinds . . . . .29 Planes of Symmetry in simple Solids ..... 29 Classification of Crystals according to their degree of Symmetry . . 30 What is meant by the word "Form" in Crystallography . . .32 Crystalline Cleavage . . . . . . .33 The Simple Forms of the Six Systems ..... 36 1. The Monometric System ...... 36 2. The Hexagonal System .... 50 3. The Dimetric System . . . . . .60 4. The Trimetric System ...... 63 5. The Monoclinic System ...... 6 6. The Triclinic System ...... 6$ xii Contents. PAGE Instances of Combinations in Crystals ..... 73 Twin Crystals . . . . . . . .75 Law connecting Crystalline Form and Chemical Composition, and excep- tions to it 79 Crystallized and Crystalline States Amorphous States . The Colloidal States The Glassy State . Allotropism and Isomerism . Other Shapes taken by Minerals The way in which Minerals break Tenacity Colour and Streak . Hardness Specific Gravity Taste and Smell 82 82 83 84 85 85 85 86 86 87 88 88 SECTION III. DESCRIPTION or ROCK-FORMING AND SOME OTHER COMMON MINERALS. A. Minerals composed of Silica ...... 88 B. Minerals composed of Silicates ...... 92 (B 1) The Felspar Group ..... 94 (B 2) Minerals allied to the Felspars in composition but differing from them in Crystalline form . . . .101 (B 3) Sodalite Group ... .104 (B 4) The Pyroxene and Amphibole Groups .... 106 (B 5) The Enstatite or Trimetric Pyroxene Group . . .111 (B 6) The Olivine Group . . . . . .113 (B 7) Mica Group ....... 114 (B 8) Hydrated Magnesian Silicates . . . . .120 (B 9) Chlorite Group ....... 122 (B 10) Silicates of Alumina . . . . . .124 (B 11) Garnet Group ....... 126 (B 12) Epidote Group . ..... 127 (B 13) Zeolite Group . . . . . . .129 (B 14) Sundry Silicates not coming under any of the preceding Groups 129 (B15) Hydrous Silicates of Alumina . . . . .133 C. Carbonates of Lime, Magnesia, and Iron . . . .134 D. Compounds of Lime other than those of the last Section . .140 E. Compounds of Barium . . . . . . .143 F. Compounds of Strontium ...... 145 G. Compounds of Iron . . . ... . . 146 (G 1) Oxides of Iron ....... 146 (& 2) Carbonates of Iron ...... 149 (G 3) Sulphides of Iron . . . . . .149 (G4) Silicates of Iron . ...... 151 (G 5) Other Compounds of Iron . . . . .152 H. Compounds of Manganese ...... 155 /. Compounds of Aluminium . . . . . .158 K. Elements ........ 160 L. Chlorides and Fluorides . 161 SECTION IV. LITHOLOGICAL CLASSIFICATION OF ROCKS. Lithological Classification of Rocks . . . . .163 Crystalline Rocks . . . . . . . .163 Non- crystalline Rocks . . . . . . .164 Contents. xiii SECTION V. PETROLOGY. PAGE Stratification ov Bedding ... ... 165 Relation between Stratification and Crystalline or Non-crystalline Texture 167 Fossiliferons and Unfossiliferous Rocks . . . . .167 Petrological Classification of Rocks . . . . . .168 Terms connected with Stratification . . . ., .168 Descriptive Geology. Summary . . . . . .169 CHAPTER III. L1THOLOGICAL DESCRIPTION OF THE NON-CRYSTALLINE ROCKS DENUDATION. Pp. 170-208. SECTION I. PRINCIPLES ON WHICH THE INQUIRY INTO THE ORIGIN or ROCKS is BASED. Principles on which the Origin of Rocks are determined . . .171 SECTION II. LITHOLOGICAL DESCRIPTION OF THE NON-CRYSTALLINE ROCKS. Texture . ....... 172 Subdivisions of the Non-crystalline Rocks ..... 172 1. Arenaceous or Sandy Rocks ..... 173 2. Argillaceous or Clayey Rocks . . . . .174 3. Calcareous Rocks or Limestones . . . . .178 4. Carbonaceous Rocks . . . . . .181 Other Rocks conveniently placed in the Non-crystalline Class . . 186 SECTION III. DENUDING AGENTS AND HOW THEY WORK. Example of the Determination of the Origin of a Rock . . .187 Denudation . . . . . . . . . 189 Enumeration of Denuding Agents . . . . . .189 The Work of Denuding Agents . . . . . .189 1. Rain ........ 190 Mechanical Action of Rain . . . . .190 Chemical Action of Rain ...... 190 2. Running Water . . . . . . .192 Rivers as Carriers of Sediment ..... 192 Denudation wrought by Rivers directly . . . .194 Underground Streams ...... 195 3. Frost and Ice . . . . . . 196 Frozen Water ....... 196 Glaciers ........ 196 Continental Ice-sheets ...... 200 Coast Ice . . . . . . . . 201 Ground Ice ....... 201 4. Action of Wind ....... 201 5. Organic Denuding Agents ...... 203 General View of Subaerial Denudation .... 203 Formation of Soil ...... 203 Removal of Soil from higher to lower Levels . . . 205 6. Marine Denudation ...... 206 Relative Importance of Subaerial and Marine Denudation . 207 xiv Contents. CHAPTER IV. WHAT BECOMES OF THE WASTE PRODUCED AND CARRIED OFF BY DENUDATION. THE METHOD OF FORMATION OF BEDDED ROCKS, AND SOME STRUCTURES IMPRESSED ON THEM AFTER THEIR FORMATION. Pp. 209-282. SECTION I. MATTER MECHANICALLY CARRIED. Arrangement of Mechanical Deposits according to Size and Weight . 210 Arrangement of Mechanical Deposits according to Mineral Composition . 210 General Arrangement of Mechanical Deposits . . . .211 Influence of Salt in promoting Deposition ..... 214 Horizontal Growth of Coarse Deposits ..... 214 Vertical Growth of Coarse Deposits . . . . .214 Drift- or Current-bedding . . . . . . .215 Ripple-drift ......... 216 Contemporaneous Erosion . . . . . . .217 Ripple-marks, Rain-drops, Sun-cracks, and Animal-tracks . .217 Summary of Characteristics of Shallow-water Deposits . . . 218 General Character of Deposits of finely-divided Matter . . . 218 Stratification, and Thickness of Beds ..... 219 Parallel between Modern Bedded Deposits and Stratified Rocks . .219 SECTION II. DISSOLVED MATTERS EXTRACTED BY THE AGENCY OF ANIMALS AND PLANTS. Foraminifera ........ 220 Coral ......... 223 Rocks formed out of Coral ....... 228 Other Limestone-secreting Animals ..... 229 Dr. Sorby's Conclusions ....... 230 Place of Pure Limestone on the Sea-bed ..... 230 Origin of Pure Limestones and Inference from their Presence . . 230 Animals and Plants secreting Silica ..... 231 Siliceous Nodules in Organic Limestones ..... 231 Glauconite Grains, Greensands . . . . . .233 Red Clay of the Atlantic ....... 233 SECTION III. MATTER CARRIED IN SOLUTION AND THROWN DOWN BY PRECIPITATION. Means by which Precipitation may be brought about . . . 236 Conditions necessary for Precipitation ..... 237 Instances of Rocks formed by Evaporation of saturated Solutions . . 238 Rocks formed by Precipitation consequent on the removal of a Solvent . 242 Rocks formed by Precipitation consequent on decrease of Temperature . 243 Rocks formed by Precipitation consequent on Chemical Reactions between Dissolved Salts ........ 243 Methods by which Dolomite and Gypsum may have been formed . . 245 Some Peculiarities of the Red Beds associated with Dolomite and Gypsum 248 Some other Rocks formed by Precipitation ..... 250 Sources of the Materials for Deposits formed by Precipitation . . 251 SECTION IV. TERRESTRIAL DEPOSITS. Soil and Rain-wash ........ 252 Screes ......... 254 Blown Sand . 256 Contents. xv PAGE Rocks of Vegetable Origin ....... 257 Coal. . 257 Partings and Sundry Irregularities in Coal Seams . . 260 Subaqueous Coal ........ 261 Cannel Coal . . ... 262 SECTION V. DEPOSITS OF ICE-FORMED DETRITUS. Distinctive Characters of Ice-borne Detritus . . . 262 Forms of Glacial Deposits ....... 263 Till ... .263 Moraines ......... 265 Glacial Mud ........ 265 Boulder Clays ...... .266 Erratics and Perched Blocks . . . . . .266 Rearranged Glacial Beds ... 266 Rocks and Deposits of Glacial Origin . 267 SECTION VI. How SEDIMENT is COMPACTED INTO ROCK. Weight of Overlying Masses . . . . . .268 Deposition of Cement ... . 268 Chemical Reactions ........ 268 Internal Heat ........ 268 Pressure ..... .268 SECTION VII. SOME STRUCTURES IMPRESSED ON ROCKS AFTER THEIR FORMATION. Slaty Cleavage ........ 270 Jointing .... .... 274 Concretions ......... 279 Concretionary Structure in Rocks ...... 280 Oolitic Structure . . . . . . . .281 Secretionary Nodules ....... 282 CHAPTER V. DEFINITION AND CLASSIFICATION OF DERIVATIVE ROCKS: AND HOW FROM A STUDY OF THEIR CHARACTERS WE CAN DETERMINE THE PHYSICAL GEOGRAPHY OF THE EARTH A T DIFFERENT PERIODS OF ITS PAST HISTOR Y. Pp. 283-303. Derivative Rocks and their Classification ..... 283 Importance of learning the Conditions under which Rocks were formed . 286 Teaching of Glacial Formations ...... 287 A . Marine Rocks Littoral Rocks ........ 288 Thalassic Rocks . . . . . . . 289 Normal Oceanic Rocks ...... 290 Erratics in Oceanic Deposits ...... 290 Chemical Deposits in Oceanic Areas ..... 291 B. Estuarine Rocks Shape in section of Deltas ...... 293 Fossils of Estuarine Beds ...... 293 Deposits formed by the Union of Deltas .... 294 Example of an Estuarine Group ..... 294 xvi Contents. C, Lacustrine Rocks PAGE Fresh-water Lacustrine Deposits ..... 298 Salt-water Lacustrine Rocks ...... 299 Example of Chemically-formed Deposits .... 299 D. Terrestrial Rocks Application to a particular instance . . . . . 301 CHAPTER VI. LITHOLOGICAL DESCRIPTION OF THE CONFUSEDLY- CRYSTALLINE ROCKS. Pp. 304-338. Texture of Crystalline Rocks ...... 304 Structures in Glassy and Devitrified Rocks .... 305 Spheroidal Structure ....... 307 Tabular or Laminated Structure ...... 307 Columnar Structure ....... 308 Felsitic Matter ........ 309 Fluid-, Glass-, Stone-, and Gas-Cavities . . . . .310 Vesicular and Amygdaloidal Structure . . . . .311 Subdivisions of the Crystalline Rocks . . . . .311 Acid Rocks ......... 312 Basic Rocks . . . . . . . . . 312 A. Acid Rocks The Granite Group . . . . . . .315 The Quartz-felsite or Elvanite Group . . . . .317 The Felstone Group . . . . . .319 The Quartz-trachyte and Rhyolite Group .... 320 Glassy Rocks of Acid composition . . . . .321 Relation of different forms of Crystalline Acid Rock to one another . 323 B. Intermediate Rocks The Syenite Group ....... 324 The Trachyte Group ....... 325 The Mica- trap Group . . . . . .327 C. Basic-crystalline Rocks Diorite Group ........ 328 Diabase Group ........ 329 The Basalt Group . . . . . .331 The Norite and Hypersthenite Group ..... 333 The Nephelinite and Leucitite Group ..... 333 Glassy Rocks of Basic composition ..... 333 Peridotite Group . ..... 334 Summary ........ 335 CHAPTER VII. VOLCANIC ROCKS. Pp. 339-389. SECTION I. INTRODUCTORY. Other points of resemblance between Volcanic Rocks and Modern Lavas . 341 Definition of Volcanic Rocks . . . . . .342 Mineral Composition of Volcanic Rocks and Modern Lavas . . 343 SECTION II. PHENOMENA OF VOLCANIC ACTION. Producing Causes of Volcanic Eruption . . . . 344 Structure of Single Cone . . . . . . .346 Contents. xvii PAGE Cessation and Repetition of Eruption ..... 346 Truncation and Breaching of Cone : Production of Crater and Cone within it ...... Subsidence after Cessation of Eruptions Development and Decay of Volcanic Activity Examples of the different Types of Volcanic Mountains Submarine Eruptions ..... 352 Dispersion of Ash and flow of Lava beyond the Cone : Prolonged Dykes . Fissure Eruptions ..... . 352 SECTION III. VOLCANIC PRODUCTS. 1. Lavas Fluidity ........ 353 Motion of Subaerial Lava Streams and Texture of their different Parts. ... .... 355 Characters of Lava which has hardened under pressure . . 357 Formation of Crystals in Lava ...... 357 Bedded Structure ... . . .358 2. Fragmeutal Products- Structure of Subaerial Ashy Deposits ... Structure of Subaqueous Ashy Deposits Torrential Accumulations of Ash .... Other fragmental Volcanic Accumulations . . . 361 3. Gaseous and Sublimed Products Hot and Mineral Springs SECTION IV. REMNANTS OF OLD VOLCANOES. Ancient Volcanic Cones ..... Remains of Central Plug of Lava ...... 364 Other proofs of Old Volcanic Action Example of Arthur's Seat .... . 364 Ancient Volcanoes of North Wales ...... 368 SECTION V. PETROLOGY OF VOLCANIC ROCKS. Distinction into Intrusive and Contemporaneous .... 369 Alteration of Neighbouring Rocks . . . . . .371 Included Fragments ....... 371 Fragmental Interbedded Rocks ... . 372 Necks of Agglomerate ....... 372 Instances of the Modes of Occurrence of Volcanic Rocks . . . 372 Circumstances under which Intrusive Sheets were injected . . 377 SECTION VI. LITHOLOGICAL VARIETIES OF VOLCANIC ROCKS. Identity of Modern and Old Volcanic Rocks .... 378 SECTION VII. MUD VOLCANOES. SALSES. 381 SECTION VIII. EARTHQUAKES. How Earthquakes are caused ...... 383 Earthquakes as Geological Agents ...... 388 xviii Contents. CHAPTER VIII. PLUTONIC ROCKS. Pp. 390-398. PAGE Differences between Plutonic and Volcanic Rocks .... 390 Plutonic Rocks necessarily intrusive ..... 391 Mode of Occurrence of Plutonic Rocks ..... 392 Granite of Devon and Cornwall ...... 392 Granite Veins ........ 393 Included Blocks in Granite ....... 395 Contact-metamorphism by Granite ...... 395 Lithological Varieties of Plutonic Rocks ..... 396 Connecting-links between Volcanic and Plutonic Rocks . . . 396 Hydrothermal Origin of Plutonic Rocks . . . . . 397 CHAPTER IX. METAMORPHIC ROCKS. Pp. 399-442. SECTION I. GENERAL VIEW AND INSTANCES OF METAMORPHISM. General Remarks on Metamorphism . . . . . 399 Metamorphic Rocks of County Donegal ..... 401 Metamorphic Rocks of the Western Territories of the United States . 402 Effects of Metamorphism ....... 403 Subdivisions of Metamorphic Rocks ..... 403 SECTION II. DESCRIPTION or THE PRINCIPAL VARIETIES or THE METAMORPHIC ROCKS. 1st Class. Those which still retain Traces of Bedding and other obvious Proofs of their originally Derivative Condition (1 a) Siliceous Members ...... 404 (1 b) Argillaceous Members ..."... 406 (1 c) Calcareous Members ...... 406 (1 d) Carbonaceous Members . . . . . .413 (1 e) Altered Volcanic Ashes ...... 414 2nd Class. Foliated or Schistose Rocks Nature of Foliation . . . . . . .414 Classification of Foliated Rocks . . . . .415 Schists ......... 415 Gneisses ........ 416 Serpentine ........ 417 Nodular and Spotted Schists . . . . . .417 How Foliation has been produced ..... 418 Degrees of Foliation ....... 418 Foliated Rocks of the Pyrenees . . . . .419 "What determines the Planes of Foliation .... 421 Artificial Production of Cleavage-Foliation .... 422 Crumpled Lamina . . . . . . .423 Intrusive Schistose Rocks ...... 424 Summary ........ 425 Early Theories about Crystalline Schists .... 426 3rd Class. Amorphous Metamorphic Rocks General Description ... ... 427 Modes of Occurrence . . . . . . .428 Relation to Plutonic Rocks ...... 429 Claystone of Chili ....... 429 Contents. xix Halleflinta ...... Rocks of Carrick in Ayrshire .... Priestlaw ...... South-west of Scotland .... Bedded Granites of Brittany .... Metamorphic Granites of the Western Territories of America Porphyroids of Nevada ..... Caution needed with Rocks of this Class PAGE 429 430 430 431 432 433 434 434 SECTION III. CAUSES OF METAMORPHISM. Local Metamorphism by Intrusive Igneous Rocks Heat one Agent Heat alone not enough Heated Vapours Water .... Pressure and Depth . Experiments of Daubree Researches of Sterry Hunt . Observations of Dr. Sorby . Variations in amount of Metamorphism Subsidiary Metamorphosing Agencies Summary .... Three Stages of Metamorphism Metamorphism no Proof of Antiquity 435 436 436 436 436 437 438 439 440 440 441 441 441 442 CHAPTER X. GENERAL VIEW OF THE CRYSTALLINE ROCKS. Pp. 443-455. Natural Classification of Crystalline Rocks . Are all Crystalline Rocks Metamorphic Products ? Instances of Irruptive behaviour by Metamorphic Rocks Counter-views on the Origin of Crystalline Rocks . Objections to Metamorphic Theory . Summary and Conclusions .... Further Classification of the Crystalline Rocks 443 444 447 448 448 449 452 CHAPTER XI. HOW THE ROCKS CAME INTO THE POSITIONS IN WHICH WE NOW FIND THEM. Pp. 456-527. SECTION I. NATURE OF THE DISPLACEMENTS WHICH ROCKS HAVE UNDERGONE. Displacements which Submarine Beds have suffered . . .456 SECTION II. VERTICAL ELEVATION. Two possible Explanations of Elevation . . . .456 Arguments against a lowering of the Sea-level . . . 457 The Land has gone up, not the Sea gone down . . 457 Denudation gives Proof of Elevation . . . .457 Instances of observed Oscillation of Land ... . 458 Submergence produced by a Polar Icecap ... . 459 xx Contents. SECTION III. DISPLACEMENT OF THE ROCKS FROM THEIR ORIGINALLY HORIZONTAL POSITION. PAGE Dip . . . . . . . . . .460 Strike ......... 460 Amount and Direction of Dip ...... 461 Measurement of Dip ....... 462 Problems connected with Dip ...... 464 Measurement of Thickness of Strata ..... 464 Outcrop ......... 468 Breadth of Outcrop ........ 472 Undulations and Contortions ...... 473 Anticlinal and Synclinal ; Dome and Basin .... 476 Anticlinal ......... 477 Dome ......... 477 Synclinal and Basin ....... 479 Parallelism of Anticlinals ....... 479 Classes of Anticlinals ....... 480 Inversion ......... 482 Outlier and Inlier . 485 SECTION IV. FAULTS. Definition of a Fault ....... 489 Slickenside ......... 491 Hade of Faults . . . . . . . .492 Course of Faults . . . . , . . .492 Parallelism of Faults ....... 492 Changes in Size and Dying out of Faults ..... 493 Effect of Faults on Outcrop ...... 496 Indirect Evidence for Faults 498 SECTION V. How THE DISPLACEMENTS OF THE ROCKS WERE PRODUCED. Character of the Movements ...... 501 Folding would produce both Elevation and Dip . . . . 503 Direction of the Folding Force . . . . . 503 Summary of the Evidence . . . . . . 510 Folding went on at great Depths . . . . . .510 Folding went on slowly . . . . . . .512 Contortions more frequent in Old than in Recent Rocks . . . 513 SECTION VI. UNCONFORMITY AND OVERLAP. "What constitutes Unconformity ...... 514 Meaning of Unconformity . . . . . . .515 Deposition on Sinking Sea-bottoms ..... 517 General Conclusions . . . . . . .518 Illustration of Unconformity ...... 518 Incidental Proofs of Unconformity ...... 521 Deceptive Appearance of Unconformity owing to Underground Dissolution of Rock. ........ 521 Deceptive Conformity ....... 521 Overlap ... .523 Practical Bearings ........ 526 Contents. xxi CHAPTER XII. MINERAL DEPOSITS AND METALLIC ORES. Pp. 528-570. SECTION I. DESCRIPTION OF THE MORE IMPORTANT METALLIC ORES AND SOME OTHER METALLIC MINERALS. PAGE 1. Ores of Copper ........ 528 (la) Sulphides of Copper ...... 529 (1 b) Oxides of Copper. . . . . . .530 (1 c) Carbonates of Copper . . . ' . . 530 (1 d) Silicates of Copper . . . . . .531 (1 e) Chloride of Copper . . . . . .531 2. Ores of Lead ........ 531 3. Ores of Zinc ........ 534 4. Tin ..... ... 536 5. Ores of Silver ........ 536 6. Ores of Cobalt and Nickel ...... 538 7. Ores of Antimony ....... 542 8. Ores of Arsenic ........ 542 9. Bismuth ........ 543 10. Ores of Mercury ....... 544 11. Gold ......... 544 12. Platinum ........ 544 13. Chromium . . . . . . . .545 14. Molybdenum ........ 545 15. Tungsten or Wolfram . . . . . . .545 16. Titanium ........ 546 17. Uranium . . . . . . . 547 SECTION II. METALLIC DEPOSITS. Nature and Forms of Mineral Deposits ..... 548 1. Lodes Definition of a Lode . . . . . . 545 Comby Lodes ........ 549 Brecciated Lodes ....... 550 Terms connected with Lodes ...... 550 Gossan ......... 550 Heaving of Lodes ....... 551 Direction of Lodes ....... 552 Dimensions and Extent of Lodes ..... 552 Relation between Contents of Lodes and adjoining Rocks . . 553 Flats ...... . . .555 Contact Deposits ....... 555 2. Stockworks Deposits simulating Lodes but without well-defined Walls . . 555 3. On the way in which Lodes and Stockworks were filled By the Chemical Action of Percolating Fluids . . .557 By Sublimation from below ..... 562 By Electro-chemical Action ...... 553 4. Masses Iron Pyrites Deposits of Andalucia ..... 564 Deposits of Haematite in Cumberland ..... 535 5. Impregnations Alderley Edge ........ 566 Kupferschiefer . . . . . . . .566 Plumbiferous Sandstone of Bleiberg in the Eifel . . 557 Native Copper of Lake Superior . . . 537 Ore-bearing Crystalline Rocks ..... 568 xxii Contents, 6. Beds Bog-iron Ore ....... 569 Clay Ironstone ........ 569 Cleveland and Northamptonshire bedded Ironstones . . . 569 Bedded Iron Ores of Canada ...... 569 7. Placers ......... 570 CHAPTER XIII. HOW THE PRESENT SURFACE OF THE GROUND HAS BEEN PRODUCED. Pp. 571-638. SECTION I. PROOFS THAT THE SHAPE OF THE SURFACE is DUE TO DENUDATION. Surface due to Denudation . . . . . . .571 Amount of Denudation ..... . 574 SECTION II. THE SHARE OF EACH DENUDING AGENT IN PRODUCING THE SHAPE OF THE SURFACE. Share of the Sea ........ 576 Plain of Marine Denudation ...... 577 Share of Subaerial Denuding Agents. Rivers .... 578 Canon of Colorado an Example of River-action .... 580 Other Subaerial Denuding Forces ...... 583 Landslips ......... 583 Basin -shaped Lie of Outliers . . . . . . 585 History of the Idea of Subaerial Denudation .... 585 SECTION III. How THE CHARACTER AND LIE OF THE UNDERLYING RoCKS AFFECT THE SHAPE OF THE GROUND. Relative Hardness . . . . . . . .586 Other Qualities which enable Eocks to resist Denudation . . . 588 Difference between Results of Marine and Subaerial Denudation . . 589 Effect of Natural Planes of Division ..... 590 Effect of the Lie of the Beds on the Shape of the Surface . . . 593 Steps in the Formation of the Surface ..... 594 Escarpment and Dip-slope ....... 597 Valleys determined by Joints ...... 601 Valleys determined by Faults ...... 601 Qualifications ........ 601 Final Results of Subaerial Denudation ..... 603 Cutting back of the Channels of Rivers ..... 603 Alluvial Plains 605 SECTION IV. RIVERS AND THEIR WAYS. Breaching of Hill-ranges by Rivers ...... 606 The Green and Yampa Rivers in Utah . . 609 Rivers running away from the Sea . . . . . .611 Reversal of the Flow of Rivers ...... 612 Breaching of great Mountain-chains by Rivers .... 612 Contents. xxiii SECTION V. FEATURES DUE TO THE ACTION OF ICE. PAGE General Aspect of Ice-worn Districts . . . . '. 614 Polished Surfaces ........ 614 Scratches . . . . . . . . . .614 Roches Moutonnees ........ 616 Moraines . . . . . . . . .617 SECTION VI. LAKES. Dammed-up Lakes . . . . . . . .617 Rock Basins. ........ 619 Rock Basins formed by Subsidence or Upheaval .... 621 Rock Basins scooped out by Ice . . . 622 SECTION VII. SURFACES NOT WHOLLY DUE TO DENUDATION. Mountain-chains ........ 625 Volcanic Cones ........ 631 Eskers ......... 632 Moraines ......... 633 Sand-dunes ......... 635 Lakes enclosed by heaped-up Mounds ..... 635 Old River-terraces ........ 635 Sea-beaches ......... 636 Raised Beaches ........ 636 Surfaces of Deltas ........ 636 Silted-up Lakes . . . . . . . .637 Prairies and Deserts ........ 637 Summary ......... 637 CHAPTER XIV. ORIGINAL FLUIDITY AND PRESENT CONDITION OF THE INTERIOR OF THE EARTH. CAUSE OF UPHEAVAL AND CONTORTION. ORIGIN OF THE HEAT REQUIRED FOR VOLCANIC ENERGY AND METAMORPHISM. REMARKS ON SPECULATIVE GEOLOGY. Pp. 639-697. SECTION I. THE PRESENT PHYSICAL CONDITION OF THE EARTH. Shape of the Earth ........ 641 Mean Density of the Earth . . . . . . . . 642 Internal Temperature of the Earth ...... 643 Inferences from the foregoing Facts ..... 644 Investigation of the Figure of the Earth on the Hypothesis of its original Fluidity ........ 645 Present State of the Earth's Interior ..... 650 Doctrine of a Thin Crust ....... 650 Sir W. Thomson's Views on Underground Temperature . . . 650 Effect of Pressure on the Fusing- point . . . . .651 Various Modes of Consolidation possible ..... 655 Argument from Precession ....... 656 Argument from Rigidity ..... .657 Objections to the preceding Arguments ..... 659 Professor Hennessy's Views . . . . . . .661 Summary ...... 663 xxiv Contents. SECTION II. WHAT THE INTERIOR OF THE EARTH is MADE OF. PAGE The Argument from Meteorites . . . . . .664 Argument from Metallic Lodes ...... 667 SECTION III. CAUSE OF UPHEAVAL AND CONTORTION. Sense in which Elevation is used . . . 667 General Structure of Mountain-chains . . . 667 Mr. Hopkins' Theory ... . .668 Theory of Scrope and Babbage . . . 669 Theory of Sir J. Herschel ... . .670 Intrusion of Granite ... . . 670 Laccolites ..... . . 671 Contraction Theory .... . . 672 SECTION IV. ORIGIN OF THE HEAT REQUIRED FOR VOLCANIC ACTION AND METAMORPHISM. Explanation on Hypothesis of a Thin Crust .... 675 Mr. Hopkins' Hypothesis . . . . . . .676 Explanations of Mr. Scrope and Rev. 0. Fisher . . . . 676 Dr. Sterry Hunt's Hypothesis . . . . . . 678 Mr. R. Mallet's Hypothesis ..... . 679 Bearing of Mr. Mallet's Hypothesis on Metamorphism . . . 684 SECTION V. THE GREAT PHYSICAL FEATURES OF THE EARTH'S SURFACE. Continents ......... 685 Distribution of Volcanoes ....... 686 Doctrine of the Permanence of Oceans and Continents . . . 687 Growth of Continents ....... 689 Association of Volcanoes with Mountain-chains .... 691 Other Hypotheses ........ 692 SECTION VI. CONCLUDING REMARKS ON SPECULATIVE GEOLOGY. Geological Time . . . . . . . .693 Former greater Intensity of Geological Action .... 694 CHAPTER XV. ON CHANGES OF CLIMATE, AND HOW THEY HAVE BEEN BROUGHT ABOUT. Pp. 698 to end. Effect of Geographical Changes on Climate ..... 699 Astronomical Causes affecting Climate ..... 702 Evidence for and against the preceding Hypothesis . . . 709 The Effect on Climate of Geographical and Astronomical Causes combined 713 Other Hypothetical Explanations of Changes of Climate . . . 715 INDEX 719 UNIVERSITY! ' CHAPTER I. THE AIM AND SCOPE OF GEOLOGY, WITH A SKETCH OF ITS RISE AND PROGRESS. "Signaler les efforts par lesquels ont etc graduellement conquises les idees theoriques que nous possedons aujourd'hui n'est pas settlement nn just hommage rendu a ceux qui ont eclaire la science par leurs travaux : c'est aussi un avertisse- inent salutaire contre les illusions speculatives. " DAUBREE. THE thing aimed at first of all by Geology was to find out what the earth on which we live is made of. It is probable that the earliest cultivators of the science did not set themselves to do any more than this, and that the only objects they had in view were the examination of the materials out of which the solid framework of the earth is built up, and the determination of their chemical composition, physical properties, manner of occurrence, and other characteristics. True these pioneers had from time to time indeed, men of their acuteness could scarcely fail to have glimpses, from the outskirts where they were labouring, of the wide geological domain that lay beyond ; but for a long lapse of time the attempts to push onward into it were few and desultory.* Geology then began, as all sciences must begin, by being a bare record of observed facts. But Geology could not, any more than other sciences, stop here. Some of the inferences to be drawn from these facts stare us in the face so palpably that they could not long escape notice. Among the facts which in this way told their own story, one of the most obvious and the first to attract attention was the occur- rence in the heart of solid rocks, and at spots far inland and high above the sea-level, of what were undoubtedly the remains of marine animals. Two most important inferences followed from this first, the rocks could not always have been there, but must have accumu- lated round the remains they now enclose ; and, secondly, the arrange- ment of land and sea must have once been different from what it is now. In this way men came to learn that the earth had not sprung into being exactly as we have it now, but that changes had passed over it from time to time ; and then there arose a further branch of Geology, which had for its object to determine what these changes had been, and how they had been brought about. The doctrine that the earth had been subject to change, which con- * See Lyell, Principles, vol. i. chap, iii., for an account of the causes that hindered the advance of Geology. A 2 Geology. stitutes the very marrow of Geology, was established in the manner just described at least as far back as the days of Pythagoras, but it was long before the science made any advance beyond this and a few elementary truths of a like nature. The attempts made to give any rational explanation of the way in which the changes had been effected were only partly true, or were wholly erroneous. Some geologists failed on account of limited experience ; they looked upon the tract they were acquainted with as a type of the whole globe, and their explanations, though well suited to local instances, were not of general application. Others were hampered by preconceived notions that geological changes had been produced all at the same time, and all by the same cause ; Noah's Deluge, for instance, was a favourite resource with this school, and was credited with far more important results than it could possibly have effected, even if the popular notion of it were correct. A third class, the most mischievous of all, " took not their material from Nature, but spun it out of themselves;" they discarded observation altogether, and amused themselves with weaving ingenious conceits as to how the earth might have been brought into its present shape ; and when they found themselves in a diffi- culty, did not hesitate to call in to their aid agencies the like of which had never either been seen or heard of, and the like of which, as far as our knowledge of the economy of Nature goes, could never have been in operation. The wild, dreamy speculations of this school, with their convulsions, cataclysms, inundations, collisions with comets' tails, and other fanciful occurrences, read like a translation of a Scan- dinavian saga without the life of the original, and it is really hard to believe that they could ever have been seriously put forward by men calling themselves scientific students. Among the other causes which hindered the progress of the science, we may specially mention a dreary controversy, which dragged its slow length along over more than a century, as to whether the things in the rocks, which appeared to be organic remains, ever had belonged to living animals, and were not rather counterfeits, moulded by Nature in some elfish mood pur- posely to lead mankind astray. It is scarcely believable that so impu- dent a notion could ever have found supporters ; but it did, and no lack of them either. The early history of Geology then consists of a record of one long string of failures, but the study of these will be by no means barren of results, if we look at them from the right point of view. When we see men of unquestionable power going on for centuries missing the mark, the only conclusion we can come to is that there was something radically wrong in the way in which they went to work. And it is easy enough to see what it was that was wrong in their method. If we want to learn how a piece of furniture, a pair of shoes, or a coat is made, we don't sit down and waste our time in barren guesses and random shots, but we go to the cabinetmaker, the shoemaker, or the tailor, and watch them at their work. Nor will one visit be enough ; if we wish really to get to the bottom of the matter, we must go again and again, till we have made ourselves masters of every step in the process of manufacture. Just so, if we want to learn how any natural Historical Sketch. 3 product arose, we must haunt the workshop of Nature, till by long and repeated study we wring from her the secrets of her trade, and gain an insight into every step of her complicated and manifold operations. JS T ow, this is just what the earlier geologists did not do. Some observed, but did not observe enough ; others shirked altogether the labour of observation, and tried to supply its place by specula- tions, which had nothing but imagination to rest upon. Practically, in spite of some advances every now and then in the right direction, Geology continued in this unsatisfactory state down to the end of the last century. Then there came on the stage Hutton, the kind of man the science had so long been in need of, and by his teaching geologists were at last started on the only path that could possibly lead them to truth. He pointed out, in words that could not be misunderstood, that if .we want to know what has happened on the earth in bygone times, we must begin by learning what is going on there now. He drove out once and for ever the imaginary agencies, which the earlier geologists had been so ready to have recourse to ; and laid down the principle that in geological speculation " no powers are to be employed that are not natural to the globe, no actions to be admitted of except those of which we know the principle, and no extraordinary events to be alleged in order to explain a common appearance."* Following out this principle, he said something like this : You have here a rock, and you want to know how it was formed. Well, what you must do is this. You must go and search whether there is anything now in the course of formation which is either identical with that rock or could be made identical with it by processes which we know Nature is capable of employing. When you find such a substance, learn what are the agents that are forming it. It will then strike you irresistibly that it is far more likely that your rock has been formed by agencies similar to those which are now pro- ducing a substance that cannot be distinguished from it, than that it was made by some imaginary, unheard-of, and improbable process. And Hutton laboured successfully to show that the forces now in action are f ally competent to form rocks, and to bring about a large portion of the changes, which we learn from Geology must have passed over the earth's surface. Here he stopped, confining himself to that portion of the earth's lifetime during which her physical con- dition has been similar to what it is at present. It is somewhat doubtful whether he even realized the probability of there having been a time when the earth was in a very different state from now ; but if he did, he declined to concern himself with the events and operations of such a period. It cannot be denied, then, that Hutton took rather a narrow view of the scope of Geology ; but in the portion to which he applied himself he may be fairly looked upon as the veritable father of the science ; and since the history of the earth as it is must be mastered before we can go on to unravel the history of the earth as it was before the present state of things was established, we may go further and call him the founder of Geology as a whole. Hutton, like most great men, was in advance of his age ; his teaching * Theory of the Earth, ii. 547. 4 Geology. fell dead,* till it was revived and illustrated by Lyell; but it is now universally recognised as the principle on which we must base all speculations relating to that part of the science of which he treated. Hutton occupied himself mainly in studying the changes that are now taking place on the earth's surface, and the means by which they are being brought about, and in demonstrating that the changes that had happened during past periods of the earth's history were of the same kind and due to the same causes as those now going on. He could not fail to realize clearly the fact, known before his day, that rocks were not all of the same age, and he describes t with rugged eloquence observations which showed him that some of the older rocks had been displaced from their original position and had suffered wear and tear before rocks of later date had been laid down upon them. But he did not go beyond these broad general facts, nor attempt to determine with any detail the order in which rocks had been formed. The first steps in this direction, sufficiently systematic to call for notice here, were made, about the same time, by two contemporaries of Hutton, Werner and William Smith. The former showed that the rocks of the part of Germany which he examined could be divided into certain groups, and that those groups came on, one over the other, in an order of succession which was everywhere the same. Thus if we call these consecutive groups a, b, and c, and note in one place that a is the undermost, b the middle, and c the uppermost of the three, we shall find these groups in the same relative position with regard to each other wherever we meet with them ; 1) will never be below a or above c. The same law holds good through- out the whole series of groups. Some members may be wanting in places, but this will not affect the place in the series of the rest ; thus if b be absent, c will rest on , never a on c. Some of Werner's sub- divisions agree pretty nearly with those of modern geologists ; others have been long ago discarded, because they were established on the strength of erroneous theories as to the way in which the rocks com- posing them had been formed. These theories were of the wildest description, wholly unsupported by observation or analogy, and, as they were put forward with a zeal and energy which gave their author great influence over his pupils, they contributed no little to hinder the pro- gress of Geology. Still it was a great step gained to have established the fact of the existence of an invariable order of succession in the rocks. While Werner was pursuing his investigations in Germany, William Smith was patiently at work among the rocks of England, paying special attention to the fossil remains of plants and animals which they contained. He found that the law which Werner had established for * How completely this was the case may be seen from the following passage, which sounds strange to the geologist of to-day: "The theory of Hutton has gradually sunk into disrepute in proportion as geological facts and observation have been more multiplied and extensive ; and it is not improbable but even the beautiful theory of Werner may share a similar fate, as some parts of it have met with considerable and powerful opposition." WESTGAIITH FOKSTEK, Section of the Strata (1821), p. 153. t Theory of the Earth, vol. i. chap. vi. Historical SketcJi. 5 the succession of rock groups in Germany was equally true for those of this country ; they were laid one upon the other in an order which was everywhere the same. His study of fossils enabled him to establish a further law of the greatest importance. He discovered that each rock group contained a number of fossils different from those in any other group, and that by means of these fossils it could be recognised and its place in the series determined, in cases where this could not be accomplished in any other way. Thus, suppose that we determine at any one spot the order in which the three groups, a, b, and c occur, and note and record the fossils found in b \ further, that at another spot we find rocks containing the same fossils as 6, but cannot see what is below or above them ; then, on the strength alone of the similarity of the fossil contents of the two rock groups, we may safely assert that these problematical rocks belong to the b group, and that below them there is either a or something lower in the series, and above them either c or something higher. From the first of the two laws just mentioned it was an easy step to show, as we shall see shortly, that the place of each rock group in the series gave the relative date of its formation, that the lowest was the oldest, the one above came next in point of time, and so on. From the second law we learn that the changes which had passed over the earth had not been confined to the inorganic portion of it, but had affected its living inhabitants as well ; that each period of its past history had had its own peculiar forms of life, and that these had from time to time died out and been replaced by new forms. Then there arose a further branch of Geology, which had for its objects to determine not merely what changes had happened formerly on the earth and how they had been brought about, but also the order in which they had occurred; and further, to describe the different living forms which had peopled the globe in former ages. These then seem to have been the main steps in the progress of Geo- logy. It began merely with the view of making out what the earth was made of with being merely a science of description and classifi- cation. Then in the pursuit of this study facts came out which told a story of former changes that had passed over the world, and geologists set to work to discover what these changes had been, and how they had been caused. Lastly, it was found possible to determine the order in which past changes had occurred, and the modifications in the forms of life by which they had been accompanied. The methods employed for these ends advanced Geology to the place of an inductive science, and their results enlarged its scope and gave rise to what may be called its historical branch. Thus there arose two main subdivisions of the science, which may be called Descriptive and Historical Geology ; and these it is still con- venient to retain. One object in giving the preceding sketch of the progress of the science has been to show that these are not mere arbitrary or even convenient divisions, but grew up with the growth of Geology itself. We have already mentioned that, even in its early or descriptive stage, many of the cultivators of Geology had foretastes of what it 6 Geology. would afterwards grow to; but the labours of Hutton and Smith, specially those of the former, may be said to have raised Geology, prac- tically at one step, from a bare record of observations to the dignity of an inductive science. Since their time it has grown apace, and no science can boast of a more rapid development.* The student will do well to mark that the great advance made by Hutton and Smith was won by systematic hard work in the field ; and he must bear in mind that no further progress can be made except in the same way : what may be called laboratory work, indispensable as it is, avails little or nothing in Geology, unless it rest on the firm basis of field investigation. Observations made out of doors need to be followed up by indoor work if we are to interpret them aright, but Geology is fundamentally an open-air study. Before going further, it will be well to inquire how much of the earth lies sufficiently within our ken to be properly the subject of geological investigation. Sea- and river cliffs, the beds of brooks, quarries, rail- way cuttings, and other artificial openings, show us what is found to a small depth below the surface, and mines enable us to feel our way a little lower down still ; but the portion of the earth's mass that we can examine by these aids alone, is evidently very small indeed. We can, however, from observations made at or near the surface, infer, with a very high degree of probability, what the composition of the earth is at depths far exceeding that of the deepest mine. Suppose, for instance, that we had proved, say by the mine shafts Fig. 1. A B 0, the presence of the three groups of rocks marked a b c in fig. 1, and had found them always to come on one over the other in the same order, and to keep a regular thickness over a considerable area, it is highly probable that these rocks, when beyond C they pass out of our sight, will preserve the same order and thickness in their underground course. Assuming this to be the case, a very simple calculation will give us the depth of any one group at a point E, and we can thus form a very probable conjecture as to the composition of the earth at a point such as E far below the bottom even of the deepest mine. In this way Van Decken has found that in parts of the Coal-basin of Saarbruck the character of the rocks may be deter- mined to a depth of more than three miles below the surface. In * For more particulars as to the history of Geology, see Lyell's Principles, vol. i. chaps, ii.-v. ; Phillips, Manual of Geology, chap. i. ; Conybeare and Phillips, Geology of England and Wales, Introduction. Carl Vogt, Lehrbuch der Geologic und Petrefactenkunde, vol. ii. 682-747. Daubree, Etudes sur le Metaraorphisme. Memoires presentes a 1'Academie des Sciences, tome xvii. Historical Sketch. 7 reasoning on a case of this sort we should feel still more confident in our conclusion if we found, as we often do find, the same groups of rocks reappearing in the same order from below, as in the shafts at F and G ; and we should then have little hesitation in showing their underground course by some such lines as the dotted ones in the figure. We should be able to determine with more accuracy the shape of these underground continuations if we could observe in the intermediate ground higher rocks, such as d e f, and g, and determine how they are lying. Thus, as Playfair remarks, "men can see much further into the interior of the globe than they are aware of, and geologists are reproached without reason for forming theories of the earth, when all they can do is but to make a few scratches on the surface." * Still, when we have pushed our investigations to the greatest depth that either direct observations, or reasoning that flows immediately from them, enable us to reach, we shall have made ourselves acquainted with no more than an outside shell or rind, not more than a few miles at the most in thickness. This shell, because it is so thin, is called the crust of the earth, and with it Geology is first of all of necessity concerned. But Geology need not stop here. When we have gathered some knowledge of the crust of the earth, we are naturally led on to make this knowledge a basis for speculations about the nature of those inac- cessible regions which lie below the crust. And when we have got together something like a history of the formation of the earth's crust, we are prompted to inquire whether there ever was a time when our planet was without a crust, what was its condition at that time, and how it passed from that condition into its present state. We shall see presently that as long as Geology confines itself to the crust of the earth, it is dealing with something corresponding to writ- ten records in history, and that the story these tell, is so clear and unmistakable that many of its conclusions are as certain as those of any other of the natural sciences ; but that, when it comes to treat of the inaccessible interior, or the history of the earth before it assumed its present state, there is nothing of the nature of a written record to guide its speculations, it has to lean mainly on a scientific use of the imagination, with little or no check from actual evidence, and its con- clusions consequently cannot rise above the rank of probabilities. Some geologists would limit Geology entirely to a study of the earth's crust, others take the wider view of its scope just given, f It will be seen that even in the more speculative parts of Geology we do not indulge in mere conjecture, but work persistently back from the seen and known to the unseen and unknown. We do not try to form any idea about the probable constitution of the invisible interior till we have made ourselves acquainted with the visible outside crust ; and we do not theorize about that part of the earth's history of which no tangible record is left, till we have diligently studied those later epochs of which some sort of written account has come down to us. * Works (ed. 1822), vol. i. p. 244. t Professor^Huxley's Anniversary Address to the Geological Society of London. Quarterly Journal, vol. xxv. ; and Geological Magazine, vol. vi. p. 275. CHAPTER II. DESCRIPTIVE GEOLOGY, " Shall we like schoolboys think that stones were made Only to cast at varlets." THE LITHIAD. SECTION I. GENERAL RESULTS ARRIVED AT BY A LITHOLOGICAL EXAMINATION OF ROCKS. WE showed in the last chapter how, with the growth of Geology, there sprang up two main subdivisions of the science, to which the names of Descriptive and Historical Geology may be given. Descriptive Geology or Petrography. The first of these, which corresponds nearly with Petrography, merely tells us what that part of the earth which is open to investigation is made of ; and is nothing more than a description and classification of the substances that make up the earth's crust. Historical Geology or Geogenie. But as men passed from examining, dissecting, and analyzing specimens of rocks indoors, to the larger views which an outdoor study of those rocks on a large scale affords, they came to see that there was every reason to believe that the crust of the earth had not been always such as it is now, but that differ- ent parts of it had been built up at different times and in different ways out of pre-existing materials. Hence arose a second great sub- division of the science, which aims at answering the questions about the origin of rocks which extended study suggests. This branch of Geology, which is pretty much the same as the Geogenie of the Germans, strives to determine first of all how the different members of the earth's crust were formed ; secondly, the order in which they were formed, and the changes in life and other events which accompanied their formation ; and, thirdly, tries to feel its way back to those dark and distant ages, when the present crust of the earth had not yet come into existence, and to form reasonable conjectures as to what the earth was like under these conditions, and as to the steps it passed through in its progress from them to its present state. Under this subdivision we shall also have to inquire into the methods by which the rocks have been brought into the positions in which we now find them, and the way in which the surface of the ground has had its present shape given to it. Subdivisions of Descriptive Geology. Descriptive Geology UNIVERSITY Lithology. consists of two parts, which may be called Lithology and Petrology (Trer/aa, rock). Lithology. Lithology describes the results which would be arrived at by a man who sat indoors in his laboratory, and examined small hand- specimens of different kinds of rock brought to him. Petrology. Petrology tells us what additional information we gain when we go out of doors and examine large masses of rock in the field. The first of these gives us accurate information as to the composi- tion and minute structure of rocks, but, except in a comparatively small number of instances, it is questionable whether purely lithological examination would suggest the notion of rocks ever having been in any way different from what they are now, and prompt questions as to their origin. Even in these few exceptional cases lithology alone would not go far towards furnishing an explanation of the way in which rocks came into existence. Petrology, however, though it belongs in part to Descriptive Geology, lands us on the threshold of the His- torical subdivision. The facts learned from it, many of them such as no amount of indoor examination of hand specimens would ever have taught us, are decidedly suggestive of the notion that rocks have not always been such as we see them now, but have been produced at various times by various causes ; and it is mainly from a knowledge of these facts that our speculations as to the probable methods by w r hich the rocks have been formed, take their rise. What we have to say under the head of Lithology will, perhaps, be better understood if we first take one or two actual instances of the lithological examination of rocks. Instances of the Lithological Examination of Rocks. We have here a bit of the rock called Granite, and have formed a clean- cut face upon it with the hammer. A very little care will show that the rock is made up of different substances, three of which we soon learn to distinguish and recognise. One is of a dull white or pinkish colour, and it breaks readily along a number of smooth parallel sur- faces, which have a pearly lustre ; the second is more like glass to look at, and breaks with the same uneven fracture as glass ; the third consists of thin plates of a dark colour, with glistening faces, and can be readily split parallel to these faces with a knife. These three substances go by the names of Felspar, Quartz, and Mica respectively. They are bound together into a solid rock by a cement or paste ; but to make out the nature of this would require more skill than we are supposed at present to possess. Here again is another rock, which we soon see is composed of rounded grains of the same substance, Quartz, as we found in Granite ; we touch it with a little dilute acid, and it effervesces briskly; no such effervescence takes place when we treat Granite in the same way. In this rock, then, we conclude that there is something present which does not enter into the composition of Granite. Break the specimen open. There is in it a hollow space lined with glistening crystals, which, when touched with acid, effervesce even more strongly than the body of the rock itself, and are therefore probably formed of the other substance which goes to make up the rock. This substance is not io Geology. unlike Felspar at the first glance, but we soon learn to distinguish between them : try to scratch both with the point of a knife, and Felspar is much the harder ; Felspar does not effervesce with acids ; lastly, while Felspar splits readily in only one, or at most in two directions, this substance splits with the greatest ease along three sets of smooth, shining planes, and can be readily broken up into a number of pieces, each of which has exactly the same geometrical form. Our new acquaintance is called Carbonate of Lime, and the rock we are examining consists of grains of Quartz bound together by Carbonate of Lime. It is called a Calcareous Sandstone. Now with regard to these substances, which we have found to form the component parts of the two rocks just mentioned, one point is important to notice. Though different parts of the specimen we first handled, differ slightly from one another in composition, some con- taining more Felspar than anything else, and some having a more plentiful supply of Mica than others, yet bits of the three substances of which the rock is made up possess the same distinctive characters from whatever part of the specimen w r e pick them, and, if we subject them to chemical analysis, we shall find them to be approximately invariable in composition. The same would be true of the second rock even to a greater degree ; some parts of the rock would be richer in Carbonate of Lime than others, but every one of the bits of clean Carbonate of Lime that could be separated would be found to be absolutely identical in every respect, no matter what part of the rock it came from. We have therefore found, in the case of the two rocks that we have been examining, that they are composed of certain substances, each of which has always pretty much the same appearance, breaks always in the same way, and keeps always the same hardness and chemical composition.* Hardness, way of breaking, and many other properties of a like nature are usually classed together under the general head of " physical characters." Further examination of the Quartz, Felspar, Mica, and Carbonate of Lime would show that each possesses, besides the physical characters mentioned, others more or less peculiar to itself ; for instance, when light passes through them, each affects it, but each affects it in a different way. These bodies, then, may be shortly described as having each a constant chemical composition and constant physical characters; other distinctive points common to all are that they have been formed by natural agencies, but not by the agency of living beings. These, and many other bodies of which the same is true, are called MINERALS, and they may be formally defined as follows : Definition of a Mineral. Minerals are bodies which have a constant chemical composition and constant physical characters, and have been formed naturally, but not by the agency of either animals or plants. There are many kinds of rocks from which we can pick out the * It will be. shown presently that these statements are in many cases not strictly true, but they may be looked upon as correct enough for our present purpose. Mineralogy. 1 1 minerals as easily as from our Granite or Sandstone; others there are the constituent minerals of which are not easily recognised ; and yet others which require very refined methods to determine what they are made of. But no matter what the rock be, it can be shown in some way or another to be made up of one or more of these substances to which the general term " mineral " is applied : it is only in com- paratively few cases that a rock consists of a single mineral, by far the larger number contain two or more. We can separate, then, rocks into minerals, and the chemist can break up minerals into their elements, but, as has been already pointed out, there is this important difference between rocks and minerals. Minerals are chemical compounds, and the elements of which they are composed are present in proportions which are always the same for the same mineral ; Quartz, for instance, is composed of two elements, Oxygen and Silicon, and every bit of Quartz contains 53*3 per cent, of Oxygen, and 46 '6 per cent, of Silicon. But if we take different specimens of the same rock, that is, of a rock composed of the same minerals, we shall not find that the percentage of any one of these minerals is the same for all the specimens : some Granites, for instance, contain a much larger percentage of Quartz than others. Further, the elements of a mineral are chemically combined, and the mineral differs totally in physical characters from its component elements ; in a rock, however, the minerals are mechanically mixed together, and each mineral preserves its individuality and distinctive characters. Bearing all these facts in mind, we arrive at the following definition of a rock : Definition of a Rock. Rocks are mechanical mixtures of minerals. We may here mention that geologists include under the name Eock all the substances, hard and soft alike, which go to make up the earth's crust ; clay or loose sand are as much rocks, in the geological sense of the word, as slate or sandstone. SECTION II. MINERALOGY. The first thing which the lithologist has to do .is to make himself acquainted with the minerals which enter into the composition of rocks; and the branch of Geology which teaches this is called Mineralogy. Number of Rock-forming Minerals. The total number of known minerals is very large indeed, but of these only a comparatively small number enter to any appreciable extent into the composition of the earth's crust, and with these only is it absolutely necessary that the geologist should make himself acquainted. Thus suppose, for instance, that we examined, in the way just described, a large number of specimens of Granite, we might find them all in the main made up of the three minerals, Quartz, Felspar, and Mica ; but besides these three predominating components, other minerals would frequently present themselves. In many cases, for example, we should detect, in addition to the three minerals just mentioned, long needle-shaped 1 2 Geology. prisms of a black mineral known as Schorl. But the presence of Schorl would not prevent us from calling the rock a Granite, and look- ing upon it as allied to other Granites which were free from Schorl, or which contained other minerals in addition to the three normal consti- tuents. We might use Granite in a generic sense, and look upon the schorlaceous and other forms of the rock as different species of that genus ; or, what would perhaps be better, we might consider these last as merely varieties of the species Granite. Minerals, such as the Schorl just mentioned, which are, as it were, superadded to the normal constituents of any rock, are called accessory or adventitious ; the advanced student will find that a study of them often throws light on the origin of rocks, but a knowledge of them is not necessary for the beginner. Again, the metallic ores, occurring as they do chiefly in narrow veins and threads, cannot be looked upon, except in some few instances, as rock-forming minerals. The mastery of these, then, the beginner may postpone till he has made some advance in his studies. Besides these two classes there are a host of minerals seldom met with, and some so rare that no one but a mineralogist, who has the most extensive opportunities of research, has ever a chance of even seeing them ; with many of these we may almost say that the geologist has nothing whatever to do ; the beginner certainly need not trouble his head about any of them. When we put aside the minerals which occur only as accessory constituents of rocks, the metallic ores, and the rare forms, the- number remaining will not be large ; and it is of these only that we need treat in an elementary work. But before we can give an account of these we must describe, as far as space will allow, the characters most important to note in minerals generally. Chemical Composition of Rock-forming Minerals. The elements known at present are sixty-four in number, but of these not more than twenty at the outside enter to any extent into the composition of the earth's crust. The names of these twenty and the symbols used by chemists to denote each are given in the table below. The ten placed in the first subdivision have been estimated to make up as much as 977-1 OOOths of the earth's crust. The rocks composed of them are such as Sandstone, Limestone, and Granite, and are so common that they may be said to be met with everywhere. Some of the remaining elements enter into the composition of rocks which are not so universally distributed, but which, where they do occur, are of considerable importance ; Rock-salt may be taken as an instance. Other elements of the second division form minerals, such as Fluor- spar, which are very common, but can hardly be said to be essential components of any rock : others enter into the composition of minerals, such as Apatite, which occur very frequently in rocks, but only in comparatively small quantity. Some, again, of the less widely diffused elements replace occasionally in small quantity a part of the essential constituents of common rock-forming minerals. Mineralogy. LIST OF THE MAIN ELEMENTARY CONSTITUENTS OF THE EARTH'S CRUST. Oxygen Hydrogen Silicon Carbon Sulphur Chlorine Fluorine Phosphorus Boron O H Si c s Cl F P B Aluminium Potassium Sodium Calcium Magnesium Iron Lithium Barium Manganese Titanium Zirconium Al K Na Ca Mg Fe Li Ba Mn Ti Zr Of these elements two only, Sulphur and Carbon, occur uncombined either as minerals or in a state of approximate purity as rock masses. All the other elements are found in combinations, of which the following are the most important : 1. Silica, Silicon Dioxide, SiO. 2 . By far the most widely diffused of rock-forrning substances, nearly one quarter of the earth's crust being composed of it. It occurs both uncombined in several minerals, and in the following compounds : Iron. Aluminium Sesquioxide, Al.jO 3 or Alumina. Potassium Monoxide, K 2 O or Potash. Silicates of Sodium Monoxide, Na. 2 O or Soda. Calcium Monoxide, CaO or Lime. Magnesium Monoxide, MgO or Magnesia. Zirconium Dioxide, ZrO 2 or Zirconia. The rocks known as Crystalline are almost wholly made up of minerals composed of these silicates. Lithium, Boron, and Fluorine occur in some silicates. 2. Calcite, Carbonate of Lime or Calcium Carbonate, CaCO 3 . Hitter Spar, Double Carbonate of Lime and Magnesia or Calcium Magnesium Carbonate, CaCO 3 .MgC0 3 , or (CaiMg^Og. These minerals are the main constituents of Limestones and Mag- uesian Limestones. 3. Water, H. 2 O. When minerals contain water in a state of chemical combination they are said to be hydrated ; when no water enters chemically into their composition they are described as anhydrous. The student must carefully distinguish between the chemically-combined water of the component minerals and the water which is present very generally in the pores and crevices of a rock. The latter, often called Quarry-damp, passes away gradually by evaporation when the rock is placed in dry air ; chemically-combined water is set free only when the minerals are partially or wholly decomposed by the action of strong heat or other methods. 1 4 Geology. 4. Spathic Iron Ore, Carbonate of Iron or Ferrous Carbonate, FeC0 3 . Specular Iron Ore, Sesquioxide of Iron or Ferric Oxide, Fe 2 O 3 . Limonite and other* hydrated Sesquioxides of Iron or Ferric Hydrates, Fe 2 O 3 + Water. Ilmenite, Titaniferous Ferric Oxide, (Fe:Ti) 2 O 3 . Magnetite, Magnetic Oxide of Iron or Ferrosoferric Oxide, Fe 3 O 4 . Rutile, Dioxide of Titanium or Titanic Oxide, Ti0 2 . Compounds of Iron occur only exceptionally as rock masses. Specular Iron Ore, Magnetite, Ilmenite, and Eutile are often present, but in comparatively small quantity, in crystalline rocks. But it is as colouring matters that iron compounds play their most important part ; the results they produce will be explained further on (see p. 154). 5. Carbonaceous substances, composed of Carbon, Oxygen, and Hydrogen, such as Coal. 6. Four, perhaps five, of the following may be said to play the part occasionally of rock-forming minerals, and the remainder are common enough to deserve mention here : (a) Common Salt, Sodium Chloride, NaCl. (b) Iron Pyrites, Bisulphide of Iron, FeS 2 . (c) Gypsum, Hydrated Calcium Sulphate, "CaS0 4 + 2H 2 O. (d) Anhydrite, Calcium Sulphate, CaSO 4 . (e) Apatite, Calcium Phosphate with Calcium Fluoride and Chloride, 3Ca 3 P 2 O 8 Ca(F:Cl) 2 . (/) Fluor-spar, Calcium Fluoride, CaF 2 . (g) Heavy Spar, Barium Sulphate, BaSO 4 . (A) Pyrolusite, Manganese Peroxide, MnO 2 . (t) Psilomelane, Mn02MnO 2 H 2 O. (a),(c), and (d) occur as rocks,but not rocks of the widely-diffused class. (b) is one of the very commonest minerals : some few cases are known where it occurs in quantities large enough to deserve the name of rock masses. (e) is very common, but in comparatively small quantity, in crystalline rocks. (/) and (g) are very common minerals. (h) and (i) occur principally as metallic ores ; they are mentioned here because thin films of them very frequently coat the walls of cracks in rocks. These films have often an outline like that of a sprig of moss or seaweed, and are then said to be Dendritic. Means of recognising Minerals. These, then, are the minerals with which we shall have to make ourselves acquainted if we wish to understand the composition of the rocks of the earth's crust, and the first question with the student will be, How am I to know one mineral from another 1 and when I come across any mineral, how can I tell what it is 1 There is one method which, with a few exceptions,t will enable us * For other Hydrates of Iron, see Prof. Brush, Silliman's Journ. , 2nd ser. xliv. 219. t The exceptions are in the case of some minerals which have the same chemical composition but differ in crystalline form and other physical characters. They will be treated of under Polymorphism (p. 79). Mineralogy. 1 5 to say without fail what any given mineral is. Each mineral has a definite chemical composition, and if we analyse the specimen we wish to determine, we shall be able in nearly all cases to assign to it its proper name. But quantitative analysis is a tedious and somewhat difficult operation, and it will evidently be a great boon if some quicker and easier way of recognising minerals can be devised. Fortunately in many cases minerals can be determined without the necessity of having recourse to analyses. Besides having a definite chemical composition, minerals have definite physical characters, and in many cases these characters are so marked and distinctive that we can by their aid alone decide what a mineral is. The determination of minerals by this method we may call the art of knowing minerals by sight. It is an art which the student must acquire ; and it is most important to bear in mind that it cannot be learned from books alone ; constant handling of actual specimens and a comparison of them with descriptions can alone impart the necessary familiarity with the look and characters of minerals. The following are the principal physical characters which enable us to recognise minerals : 1. The shape they take. Under this head what is known as Crystalline Form is most important. Crystallography, or the description of Crystals, looks at first sight a somewhat alarming subject, and it does present certain difficulties to a beginner, especially if he. has not much previous knowledge of Geometry. These difficulties can, how- ever, be easily got over by the aid of a little patience and close attention, and then the subject will be found neither hard nor uninviting. A short sketch of the elements of Crystallography will be shortly given, which we will endeavour to make full enough to meet the needs of those who wish to have merely a general acquaintance with the subject, and which, in the case of those who aim at a more thorough knowledge, will, it is hoped, serve to smooth over the difficulties which beset the path of the beginner, and clear his way for entering on a study of works specially devoted to the subject. 2. The ivay in which minerals break, under which head Crystalline Cleavage is the most important item, and then relative Tenacity. 3. The behaviour of minerals before the Blowpipe. The blowpipe characters of all the minerals described will be given, but for the method of using the instrument reference must be made to works specially devoted to the subject. A knowledge of the more important blowpipe reactions is most essential. 4. Colour, Streak, and Lustre. 5. Relative Hardness, and Feel or Touch. 6. Specific Gravity. 7. Taste and Smell. 8. Properties connected with Light, Heat, Electricity, and Magnetism. These we shall be able to do no more than very shortly call attention to. Under the first head Crystalline Shape is the most important, and we will begin our account of the various shapes assumed by minerals with a sketch of the element of Crystallography. Definition of a Crystal. Crystals are solid bodies, of more or 1 6 Geology. less regular shape, bounded by plane surfaces, which are often polished and glistening.* The planes that bound crystals are called their "faces ;"f the inter- section of two adjacent faces is called an " edge ;" a point where three or more faces meet is called an "angle ;" the inclination of two faces to one another is called " an interfacial angle," or " the angle of the edge " in which those faces intersect. The faces of some crystals are curved. Simple Crystalline Forms. The following are some of the simplest shapes assumed by Crystals : 1. PRISMATIC SHAPES. If we take any two plane rectilineal figures alike in all respects, place them so that each side of the one shall be parallel to the corresponding side of the other, and join the correspond- ing angles by straight lines, the solid so enclosed is a Prism. The two plane figures are called the "ends" or " terminal faces ;" the other faces, which it is easy to see are parallelograms, are called the "lateral faces," and their intersections are the "lateral edges." The line joining the centres of the ends is called the " longitudinal axis ;" a plane parallel to the ends through the middle point of the longitudinal axis is called the " base." Prisms are classed according to the shape of their ends ; a prism with square ends is a square prism, one with hexagonal ends an hexagonal prism. Further, if the longitudinal axis is perpendicular to the ends, the prism is a "right prism;" if not, an "oblique prism." The most important right prismatic shapes are The Cube, bounded by six equal squares (fig. 21, a). The Right Square Prism, ends squares, lateral faces rectangles. The Right Rectangular Prisms, ends rectangles. The Right Rhombic Prism, ends rhombuses. The Right Rhomboidal Prism, ends rhomboids. In the two last solids the longer diagonal of the base is called the "macro-diagonal," the shorter the "brachy-diagonal." Fig. 2 shows the general aspect of these prisms, which differ only in ^ ^ the shape of the ends. The Right Hexagonal Prism, ends regular hexagons (fig. 70). The Right Dihexagonal Prism, differs from the last in having equilateral dodecagons for its ends. Among oblique prisms the following are the most im- portant : The Monoclinic Rhombic, Prism (fig. 17), ends rhoni- Fi 2 buses. The longitudinal axis is perpendicular to one of the diagonals BB' of the ends, which is called the " Ortho-diagonal," and oblique to the other A A' which is called the " Clino-diagonal. * This definition will serve for our present purpose. We shall see shortly that it involves much more than is here stated. + Never speak of the "side" of a crystal. The lines which bound plane figures are their sides ; solids are bounded by surfaces, which in the case of most crystals, are plane surfaces ; in the case of a solid always speak of "faces" and "edges." Mineralogy. The Triclinic Rhomboidal Prism, ends rhomboids, longitudinal axis oblique to both diagonals of the base. The Rhombohedron, bounded by six equal and similar rhombuses. Figs. 20a, 66, and 67. 2. PYRAMIDAL FORMS. If from a point above a plane rectilineal figure we draw straight lines to each of its angles, the solid so formed is the Pyramid of Geometry. The Crystallographic Pyramid consists of two such pyramids, alike in every respect, placed base to base in corresponding positions. The plane figure is called the "base;" its sides are the "lateral or basal edges;" the point is called the "vertex" or "pole;" the edges which pass through it the "longitudinal, or vertical, or polar edges." The line joining the two vertices is the " longitudinal or vertical axis ;" if it is perpendicular to the base, the Pyramid is a " Right Pyramid ;" if not, an " Oblique Pyramid." Many of the Pyramids that occur in Crystallography have a four- sided base, and being bounded by eight triangles are called Octa- hedrons. The following are the most important : The Regular Octahedron, bounded by eight equilateral triangles (A A' BB' CO' fig. 22). The Square Octahedron, square base, bounded by eight equal isosceles triangles (fig. 18). The Right Rhombic Octahedron, base a rhombus (A A' BB' CO' fig. 22, if ABA'B' be a rhombus). The Monoclinic Octahedron, base a rhombus, longitudinal axis perpen- dicular to one and oblique to the other diagonal of the base (fig. 3, if ABB' A' be a rhombus, 00 perpen- dicular to AA and oblique to BB'). The Triclinic Octahedron, base a rhomboid, longitudinal axis oblique to both diagonals of the base (fig. 3, if ABB' A be a rhomboid, 00 oblique to both AA and BB'). The last three are bounded by eight scalene triangles. Fig. 3. We have also the Right Hexagonal Pyramid, base a regular hexa- gon. The ends of the crystal in fig. 5 are Right Hexagonal Pyramids. The Right Dihexagonal Pyramid (fig. 38), base an equilateral dode- cagon. 3. SCALENOHEDRONS. The Hexagonal Scalenohedron, bounded by twelve equal and similar scalene triangles (fig. 40); the Tetragonal Scalenohedron, bounded by eight equal and similar scalene triangles (fig. 43). 4. TETRAHEDRONS, bounded by four equal and similar triangles (fig. 22a). In the Regular Tetrahedron the triangles are equilateral; in the Geology. Tetragonal Tetrahedron or Sphenoid they are isosceles ; in the Rhombic Tetrahedron or Sphenoid they are scalene. 5. DODECAHEDRONS. Rhombic Dodecahedron, bounded by twelve equal and similar rhombuses (fig. 24); Pen- tagonal Dodecahedron (fig. 29), bounded by twelve equal and similar pentagons. 6. DELTOHEDRONS and TRAPEZOHEDRONS. >c A Deltoid is a four-sided figure ABCD (fig. 4), in which AB=AC and DB=DC, but all the sides are not equal. A deltohedron is bounded by twelve or twenty-four equal and similar deltoids (see fig. 31). We have also the Tetragonal Trapezohedron, bounded by eight equal and similar deltoids (fig. 44) ; and the Hexagonal Trapezohedron, bounded by twelve equal and similar deltoids. How the crystalline shape of a Mineral helps us to recognise it. Fig. 5 shows a shape very often assumed by the mineral quartz. The middle part of the figure is a prism, and if we cut it across by a plane perpendicular to the vertical edges, the section is a regular hexagon each of whose angles is 1 20 (fig. 5a) ; the middle part of the Fig. 5. Fig. 5. crystal is therefore a Right Hexagonal Prism. We can easily find out in the same way that the ends are Right Hexagonal Pyramids. The angle between each pair of adjacent faces of the pyramid is 133 44'; each face of the pyramid makes with the adjacent face of the prism an angle of 141 47'. Now take any number you like of crystals of quartz having the above form, measure the inclinations of these planes to one another, and they will come out exactly the same in every case. What is more, though other minerals crystallize in hexagonal prisms capped by hexagonal pyramids, these angles are in none of them the same as in Quartz. Each mineral has a set of angles Mineralogy. of its own, and the angles are the same in the same mineral and different in different minerals. Clearly then if we knew these angles for any mineral, and find a crystal which possesses these angles, we may safely say that that crystal is formed of the mineral in question. The exact measurement of the angles of a crystal requires great prac- tice and skill ; but in the case of the common minerals the eye becomes by use sufficiently familiar with the shapes they usually assume to enable one to make at least a likely guess as to the mineral from its crystalline shape alone ; other tests are then used to confirm this approximate determination. The crystal in fig. 5 is very regular in shape ; all the prismatic faces are equal rectangles, and all the pyramidal faces are equal triangles. Such a crystal may be called an Ideal or Model Crystal. Natural Crystals seldom show this perfect regularity ; usually some of the faces are extended at the expense of the others. Thus in fig. 6 the crystal, though it closely re- sembles the ideal shape in fig. 5, differs somewhat from it. The face B' is extended and the faces A and C are narrowed, the faces A, B, and C are no longer rect- angles, and the faces a and b are no longer triangles. A cross sec- tion of the prism is given in fig. 6a. Now note that, though the shape is somewhat altered, the angles remain unchanged : b is still inclined to c and a at angles of 133 44', and to B' at an angle of 141 47'; and though the cross section (fig. Qa) is no longer a regular hexagon, the angles be- tween B' and C and between B' and A are still 120. We should have no hesitation therefore in calling this a crystal of Quartz, in spite of its deviation from the ideal form. Quartz crystallizes in many other shapes, some far more complicated than the two just given, but all can be shown to be closely related to the ideal form in fig. 5. For instance, when the faces are so numerous that no resemblance can at first sight be detected to a Right Hexagonal Pyramid, we can often pick out six faces which do not actually intersect, but which, if they were produced till they did intersect, would enclose a solid of that shape with each pair of adjacent faces inclined tp one another at an angle of 133 44'. Again, the crystal outlined by strong lines in fig. 7 does not bear much resemblance to a regular octahedron. Two of its faces are Fig. 6. 20 Geology. hexagons, and the other six four-sided figures. But it is found, when the angles are measured, that the angles between adjacent faces are exactly those of a regular octahedron. Its relationship to a regular octahedron may be further shown thus. The faces O ly 4 , O s intersect in a point B ; 0^ and the faces opposite to (9 4 and 8 intersect in a point A, and by taking the faces all round three and three, we determine by their intersections the points A', B', C, and C". When these points are joined a regular octahedron is formed, six of whose faces coincide with 4 , O s , 5 , and the faces opposite to these, while the other two are parallel to 1 and the face opposite to O r In fact, if we start with a regular octa- hedron, caca'b' in which a'b is equal to the breadth of the face 4 , and if we suppose the face abc and the face opposite to it to be prolonged right and left, and if then we keep adding plates on to the other faces of the octahe- dron, each plate being bounded by the prolongations of abc and the opposite face, we shall ob- tain the solid bounded by the strong lines. It is an octahedron which has grown by additions to six of its faces, while no additions have been made on the remaining two. Again, in this solid some of the faces may grow faster than others ; we should then obtain a solid whose faces would no longer intersect in the angles of a regular octa- hedron, but whose faces would be parallel to those of a regular octahedron. In either case we look upon the solid as a regular octa- hedron with its faces unequally developed. In the study of Crystallography it is easiest to begin with regular ideal shapes, and in the introductory sketch here given these only will be for the most part considered. The law of which we have just given illustrations may be thus enunciated : Substances having the same chemical composition crystallize in shapes the corresponding interfacial angles of which are always the same. Some exceptions to this law will be noticed further on. Methods of denoting shortly the Faces of a Crystal. One's feeling on examining for the first time a collection of crystals is one of utter bewilderment. The number of different shapes is so large, and the faces by which they are bounded are so numerous, that it seems quite hopeless to attempt to convey by words any idea of even the general aspect of the crystals. And a verbal description would be so long and unutterably complex that it would be worse than valueless for this purpose. But short symbolical notations have been devised by which each face can be denoted so concisely and in so lucid and Mineralogy. 21 unmistakable a way, that any one familiar with the symbolical language used can picture to himself from a row of apparently unmeaning figures or letters the relative positions of the bounding faces as clearly as if he had before him a careful drawing of the crystal, and almost as clearly as if the crystal itself lay in his hand. Invaluable however as such a notation is to one who understands it, it must be confessed that it wears a most uninviting look to one who does not ; and many people are deterred from a study of Crystallography by the fear that a symbolical language, apparently so hopelessly unintelligible as that used in this science, must be very hard to master. That this is not so we hope now to be able to show. Various forms of symbolical notation have been devised by different authors ; the following is the easiest to understand and the one best suited for beginners. Let (fig. 8) be a point with- in a crystal; Ox, Oy, Oz, three straight lines, not all in the same plane, meeting in 0. Let any face of the crystal, produced if necessary, cut Ox, Oy, Oz in A, B, C. Then if we know the lengths of OA, OB, 00, we shall know the position of the face in ques- tion. Ox, Oy, Oz are called "Co- ordinate Axes." The planes xOy, yOz, zOx, each of which contains two of the axes, are called "Axial Planes." The axial planes are usually selected so as to lie parallel to three Fig. 8. faces of the crystal. Let OA = a, OB = b, OC=c, then a, b, c are called the "Intercepts" of the face on the axes. The first point to notice is that we do not require for Crystallo- graphic purposes to know the actual length of the intercepts of the faces of a crystal ; it will be sufficient if we know the ratio of two of them to the third, i.e. if we know the numerical values of ^ and ^. This will appear from the following considerations. Suppose that we were told the actual lengths of the intercepts of the different faces of any crystal on a given set of axes, and the order in which the faces occur. Then by measuring off the intercepts of any face (A) on the axes we could determine the position .of that face ; in the same way we could determine the position of each of the adjacent faces. This would enable us to fix the positions and lengths of the edges in which the face (A) is cut by the adjacent faces, that is, it would tell us the shape and size of the face (A). In the same way we could determine the shape and size of every face, and could make a model the faces 22 Geology. of which would have exactly the shape and size of the faces of the crystal. But this would be more than the crystallographer wants. In Crystallography we do not care about the shape and size of the faces, for we have seen that different crystals of the same substance have often their corresponding faces of very different shapes and sizes (p. 19). What we do care about are the angles which the faces make with one another, for these are constant for the same substance. Now it so happens that the angle between two faces does not depend on the actual lengths of their intercepts. This is a point on which we can easily satisfy ourselves. Let P and P be two parallel faces ; Q and Q' two faces parallel to one another but inclined to P. P, P', Q, Q', may all have different values of a, b, c, but in spite of this the angle between P and Q will be the same as the angle between P' and Q'. Clearly, then, the angle between two faces does not depend on the actual lengths of a, b, and c; what it does depend upon is the ratio of a b c to b, and a to c, or the value of the fractions ^ and ^ ; for if we know the values of these fractions for two faces, we can calculate the size of the angle between those faces. * In short, Crystallography is not concerned with the linear dimen- sions and shapes of the faces of crystals, but solely with the angles between the faces ; and this is the same thing as saying that it requires to know not the actual values of the intercepts of the faces, but the ratios which any two of these intercepts bear to the third. All then we need know about a face is that its intercepts on the axes are in the ratio of a : b : c ; -, b c or * : a ' a > and such a face may be very shortly denoted by the symbol M ^\ It will be necessary to agree upon some rule which shall tell us to which of the three axes each of the numbers in the bracket refers, and the following is the convention we will adopt : the number placed first is proportional to the intercept on Ox, that placed second to the intercept on Oy, and that placed third to the intercept on Oz. To make this quite clear let us take one or two instances. In the face (3, 1, 2) the index 1 refers to Oy, I therefore measure off any distance OB on that axis; the index 3 refers to Ox, I therefore measure off on that axis OA = 30B ; the index 2 refers to Oz, I therefore measure off along Oz, OC=20B. Then the face in question is ABO. Take again the face (5, 4, 3): here if we make OA = 5, OB = 3, 0(7=4, the face is ABC. Or we may proceed thus : since 5 4 5 = 3 4 = 3 3131 * This will be evident if the student has a little knowledge of Solid Geometry. If he has not, he must take the statement on faith. Mineralogy. 23 the symbol may be written (|, ^, 1J- Take any length 00' on Oz, make OA' on 0^ = kO(7', and OB' on Oy= f> OG'\ then the face is It is easy to see that ABC and A'B'C' are parallel to one another, and they will therefore make the same angle with any other face of the crystal, and hence, since this angle is all we care about, it does not matter which of the two methods we adopt. In the same way, if the intercepts of any other face are in the ratio of a' : V : c f , we might denote it by the symbol (l ^'a')' , ., b' be' c and if -,= 01-, -, =n- this symbol might be written f I, m~, n~\ If we adopt this last form we can still further abbreviate our symbols. Let us take one face, say the face (l - ^J for a standard to which all other faces are referred. Then the symbols for other faces will be T ,b ,c\ 1 m a n a) i ,,b ,,c\ lm a n a) and so on. Since - and - are the same in all the symbols, we may omit them, and write the symbol for any face (1 m n). But we must always carefully bear in mind, if we use symbols thus abbreviated, that m does not stand for the number m, but for m times -, and n stands in like manner for n times - r When we have agreed which face shall be taken as a standard face in any crystal, the numbers 1 - ^ are called the " Parameters " of that crystal, and the standard face is called the Parametral Plane ; 1 m n are called the " Indices " of the face to which they belong. It will be seen shortly that this method has certain advantages over that which would denote every face by an independent symbol. In symbols of the form (1 m n) we still agree that the number 24 Geology. placed first shall refer to Ox, that placed second to Oy, and that placed third to Oz. One numerical example may be given by way of illustration. In a crystal of Heavy Spar | = 1'23 and | = 1-61 for the face selected as the standard face. The symbol of another face is (2, 4, 1). The intercepts of this face on the axes are proportional to 9 a t 4x1-23, 1-61; or to 9 r* 4-92, 1-61; or to 9 *> 5, 1*6 nearly; or to 1, 2-5, 8 nearly. Therefore if I take any length OA on Ox, measure along Oy OB = '2~50A, and along Oz OC='SOA, the face in question is parallel to ABC. If a plane is parallel to one of the axes, its intercept on that axis becomes infinitely great and the index corresponding to that axis becomes infinite. The symbol oo or i is used to denote infinity. Thus the plane (1 oo n) is parallel to Oy; the plane (1 GO oo) is parallel both to Oy and Oz, that is, it is parallel to the plane yOz. Suppose that in fig. 8 the axes are produced backwards to x'y'z, and that the axial planes are extended in these directions. The axial planes now evidently divide the space round into eight com- partments or octants. The plane shown in fig. 6 lies in the octant zOxy. We must have some way of denoting the planes which lie in the other seven octants. The following convention enables us to do this : any line measured along Ox, Oy', or Oz is reckoned negative, and if its length is m, is denoted by m. Usually zOz is placed vertically, and lines measured downwards from are reckoned negative ; xOx' runs from right to left, and lines measured to the left of are reckoned negative ; yOy' runs from front to back, and lines measured backwards from are reckoned negative. Take the case of a plane in the octant zOx'y, it cuts x'Ox on the left of 0, therefore its intercept on that axis is negative, arid its symbol will be of the form (I m n). In the same way a plane in the octant y'Oxz has its intercepts on yy' and zz' negative, and its symbol will be of the form (1 m n) ; a plane in the octant x'Ot/'z has all its intercepts negative, and its symbol will be of the form (I m n). It is convenient to number the octants in the following order : Upper four octants, zOxy 1st, zOxy 2nd, zOxy 3rd, zOxy 4th. Lower four octants, zOxy 5th, zOxy 6th, zOxy 7th, zOxy 8th. In each case, if a watch were laid on the plane yOx, we count round in the direction opposite to that in which the hands revolve. It is of course impossible, in the case of an actual crystal, to measure the lengths of the intercepts of the faces, and so determine their in- dices ; but the angles between the faces can be measured, and when these angles are known, the parameters and the indices can be calcu- Vv v w Mineralogy- 25 lated by mathematical formulae. A knowledge of the indices not only enables us to describe the faces concisely, but is also useful in drawing figures of crystals (see pp. 44 and 48). The indices of the faces of a very large number of crystals have been thus determined, and in scarcely any case have they been found to be large. The highest index seldom exceeds 6. Also in every crystal yet examined the indices are what are called rational quantities ;* that is, it is always possible to express their ratio to one another exactly by two whole numbers. This is called the Law of the Rationality of the Indices. Zones. It frequently happens that a number of faces of a crystal intersect, or would intersect if they were produced till they meet, in parallel straight lines. Such a group of faces are said to form a zone, or to lie in the same zone, or to be tautozonal. A plane perpendicular to the parallel intersections is called a zone plane, and a straight line perpendicular to it a zone axis. All the faces in a zone are evidently perpendicular to the zone plane. In the crystal of quartz (fig. 5) the faces of the hexagonal prism are tautozonal. , If a cube be placed on a table, the four vertical faces are in the same zone; the top and bottom, faces, together with each pair of opposite vertical faces, form two more zones. A recognition of zones often helps us to detect symmetry in crystals, and in the higher branches of Crystallography zones are of the utmost service. Symmetry of Crystals. If the task of devising a short and easy method of denoting the relative position of the faces of a com- plicated crystal looks at first utterly hopeless, any attempt to discover a law or an order according to which they were arranged would seem for a while an endeavour still more vain. Gradually however light would dawn on the patient student. He would notice perhaps two faces on opposite sides of the crystal which seemed in a manner to correspond with one another ; other corresponding pairs would reveal themselves ; and at last he would realize that if the crystal was cut in two by a certain plane, the two portions, to say the least, resembled one another very closely. He would have discovered a plane on opposite sides of which the faces were similarly grouped. Such a plane is called a Plane of Symmetry ; but before saying anything more of such planes it will be desirable to illustrate by a few examples the meaning of the term Symmetry. In fig. 9, ABC is an isosceles triangle, Aa perpendicular to EC. Aa divides the triangle into two halves, which in shape, size, and position are the exact counterpart of one another. If we cut out the triangle in thin cardboard, cut the card half through along A a, and * If we try to determine the ratio of the circumference of a circle to its diameter, we shall not be able to find two whole numbers to express it : f^ is near to it, f ^^ nearer, f^T^ still nearer, xiririTt $ nearer still. And so we may go on, but we shall never find two whole numbers which exactly express the ratio. Such a ratio is called irrational. 26 Geology. then turn ABa over round Aa till it comes again into the plane of the A card, it will exactly cover AC a. If we cut out ABa, and hold it against a vertical looking- glass, with its plane horizontal and Aa touch- ing the glass, ABa and its reflection will exactly reproduce the triangle ABC. This last test is the same as saying that if from a point P in AB we draw a perpendicular PD to Aa and produce it to Q making DQ equal to DP, then Q lies on AC, and that this is true for every point on AB. These are only different ways of stating the same fact, and that fact is more shortly expressed by saying that the triangle is Fig. 9. symmetrical about the line Aa. We will now consider a few cases of symmetry in plane figures. A circle is symmetrical about every diameter, there are therefore an infinite number of lines which divide it symmetrically, and it possesses the highest possible degree of symmetry. A regular hexagon, ABCDEF (fig. 10), is symmetrical about each of the lines AD, BE, CF, which join opposite angles, and also about the lines ad, be, cf, which join the middle points of opposite sides. It has therefore six directions of symmetry, and they make angles of 60 with each other. A square, ABCD (fig. 11), is symmetrical about the diagonals AC, DcC CL Fig. 11. BD, and also about the lines ac, bd, which join the middle points of opposite sides. It has therefore four directions of symmetry, and they are inclined to one another at angles of 45. An equilateral triangle, ABC (fig. 12), is symmetrical about each of the lines Aa, Bb, Cc, drawn from an angle perpendicular to the oppo- site side. It has therefore three directions of symmetry, inclined to one another at angles of 120. An oblong, ABCD (fig. 13), is symmetrical about the two lines ac, bd, joining the middle points of opposite sides, which are at right angles to one another. It is not symmetrical about the diagonals. For draw DP perpendicular to AC and produce it to D', making PD' = PD; then if ADC be turned round AC, it will come into the position A D'C and will not cover ABC. Mineralogy. 27 A rhombus, A BCD (fig. 14), is symmetrical about its diagonals, for Fis. 12. Fig. 13. they are at right angles to one another, and if ABD be turned round BD, A will come to C and ABD will cover BCD exactly. An isosceles triangle has been shown to have one line of symmetry, and this is its only one. There is no line about which a rhomboid, ABCD (fig. 15), is sym- Fig. 14. Fig. 15. metrical, but it possesses an inferior degree of symmetry, called symmetry about a point, which may be thus defined. A plane figure is symmetrical about a point 0, if, when we take any point P on the boundary of the figure, join PO, and produce PO to Q making OQ = OP, Q also lies on the boundary. Now this is obviously true for the rhomboid, if be the intersection of the diagonals, for the triangles POD, BOQ are equal in all respects, and therefore OP= OQ. A scalene triangle is symmetrical neither about a line nor about a point. Corresponding to the Symmetry of Plane Figures about Straight Lines is the Symmetry of Solid Figures about Planes. A plane of symmetry is a plane which divides a solid into two halves which are in shape, size, and position the exact counterpart of one another, or to give a more exact definition, if from any point P on the surface of the solid we draw PO perpendicular to a plane of symmetry, and produce PO to Q, making OQ = OP, then Q also lies on the surface of the solid. Note that it is not sufficient that this be true for one pair, or two pairs, or three pairs of points. It must be true for every point on the surface. 28 Geology. A sphere possesses an infinite number of planes of symmetry, for it is symmetrically divided by every plane through its centre. A cube has nine planes of symmetry. Three of these pass through the centre and are parallel to the faces : they are perpendicular to one another, and intersect in three straight lines, AOA', BOB', COG' (fig. 16), per- pendicular to one another. The cube is also symmetrical about planes, such as PjP 3 P 7 P 5 , P 3 P 4 P 5 P 6 , passing through the parallel diagonals of oppo- site faces. Each pair of faces gives two of these, and as there are three pairs of parallel faces, the whole number is six. If we join the middle points of the four vertical edges of a cube with the centres of the top and bottom faces, we get a p. lfi Regular Octahedron, and this is evidently symmetrical about the same nine planes as the cube. A regular hexagon has six lines of symmetry. Consider a Right Hexagonal Prism or Pyramid : it is evidently symmetrical about six planes, each of which passes through the longitudinal axis and one of the lines of symmetry of the hexagonal base. The base itself is also a plane of symmetry. A square has four lines of symmetry. Consider a Right Square Octahedron or Prism : it will be symmetrical about four planes, each of which passes through the longitudinal axis and one of the lines of symmetry of the square base. The base itself is also a plane of symmetry. In the same way a Right Rhombic Octahedron or Prism has three planes of symmetry ; the rhombic base, and planes passing through the longitudinal axis and each of the diagonals of the base. A Monoclinic Octahedron or Prism has only one plane of symmetry. The base is evidently not one, for let ABA'B' be the Fig. 17. rhombic base (fig. 17) and AOA' the clino-diagonal, then a perpendi- cular, say from P, on the plane ABA'B' will fall altogether outside the solid, and can never pass to a corresponding point on its surface. But the plane CO A passing through the longitudinal axis and the Mineralogy. clino-diagonal is a plane of symmetry, for it is perpendicular to the base ; hence all perpendiculars to it, such as EFe, are parallel to the base, and will cut the surface of the solid in points which are equi- distant from the plane. A Triclinic Octahedron or Prism has no plane of symmetry, but it is symmetrical about the middle point of the longitudinal axis. A straight line perpendicular to a plane is called a Normal to that plane. A normal to a Plane of Symmetry is called an Axis of Symmetry. In using our symbolical notation for the faces of crystals, it is con- venient to take Axes of Symmetry for the co-ordinate axes. Planes of Symmetry are of two kinds. This will appear best from an illustration. In fig. 18 CABA'B'C'is&right square octahedron, ABA'B', CAC'A, CB G'B' are planes of symmetry, and AO, BO, CO axes of symmetry. Now there is an important difference between the axes OA, OB, and the axis OC, for if we turn the figure round OC till OA comes into the position OB, the solid will not be altered. OA and OB are called Interchange- able Axes of Symmetry. We cannot turn the solid round OA till OB coincides with OC without altering the solid. Therefore OB and OC are not interchangeable. The plane ABA'B' contains two in- terchangeable axes. A Plane of Symmetry which con- tains two or more Interchangeable Axes is called a Principal Plane of Symmetry, and a normal to it a Prin- cipal Axis of Symmetry. The planes AOC, BOG each contain two axes, but they are not interchange- able axes. A Plane of Symmetry which does not contain at least two inter- changeable axes is called an Ordinary Plane of Symmetry, and a nor- mal to it an Ordinary Axis of Symmetry. OC being normal to a Principal Plane of Symmetry is a Principal Axis of Symmetry; AO, BO are Ordinary Axes of Symmetry. Planes of Symmetry in simple Solids. Let us now see how many Principal Planes of Symmetry the solids we have been so far dealing with possess. In the cube (fig. 16) OA, OB, OC are axes of symmetry, and they are evidently interchangeable. Hence each of the planes xOy, yOz, zOx are Principal Planes of Symmetry. In the Hexagonal Pyramid, the base of which is shown in fig. 10, 30 Geology. there are six axes of symmetry lying in the base ; of these AD, BE, and CF are interchangeable, as are also ad, be, cf. The base is there- fore a Principal Plane of Symmetry, and the Longitudinal Axis a Principal Axis of Symmetry. The longitudinal axis is not interchangeable with any of the axes in the base, there is therefore no other Principal Plane of Sym- metry. In just the same way we may show that the Square Octahedron has only one Principal Axis of Symmetry, viz. its longitudinal axis. In the Rhombic Octahedron there are three axes of symmetry, but they are none of them interchangeable. It has therefore no Principal Plane of Symmetry. In the Monoclinic and Triclinic Octahedrons there can evidently be no Principal Plane of Symmetry. Classification of Crystals according to their degree of Symmetry. The instances given in the last paragraphs show that the solids treated of possess different degrees or different types of Symmetry, and it would evidently be possible to group them into classes or systems according to the amount of Symmetry they exhibit. Now though the shapes which crystals present are almost countless, and though scarcely any two of these shapes might seem to the superficial observer to have anything in common, yet really careful examination shows that a very large number possess exactly the same type of symmetry. The cube and regular octahedron, for instance, have each three principal planes of symmetry. These two shapes, then, and any others we meet with that have the same type of symmet^, may be all put in the same class. A second class takes in all crystals which have the same type of symmetry as the Right Hexagonal Pyramid or Prism, and so on. In this way it has been found possible to group all crystals together into the following six classes or systems. 1. Monometric, Isometric, Tessera], Cubic, Octahedral, or Regular System. Crystals framed on the type of the Cube and Regular Octahedron, with three Principal and six Ordinary Planes of Symmetry. It is convenient to take the three Principal Axes of Symmetry for the Co-ordinate Axes. A face of the Regular Octahedron is taken for the Parametral Plane, and the Parameters are evidently all equal. We next come to crystals having only one Principal Axis of Sym- metry, and these fall into two classes according as they possess six or four Ordinary Axes of Symmetry. 2. Hexagonal, Rhombohedral, Three-and-one axial (Drei-und-ein- axige}, or Monotrimetric System. Crystals framed on the type of the Right Hexagonal Pyramid, with one principal Axis of Symmetry, viz. the longitudinal axis of the Pyramid, and six Ordinary Axes of Symmetry. It is convenient to refer these crystals to four instead of three co- ordinate axes. One axis, called the longitudinal or vertical axis, is the Mineralogy, 3 r Principal Axis of Symmetry ; the other three, called lateral axes, are those three Ordinary Axes of Symmetry which pass through the angles of the hexagonal base (OA, 00, and OE in fig. 10). The lateral axes are equal to one another, and the angle between each pair is 120; the vertical axis is at right angles to the lateral axes, but not necessarily equal to them. A face of an Hexagonal Pyramid is taken for the Parametral Plane, and of the Parameters only two will usually be equal 3. Dimetric, Monodimetric, Pyramidal, Prismatic, Tetragonal, Quad- ratic, Two-and-one axial (Zwei-und-einaxige) System. Crystals having the same type of symmetry as the Eight Square Octahedron, viz. one Principal and four Ordinary Axes of sym- metry. They are referred to the Principal Axis of symmetry and the diago- nals of the square base as Co-ordinate Axes (0(7, OA, and OB in fig. 8). The first is called the longitudinal or vertical axis, the others the lateral axes. The Co-ordinate Axes are therefore at right angles to one another. A face of a Square Octahedron is taken for the Parametral Plane, and usually only two of the Parameters are equal. There remain crystals without a Principal Plane of Symmetry, and these fall into three classes. 4. Trimetric, Orthorhombic, Rhombic, Orthotype, or One-and-one axial (Ein-und-einaxige) System. Crystals framed after the same type of symmetry as the Eight Rhombic Octahedron, viz. with three Ordinary Planes of Symmetry. They are referred to the three axes of symmetry, viz. the diagonals of the rhombic base and the longitudinal axis of the Octahedron, as Co-ordinate Axes. These axes are therefore at right angles to one another. A face of a Ehombic Octahedron is taken for the Parametral Plane, and of the three Parameters two must, and usually all three will, be un- equal. For the diagonals are always unequal, and the longitudinal axis is in most cases equal to neither of them. 5. Monoclinic, Monosymmetric, Oblique, Clinorhombic, Monoclino- hedral, Hemiorthotype, or Two-and-one membered (Zwei-und-einylie- derige) System. Crystals having the same type of symmetry as the Monoclinic Rhombic Octahedron, viz. only one plane of symmetry. They are referred to the diagonals of the rhombic base and the longitudinal axis of the Octahedron as Co-ordinate Axes (OA, OB, and OG in fig. 3). Of the angles then between these axes two are right angles and one an oblique angle. A face of a Monoclinic Ehombic Octahedron is taken for the Para- metral Plane, and of the Parameters two must, and usually all three will, be unequal. 6. Triclinic, Asymmetric, Anorthic, Doubly oblique, Triclinohedral, One-and-one membered (Ein-und-eingliederige) System. Crystals framed on the type of the Triclinic Octahedron, viz. with symmetry about a point alone. 32 Geology. They are referred to the diagonals of the rhomboidal base and the longitudinal axis of the Octahedron as Co-ordinate Axes. The angles between these axes are therefore all of them oblique. A face of a Triclinic Octahedron is taken for the Parametral Plane. Of the Parameters therefore two must, and usually all three will, be unequal. What is meant by the word " Form " in Crystallo- graphy. The word " Form," which in ordinary language is of the vaguest of expressions, has a definite and restricted meaning in Crystallography, which will be best explained by an example. Suppose that we are told that a crystal belongs to the Monometric System, and that one of its faces is (oo oo 1). Let us inquire what this information amounts to. The face (oo oo 1) is parallel to Ox and Oy, and cuts Oz at a distance 1 : it is the face P^P^ in fig. 16. yOx is a plane of symmetry, and there must therefore be a corre- sponding face on the opposite side of ; the face P 5 P 6 P 7 P 8 , namely. Its symbol is (oo oo 1). Again Oz and Ox are interchangeable axes, and if the crystal be turned round Oy till Oz coincides with Ox, its shape must not be altered. By turning it thus, P^PJP^P^ will come into the position P^PgPg. This then must be a face of the crystal Its symbol is (1 oo oo ). yOz being a plane of symmetry, there must be a face to correspond with PjPgPgPg on the opposite side of 0\ the face P 3 P 4 P 8 P 7 , namely. Its symbol is (1 oo oo). Exactly the same line of reasoning will show that the crystal must possess the two faces P^PgPg, P 2 P 3 P 7 P 6 , whose symbols are (oo 1 oo) and (oo 1 oo ). The solid is a cube. In this case then the presence^ of the face (oo oo 1) makes it necessary that the five other faces (oo oo I) (1 oo 00 ) (I oo oo) (oo 1 oo) (oo 1 oo) shall also be present. In crystals belonging to other systems we should find, in exactly the same way, that if one face is present, symmetry makes it necessary that certain other faces shall also be present. If we bear this in mind we shall see that it is not necessary in describing a crystal to write down the symbol for every separate face. The one symbol (oo oo 1) denotes not only the face PiP 2 P 3 P 4 , but really all the other faces of the cube, and is alone sufficient to tell us that the crystal is a cube. Also note that the symbols for the other faces are obtained merely by changing the order of the indices. The group of faces which symmetry makes it necessaiy should coexist in a crystal, and which are denoted by different arrangements of the indices in a symbol, is called a Form. A Form may evidently be denoted by the symbol of one of its faces, but to distinguish between the two cases we will enclose the symbols of faces in brackets of this shape ( ), and the symbols of forms in brackets of this shape { }. If all the faces that can coexist are present, the Form is said to be Holohedral. Mineralogy. 33 In some cases only one-half of the faces belonging to a Form are present : the form thus obtained is called Hemihedral. If only one-fourth of the faces are present, the form is Tetartohedral. Sometimes only the faces of a single Form appear on a crystal, as in the cube just taken as an illustration. The crystal is then said to be Simple. More frequently faces belonging to two or more forms are present in the same crys- tal. These are called Combinations. In fig. 19, for instance, the faces marked H are those of a cube ; the faces marked would, if they were produced to meet, form a regular octahedron. Here then faces belonging to two forms occur. Different parts of a crystal cannot possibly crystallize in different systems. Hence in a Com- pj 19 bination the different Forms must all belong to the same system. From this a very important law follows. In the Combination in fig. 19 the faces marked would, if produced till they met, enclose a Regular Octahedron, and those marked H would enclose a Cube. If we look at one angle of the Cube we see that it has, so to speak, been cut off by a plane of the octahedron ; but if this plane be present, all the other seven planes of the octahedron must be present also, and therefore all the other seven angles of the cube must be cut off in the same way. Again, in fig. 25 the planes marked D are those of a Rhombic Dodecahedron, and they cut off the edges of the cube ; and here, as before, if one edge be cut off, all must be cut off after the same fashion. It is easy to see that this must always be the case, and we arrive at a general law that if one angle or edge of a Form be modified, all the similar angles or edges must be modified in a similar manner. The only exceptions are when one of the Combining Forms is Hemihedral, and then only one-half loill be modified, and when one of the Com- bining Forms is Tetartohedral, when only one-fourth will be modified. Of course in Forms possessing a lower degree of symmetry than the Cube, it will not be necessary for all edges to be modified, only all similar edges. For instance, in the Square Octahedron in fig. 18, if one of the lateral edges AB be modified, the other three lateral edges BA', A'B', B'A must be similarly modified. But the vertical edges CA, CB, CA', CB' need not be modified, because the lateral edges are. If one vertical edge however is modified, all must be similarly modified. Crystalline Cleavage. The meaning of this term will be gathered from the following example. Here is a piece of a mineral known as Calc-spar or Calcite (fig. 20, a). It is a solid bounded by smooth, glistening faces, each of which is a rhombus, and all these rhombuses are of exactly the same shape, that is, the corresponding angles are the same for every one. The solid is called a Rhombohedron. Knock a bit off one corner ; it falls away in the shape of a rhombohedron, smaller than the one from which it has been broken off, but in every other respect exactly similar (figs. 20, 6 and c). The rhombic faces of the detached frag- c 34 Geology. ment have their angles exactly equal to the corresponding angles of the rhombuses that bounded the original block, and the corresponding interfacial angles of the two specimens are exactly equal. Further, the bit we have broken off can itself be further broken up into rhombohedrons (fig. 20, d), these again into still smaller rhoinbobe- drons, and in every case the shape of the fragments will be identically that of the block we started with. This property which the mineral has of breaking more readily in certain directions than in others, and of breaking in these directions with a smooth face, is called Crystalline Cleavage, and the smooth faces thus obtained are called Planes of Cleavage. We might by force make Calc-spar break in other directions than those of the Cleavage Planes, but the surfaces we should thus obtain would be no longer smooth and shining, but rough and dull. But Calc-spar is found crystallized not only in the shape of a rhombohedron, but in hundreds of other forms besides. One of the commonest of these is called, from the pointed shape of the crystals, Dog-tooth Spar (see fig. 40). Take one of these tooth-shaped crystals, tap it gently with the hammer, and it will fall into a number of rhom- bohedrons identical in every respect except size with those of our first specimen. This crystal, seemingly so different from the rhombohedron we were just now handling, is really built up of a number of elements agreeing with it exactly in their geometrical form. This will also be found to be the case with all the different crystalline shapes under which Calc-spar is found. Calc-spar has been chosen for the foregoing example because it cleaves readily parallel to all the faces of the rhombohedron. In other crystallized substances cleavage can be obtained only in two directions ; in others only in one ; in some not at all, or only with the greatest difficulty. Fluor-spar furnishes another good instance of Crystalline Cleavage, and also yields a very instructive example of the passage of one crystalline form into another. This mineral is found frequently crystallized in cubes, such as fig. Mineralogy. 35 21, a. If a knife be placed near one of the angles of the cube, touching the face A BCD, and equally inclined to the three faces that meet in B, and then firmly pressed against the crystal, a bit will fly off, (a) U) V ,r Fig. 21. bounded by triangular faces, and the crystal be reduce i to the shape in fig. 21, b. If each of the angles be treated in the same way, the whole crystal may be similarly modified. We shall then find that we may continue to split off slices parallel to faces corresponding to abc, till the crystal assumes the shape of fig. 19 ; by continuing the process it is reduced to the shape in fig. 21, c, which is a solid bounded by six squares, each of which is formed by joining the middle points of the edges of each face in the original cube (the dotted figure efgh, in fact, of fig. 21, 6), and eight equilateral triangles. By still further continuing the process the square faces grow smaller and smaller, till the crystal is at last reduced to the shape of fig. 21, c?, which is a regular octahedron, bounded by eight faces, each of which is an equi- lateral triangle. Each of the angles of the octahedron occupies what was the centre of a face of the original cube, the relative position of which is shown by dotted lines. The Cleavage of Calcite being parallel to the faces of a rhombohedron is called Khombohedral, that of Fluor-spar is from a similar reason called Octahedral. 36 Geology. In other minerals cleavage takes place parallel to the faces of other crystalline forms, but no mineral can possess more than one type of cleavage. Cleavage then is evidently a physical character which will aid us in recognising minerals. The Simple Forms of the Six Systems. We will now describe the more important simple forms in the six crystallographic systems. There is a system of crystallographic notation, invented by Naumann, which is so very widely used that the student will do well to familiar- ize himself with it. It differs but little from the notation we have been employing, and we will, in the case of each form, give Naumann's symbols in addition to our own. Another system, first used by Whewell and developed by Miller, differs but little from the one described on p. 21. If a, b, c be the parameters for any mineral and ^, , 7> Z> ~D f) 7> Z> ft * r T> *>* 8* r 5 r< to r *L r T These diagonals therefore are the edges of the form. They are all equal. The form is therefore bounded by four equal equilateral triangles, that is it is a Regular Tetrahedron: it is shown in fig. 22a. A second Tetrahedron can be obtained by discarding the faces retained above, and producing those omitted. One Tetrahedron is called posi- tive, the other negative. The student should draw a figure showing the other Tetrahedron. The symbols for the Regular Tetrahedrons are {in} or - , and - {in} O or- 2 In Jordan's " Elementary Crystallography " (Murby's Science and Art Department Series of Text-Books) full directions are given for making cardboard models of crystals. Coaguline is better than gum for fastening the models together. Let the student make a model of a Regular Octahedron, and also one of a Regular Tetrahedron with edges twice as long as those of the Octahedron. If one face of the Tetrahedron be left unfastened, the Octahedron can be slipped inside, and the relation between the two is instantly seen. (2) {loooo}, H, oo oo. This form, we have already shown (p. 32), is a Cube or Hexahedron. * CAB, CA'B' are like the two sides of the roof of a house, and their line of intersection corresponds to the ridge of the roof. Mineralogy. (3) {1 loo}, [GO], oo 0. 39 Here each face is parallel to one of the axes, and cuts the other two at equal distances from 0. There will be four faces parallel to Oz, and therefore lying in the same zone, viz. (1 1 oo), (1 1 GO), (1 1 oo), (1 1 oo). There will be also a zone of four faces parallel to Oy, and another of four faces parallel to Ox. Twelve faces therefore in all. It is not difficult to see from the perfect symmetry of the form that all the edges must be equal, and also that the opposite edges of each face must be parallel ; the faces are therefore equal rhombuses and the form is called a Rhombic Dodecahedron. The shape of the form may also be conveniently found thus. Let P^PzPP 5 P 6 P 7 Ps (fig. 23) be a cube having Ox, Oy, Oz for its axes. Fig. 23. Through every edge of this cube draw a plane equally inclined to the two faces which intersect on that edge. These planes will be the faces of the form. For consider the plane passing through P^. It is parallel to Oy, for P^ is parallel to Oy. It cuts Ox and Oz at equal distances from 0, for it is equally inclined to P^PJP^ and P,P. 2 P S P 4 , and these planes are parallel to Oz and Oy, so that it is equally inclined to Oz and Oy. Similar reasoning will show that each of the said planes is a face of the form. Three faces meet in each of the points P. Therefore each of these points is an angle of the form. Again it is easy to see that the faces through P^, P. 2 P 6 , P 6 P 5 , all meet Ox on the same point A, OA being =P 1 P 2 ; and that 4O Geology. four corresponding faces all cut Ox in the same point A', OA' being = OA. B, B\ (7, C' are corresponding points on Oy and Oz. Hence four faces meet in each of the points A, A', B, B', C, C'. Therefore each of these points is an angle of the form. Hence the edges are P^A, P.B, Pf, P 2 A, P 2 B', P 2 <7, P 3 A e , P 3 C, P 3 B', PA', P 4 , P 4 C, in the upper four octants, and similar lines obtained by changing C into C' in the lower four octants. Fig. 24 shows the complete form. Fig. 24. We have explained a way of drawing this form at length ; if the student will thoroughly master the method, he will have no diffi- culty in following the similar explanations that follow, which have been necessarily somewhat condensed. He should draw with the utmost care fig. 23 on a large scale, on a sheet of good drawing- paper, adding the faces not shown in that figure. First let him draw the whole in thin pencil lines, then he may ink in the edges, making those which are in view strong, and representing those which would be seen only if the solid were transparent by faint or dotted lines. He must also, from the instructions in Jordan's " Elementary Crystal- lography," construct a cardboard model of the form.* By making very careful drawings on a large scale and models of the * The " nets " for these models consist of the same figure or group of figures several times repeated. It saves trouble and ensures accurate fitting to draw this figure carefully on the cardboard. Then copy it on to a bit of tracing- paper. Then lay this tracing in the positions which it has successively to occupy, and prick the angles through on to the cardboard with a fine needle- point. Mineralogy. different forms as he goes along, the student will gradually grow so familiar with the meaning of crystallographic symbols that he will acquire the power of picturing in his mind's eye from these symbols alone the shape of every form, and so become independent of figures and models. He should from time to time test his progress by trying to realize what the shape of the form he is going to study will be before drawing it ; but till he can do this, he should steadily persevere in drawing and modelling till the power comes. It is important that the student should realize the intimate relation- ship which exists between the Octahedron, the Cube, and the Rhombic Dodecahedron. We have given two ways in which the Octahedron may be obtained from the Cube (p. 28, and p. 35). We have also just shown one relationship between the Cube and the Rhombic Dode- cahedron ; here is another. Take a cube and cut off each of the edges by a plane equally inclined to the faces which meet in that edge. We get a solid of the shape shown in fig. 25. The faces D have the shape abcdef (fig. , the edges be, dc being parallel to ef, af. Take off another f< Fig. 25. Fig. 26a. Fig. 266. set of slices, keeping the knife parallel to D. The crystal now assumes the shape in fig. 26 ; the faces H are smaller, the faces D larger than before ; each of the faces D has the shape shown in fig. 26a ; the edges ed, ab are shortened, the others longer than before. By con- tinuing to take off slice after slice in this way, the edges ed and ab grow shorter and shorter till at last they disappear. The faces H disappear with them, and the faces D assume the shape bfgc (fig. 266), which is a rhombus. Each edge gives place to a rhombic face, and since there are twelve edges, we obtain a solid bounded by twelve rhombuses. The solid CABC'A'B', fig. 24, is a Regular Octahedron, and by slicing off each angle of the Dodecahedron in that figure by planes equally inclined to the three axes, we convert it into this Octahedron. With a little care the student will find no difficulty in cutting out 4 2 Geology. a cube in chalk, an apple, or a turnip, and transforming it either into a Regular Octahedron or a Rhombic Dodecahedron. (4) {1 mac}, [woo], oo m. Let a, ', b, b f , c, c' (fig. 23) be the middle points* of the faces of the cube P^P^P.P^. Produce Oa, Oa, Ob, etc., to A, A', B, etc., making a A = a A' = Ill Join A and the angles P sponding angles of the cube. The resulting solid will be this form. ^ and join A', B, B', C, C' to corre- Fig. 27. For let D (fig. 27) be the middle point of P^P^ then CD is the in- tersection of the plane CP^P, 2 with the plane of yOx ; produce CD to cut Ox in E, then OE DC Oc <9c OC Cc Cc Oc But OE and 0(7 are the intercepts of the plane C f P 1 P 2 on Ox and 6ty, and they are in the ratio of m to 1. The plane is also parallel to Oy. Therefore its symbol is (m oo 1). 2 In the same way we may show that the other faces have corresponding symbols. The form may be evidently obtained by A placing right pyramids of equal height on each face of a cube, the vertex of each pyramid t being on the axis passing through the face on which it stands. Hence the form is called the Four-faced Cube. It is also known as the Tetrakis Hexahedron, or Tetrahexahedron, or Fluoride. Fig. 28 shows its shape. Form of the Four-faced Cube is obtained by Fig. 28. The Hemihedral omitting the shaded faces in fig. 28. * The letters are the same as in fig. 23. In that figure m=l. The student m easily draw a similar figure in which m is some integer othei ler than unity. Mineralogy. 43 The two planes BP^P^ BP 4 P 8 are parallel to Oz. Therefore when extended they will intersect in a line 77' (fig. 29), through B parallel to Oz\ 7 and 7', the ex- tremities of this edge, will be the points where the line meets the planes CP^P^ CP S P 5 . The corresponding faces on the opposite side of will intersect in a line occupying a similar position. There will therefore be a pair of edges through B and B' parallel to Oz. There will evidently be also a pair of edges through A and A' parallel to Oy, and a pair through C and C" parallel to Ox : let ftp be Fig. 29. the edge through A. All the points P will be angles of the form, for three of the faces retained meet on them. Three of the edges then of the face formed by the extension of BP^ 5 will be 77', 7P 1? y'P 5 , and the other two will be the lines along which it meets AP^P^ AP 5 P 6 , viz. ftP^ and (3P 5 . The face will therefore be a pentagon y'yP^P^ There will be twelve faces and the form is a Pentagonal Dodeca- hedron. The edges yP lt P$, /3P 5 , P 5 y' are all equal : 77' is not equal to them. The student will have no difficulty in making models of the Four- faced Cube and Pentagonal Dodecahedron from Mr. Jordan's figures. It is not so easy to make them of such a size that the Dodecahedron is the Hemihedral Form of the particular Four-faced Cube chosen. With the following dimensions however this will be the case. AP 2 = 3, P^ = 2-3, ^7 = 1-9, 77=1-7, The " nets " are constructed thus. The " net " for the Four-faced Cube consists of a group of four equal isosceles triangles repeated six times over. It may be easily constructed of the dimensions given in this way. Describe a circle with centre and radius = 2'3 : open a pair of dividers till the distance between the points is 3, and with them place four chords, AJB, BC, CD, DE, each = 3, in the circle. OAB, OBC, OCD, ODE will be one group of triangles. Copy this carefully on a bit of tracing-paper, lay the tracing-paper in the position of the second group, and prick through the angles with a fine needle on to the card- board, and repeat the process till the " net " is complete. The Pentagonal face of the Hemihedral Form is constructed thus. Take a line 77' =1 '7 : draw yM, y'N, making angles of 110 with 77'. On yM take yP l = 1 '9, and on y'N take y'P 6 = 1 '9. With centres P^ describe two circles each with radius 1 '9. These circles will intersect in J3. This Pentagon must now be traced, and the " net " constructed by laying it in the successive positions indicated in Mr. Jordan's figure, and pricking the angles on to the cardboard. Next mark on the models the letters used in figs. 28 and 29. In 44 Geology. the Pentagonal Dodecahedron the P's are easily found. B is the middle point of 77', and A, A', B', C, C' the middle points of the other corresponding edges. The lettering of the Pentagonal Dodecahedron being completed, join BP^ P 5 , P^, and so obtain the portion of the face yP-ifiPsy', which corresponds to the face BP^P 5 in the Four-faced Cube. Do the same for each face. Then colour with the same colour the faces retained in the Four-faced Cube and the portions of the faces of the Petagonal Dodecahedron which correspond to them. The rela- tion between the two becomes then clear enough. (5) {1 lm},[m],mO. Since the axes are interchangeable we must transpose the three indices in every possible way in order to get all the faces required by symmetry. We find by doing this that there are in the first octant the three faces (1 1 m), (1 m 1), (m 1 1). On the axes take OA = OB = OC, and OA' = 0ff = OC' = mOA (fig. 30). The plane (I m 1) cuts yOx inAB', the plane (m I 1) cuts yOx in A'B. Therefore if c be the intersection of AB' and A'B, (1 m 1) and (ml 1) both pass through c. They also both pass through C. Therefore Cc is the line of their intersection or the line of an edge. OF THE UNIVEESIT Mineralogy. \ ^v 4&r ^K In the same way we can show that Aa' ', ^?&' are edges produced. The lines ^4a', .56', Cc' all meet in a point P. P' then is the intersection of three edges, that is an angle. AB is an edge, for it is the intersection of the faces (11 m), (1 1 m) ; and for similar reasons EC and CA are edges. Hence A, JB, and C are angles; and since P is an angle, PA, PE, PC are edges. Hence the faces in the first octant are three equal and similar isosceles triangles, P AB, PEC, PC A. The complete form is bounded by 3 x 8 = 24 equal and similar isosceles triangles ; it is called the Three-faced or Triakis Octahedron, the Trigonal Trisoctahedron, or the Galenoid. The student must draw on a large scale a figure which shows the faces in all the octants. ABC is the face of a Regular Octahedron, and the form maybe obtained by placing pyramids of equal height on each face of a Regular Octahedron, the vertex of each pyramid being on a line through and the middle point of the face on which it stands. AP^CP^ is the face of a Rhombic Dodecahedron, and we may also obtain the form by placing pyramids of equal height on each face of a Rhombic Dodeca- hedron, the vertex of each pyramid being on a line through and the intersection of the diagonals of the face on which it stands. By leaving out all the faces in each alternate octant and producing the remaining faces, we get the Hemihedral form of the Three-faced Octahedron. We may retain P^AE, P^BC, PA, P^A'B', P Z E'C, P S CA' in the upper, and P^AB' , P^E'C', PffA, P^A'B, P S C', P 8 C'A' in the lower octants. The points P lt P 3 , P 6 , P s will be angles of the form, for three of the retained faces meet in each of them. Also A, A', B, B, C, C' will be angles, for four of the retained faces meet in each of them. We may determine the shape of the face produced by the extension of P^AC thus. P^A, P^C are evidently two of its edges. The face must be prolonged till it meets PJB'C and P^B'A. Let $ 2 * be the intersection of the three planes, P^AC, P. 3 'C, P^B'A. Then two other edges of the face produced by the extension of P^AC are Q. 2 A, Q 2 C. It may be shown that P^A = P^C, and Q. 2 A = Q%C. Therefore the face in question is a Deltoid. The complete form is bounded by twelve equal and similar deltoids. It is called the Deltoidal Dodecahedron or Twelve-faced Deltohedron. Even the most carefully-drawn figures of forms like the Three-faced Octahedron and the Deltohedron are more or less confusing. We shall therefore leave it to the student to construct models for himself of these forms. With the models before him he will easily follow the preceding demonstrations. *$ 2 because it lies in the second octant: the other angles corresponding to #2 lie in the fourth, fifth, and seventh octants, and are lettered Q, Q 5 , Q 7 . 5 Geology. The Three-faced Octahedron may have the following dimensions. P^ = 2-l. (7^=3-6. The Deltohedron derived from this has the dimensions. The holohedral form cannot in this case be well placed inside the hemihedral, but their relations may be thus made clear. Colour the omitted faces black and the retained faces red on the holohedral model. On the hemihedral model mark the points P lt P 3 , P 6 , PS, A, A', .Z?, B', C, C', and then colour red the portion of each face which coincides with the corresponding face of the holohedral model. The making of models is a little tedious, but the student will by their aid realize the shapes of the different forms and their relations to one another so rapidly that he will in the end save much more time than has been spent in constructing them. (6) {1m m}, [m m], m m. The construction by which the shape of this form is determined is so exactly similar to that employed in the case of the Three-faced Octahedron that the student will have no difficulty in working it out Fig. 31, by the aid of fig. 31, and showing that there are in each quadrant three faces each of which is a deltoid, and that the complete form is bounded by twenty-four equal and similar deltoids. Mineralogy. 47 The form is called the Tetragonal Trisoctahedron, the Ikositetra- hedron, and the Deltohedron. The following is a convenient way of lettering the angles. The angle on Ox is A, that on Ox is A' \ and the angles on yOy ', zOz are B, B' and (7, C". The angles opposite to are all P's and are distin- guished by suffix-numbers ; they are numbered in the same order as the angles of the cube, P x being the angle in the octant zOxy, P 2 the angle in the octant zOxy', P 3 the angle in the octant zOx'y', and so on. There are four angles lying in the plane xOy, these are all denoted by c's : q is the angle in the quadrant yOx, c 2 the angle in the quadrant xOt/j c 3 the angle in the quadrant y'Ox', c 4 the angle in the quadrant x'Oy ; if a watch were laid on the plane yOx, the counting goes round in the direction opposite to that in which its hands move. The four angles in the plane yOz are denoted by a's, and the four in the plane zOx by 6's ; the a's and b's are numbered in the same way as the c's. In the Hemihedral Form of the Deltohedron all the faces in alter- nate octants are omitted. Consider the face P^Cb^ (mm 1) and the corresponding face P ; j6 2 (7a 4 (in m 1) in the octant xzy. Exactly as in the case of the Regular Tetrahedron (fig. 22), we can show that these two faces will intersect in that diagonal of the face of the circumscribing cube which passes through C and is parallel to AB. And by taking the other pairs of faces which meet on the axes, we can show that they will intersect in corresponding diagonals. Hence six of the edges of the form will be edges of a regular tetra- hedron, Q. 2 QQ 5 Q 7 (suffix-figures denote the octant in which each Q lies). One edge then of the face produced by the extension of P-^Cb^ will be the edge of this tetrahedron which passes through (7, i.e. Q 2 Q 4 : the other edges will evidently be obtained by producing P^, P^ to the extremities of this edge of the tetrahedron. This face will be an isosceles triangle P^Q^Q^. Each of the other faces retained gives rise to a similar triangle, and the complete hemihedral form is bounded by twelve equal and similar isosceles triangles. It could evidently be constructed by placing equal pyramids on the four faces of a regular tetrahedron. It is bounded by 4 x 3 = 12 equal isosceles triangles, and is called the Three-faced or Triakis Tetrahedron, or the Kuproid. The student must make models of the Holohedral and Hemihedral forms of the following dimensions. PA = 1-2, ^=1-8, and distinguish the retained faces of the holohedral form, and the portions of the faces of the hemihedral form which correspond to them, by tinting them the same colour. (7) {1 m n}, [m n], m n. Since the axes are interchangeable, we must transpose the three indices in every possible way in order to get the faces required by symmetry. 48 Geology. In the first octant there will be the six faces (1 m n), (1 n m), (m 1 n), (m n 1), (n 1 m), (n m 1). On the axes take (fig. 32) Fig. 32. The face (1 m n) cuts yOx in A2$ u and the face (m 1 n) cuts yOx in AJB. Therefore if c x be the intersection of AB A^, these faces botli pass through c v They also both pass through C f Therefore C^ is the line of their intersection or an edge. In the same way we may show that A 2 a^ B. 2 \ are edges. A 2 a^ B 2 bv C& all meet in a point P lt which is therefore an angle ; and /Vzu Pjftj, P^ are evidently edges. Also P^A is an edge, for it is the intersection of (1 m n) and (1 n m)-, and for similar reasons P^B, P^C are edges. Further, Ca^ is an edge, for it is the intersection of (n m 1) and (n m 1); and for similar reasons a^B, Ec^ c^A, Ab^ b^C are edges. The faces in the first octant are therefore the six scalene triangles P.Ac,, P&S, PJaH P&C, Pb P&A. The complete form is bounded by 6 x 8 = 48 scalene triangles ; it is called the Six-faced or Hexakis Octahedron, or the Adamantoid. The angles are lettered exactly as in the Ikositetrahedron. ABC is the face of a Regular Octahedron; the points Pare the angles of a cube; AP^CP^ is the face of a Rhombic Dodecahedron: and the form may be obtained by placing pyramids of equal height on Mineralogy. 49 each face of a Rhombic Dodecahedron, the vertix of each pyramid being on the line passing through and the intersection of the diagonals of the face on which it stands. The form {124} has the following dimensions : From these figures the student should construct a model. There are three Hemihedral Forms derived from the Six-faced Octa- hedron. One is obtained by omitting all the faces in alternate octants. A, A', B, ff, C, C", /> P 3 , P 6 , P 8 will evidently be angles. It can be proved that the faces P&C (m n 1), P&A (In m), P^C (n m 1), P 3 a 2 B' (n I m), P 6 Ac 2 (1 m n) PB'c. 2 (m I n), all meet in a point Q 2 . This being so, it is clear enough that the face formed by the exten- sion of P-fi^C will be bounded by P^Q^ Q%C, CP^ It will be a scalene triangle. Each retained face will give rise to a similar triangle, and the whole form will be bounded by twenty-four equal and similar scalene triangles. $ 2 lies in the second octant, and there will be corresponding angles Qv Qfr Q? m the fourth, fifth, and seventh octants. Q 2 , Q 4J Q- a , Q 7 are the angles of a Regular Tetrahedron. The form is called the Six-faced or Hexakis Tetrahedron. The Hexakis Tetrahedron corresponding to the dimensions given for the form {1 2 4} has the following dimensions : The student should make a model, and colour on each face the portions which coincide with the corresponding face of the Holohedral Form. A second Hemihedral Form is obtained thus. Of the eight faces that meet in each of the angles, A, A', B, B', C, C ', omit every alternate pair, and omit them in such order that each retained pair of the faces meet- ing, say in A, adjoins an omitted pair of the faces meeting in B. Thus if we retain of the faces meeting in A, AP 2 c 2 , AP K c AP&, AP 5 c^ we retain of the faces meeting in B, BP^a^ BP 5 a BP 4 a } , BP B a, and similarly for the other angles. A, A', B, B', C, C', and the points P will evidently le angles. The four faces AP&, AP 2 c 2 , CP^, CP^ meet in a point Y lying on the plane xOz. Also the four faces BP^, BP^, CPfa, CPJ>, meet in a point X lying on the plane yOz. This being so, it is clear that the face formed by the extension of CPA will be CTiPiXi. It is a quadrilateral in which P 1 X 1 = P 1 Y 1 , but the other two edge are neither equal to one another nor to P^. Each retained face will give rise to a similar figure, and the form is bounded by twenty-four equal and similar quadrilaterals. It is called the Diakis Dodecahedron, the Trapezohedron, or Diploid. There will be four angles lying on the axial plane that contains Ox and Oy, these may be denoted by Z^ Z 2 , ^ 3 , Z 4 ; and the angles lying on the two other axial planes may be denoted X^ X v X,, X^ 7,, Y& F 3 , Y 4 . SO Geology. The following are the dimensions for a model of the Diakis Dodeca- hedron derived from the Hexakis Octahedron of p. 49 : CTi = 2-4, P^ = ^ = 1-6, ^(7 = 1-5. Angle ^(7^ = 83 47'. A third Hemihedral Form is thus obtained. Of the eight faces that meet in each of the points A, A', B, B', C, C', omit every other face, and omit in such order that no two of the retained faces have a common edge. Thus of the faces meeting on A we may retain Ac.-^P^ Ab^P. 2 , Ac. 2 P M Ab 4 P 5 . Of the faces meeting on B we must then omit BP^, because it has an edge Pc-i common to Ac^P^ and retain BP^, and so on. On a model there is no difficulty in seeing that there will be over each omitted face a point in which the three adjacent retained faces meet. Let a be the point over Ba^P^ a the point over Ca-^P^ /3 the point over Cb z P 4 . Then the face fonned by the extension of Ca^P^ is P 4 /3Caa, a Pentagon. The form is therefore bounded by twenty-four equal and similar pentagons. This form is crystallographically possible, but it has not been observed to occur. 2. THE HEXAGONAL SYSTEM. There are three lateral axes, OA, OB, OC, in fig. 33, lying in one plane and making angles of 120 with one another; and a longitudinal or vertical axis, which we will denote by VOV, at right angles to the lateral axes. The last is the only Principal Axis of Sym- metry. The Parameters are usually only two of them equal. The demands of symmetry are as follows. Produce AO, BO, CO backwards to A', B', C'; then the planes passing through the vertical axis and A A', BB', CC', together with the plane of p. 03 the lateral axes, divide the space round into twelve compartments. The lateral axes are interchangeable. Now a rotation round V through 60 will bring OA to OC', and this must not alter the shape of the form ; a further turn through 60 brings OA to OB, and this must not alter the shape of the form ; and by going on thus we can show that the form must not be altered by six turns round V of 60 each. This is equivalent to saying that whatever faces are present in one of the six upper compartments, a similar set must be present in the remaining five. Further, since the plane of the lateral axes is a plane of symmetry, the faces in the lower six compartments must be the same as in the upper six. In short then whatever faces occur in one compartment, a similar set must be present in each of the remaining eleven. We shall need therefore to find the faces in one compartment only. V Mineralogy. 5 1 The symbols in this system will require a word or two of explana- tion. In the first place, as there are four axes, we might be supposed to need four indices ; four indices we may use if we like, but we will show that three are sufficient. Again, in the Monoinetric System we agreed to take a face of the Eegular Octahedron for the Parametral Plane, and when this was settled, there was only one Parametral Plane possible. In the present system we have agreed to take a face of an Hexagonal Pyramid for our Parametral Plane, but as there may be any number of Hexagonal Pyramids on the same base, this alone does not fix our Parametral Plane definitely. We must know which of these Pyramids is the one whose face is taken for the Parametral Plane. Now if we take any mineral which crystallizes in Hexagonal Forms and examine a number of its crystals, we often find on them a set of faces, which, if they were produced till they met, would enclose an Hexagonal Pyramid. A second set would give us another Hexagonal Pyramid, and so on. One of the Hexagonal Pyramids is chosen as the Unit or Standard Pyramid, and a face of it is the Parametral Plane for the mineral we are dealing with. Suppose that for a face of this Pyramid the inter- cepts on the lateral and vertical axes are in the ratio of a : a : c. Suppose also that for another Pyramid the intercepts of a face on the axes are as a : ma : ma : nc, the last being proportional to the inter- cepts or the vertical axis. Then we might write the symbol of this face (a, ma, m'a, nc\ or (l, m, m, | This is usually further shortened into (1, m, m, n). But if we use the symbol under this last form, we must carefully recollect that n does not mean the simple number n, but n times |. To remind us of this an accent is sometimes placed over the index which belongs to the vertical axis, thus (1, m, m', n). For example, in the Pyramid chosen for the unit pyramid of Beryl fj = 0'5 and the form (1 | 3 &) occurs. Here the intercepts on the three lateral axes are on the ratio i .3. 3 1 2 ' ' and the intercept on the vertical axis is proportional to 3 x 0*5 = Q 1 -5 = |, so that the four intercepts are as 1.3.0.3 1 2 5 ' 2* But the symbols in this system may be still further simplified. Let OA, OB, OC (fig. 34) be the positive directions of the lateral axes, Geology. and produce these axes backwards to A', Z', C' \ then any intercepts measured along OA', OB\ OC' will be negative. Now if any face cut two of the axes on a and b, it is quite clear that in order to find the intercept on the third axis we have only to produce al> till it cuts that axis in c and measure Oc\ so that if we know two of the quan- tities 1, ra, m, we can find the third. Only two of these quantities therefore need be expressed in the symbol. The following relations enable us, when we know- two of the quantities 1, m, Fi S- 34> m, to find the third. If every possible position be given to the line abc, it is easy to see that of 1, m, m, Either two are positive and one negative, Or two are negative and one positive. It is also easily seen that the two which have the same sign are the two largest of the three. It is convenient to agree that m shall be greater than 1, and m' greater than m, and then m and m' have the same sign, and 1 has the opposite sign. Further, on the above agreement it can be proved that m cannot be greater than 2, and that, if the indices have their proper signs given them, m m 1 Draw (fig. 34) NcM at right angles to OC'. Then if we turn nbc round r, Ob (i.e. m) keeps increasing till abc coincides with NcM, and then ~ oc~ We cannot turn abc beyond NcM, for, if we did, Ob would become greater than Oa, or m greater than m'; so m cannot be greater than 2. Also, if bcC' a, 1 _0c_sin. (a- 60) JL_^i_ sin - (<* + 60) m ~~ Ub sin. a in: Oa ~ sin. a .-. 1 + =2 cos. 60 = 1; or + -1 = 0. m m' in m Now either m and m'are both positive and 1 negative, in which case -f __ (-1) = 0, i.e. + ' + 1=0; 79i 'in' m in or m and m' are both negative and 1 positive, in which case w m' m w that is, + - + 1=0 always. m m Mineralogy. 5 3 Since then we can always find m, if we wish, from this equation when m is known, we may discard m from the symbol and write it (1 m h), where 1 and m are proportional to the two shortest intercepts on the lateral axes, and have always opposite signs. Now this is evidently the same as saying that we need only find the faces in one compartment. For of the two lateral axes that bound a compartment one is positive and the other negative, and the intercept on the third lateral axis is obviously longer than the intercepts on the two axes which bound a compartment. Since the lateral axes are interchangeable, we must make 1 and m change places in the symbol to get all the possible faces. We must not allow 1 or m to change places with n, because the lateral and vertical axes are not interchangeable. The greatest number of possible faces in each compartment is there- fore two, viz. (1 m A), (m 1 &), and the greatest number of possible faces in any form is 2 x 12 = 24. In this and the following systems Naumann uses the symbol P to denote a form, the intercepts of whose faces are proportional to the parameters, the Unit or Standard Form in fact. In the other forms he writes the index which belongs to the vertical axis first, and denotes it by m. In this notation then P represents a form the intercepts of whose faces are in the ratio of and m P n represents a form the intercepts of whose faces are in the ratio of In Miller's notation a Rhombohedron, which Naumann looks upon as the Hemihedral Form of an Hexagonal Pyramid, is taken for the Standard Form, and the axes are lines perpendicular to the three faces which meet in its vertex. The following are the more important forms in this system. (1) {11*}. [n]. mP. The first two indices, bear in mind, must have opposite signs. Take (fig. 33) Oa on OA=Ob' on OB', then Oa and Ob' are equal and of opposite sign, and the only face in the first compartment is Vab', an isosceles triangle. The form is therefore bounded by twelve equal isosceles triangles, and is a Right Pyramid with a Regular Hexagon for its base. Here, to find m', we must put m== 1 in the equation m m' ' and we get ^=0, or m is infinitely great, that is each face is parallel to one axis. The figure shows this is the case. 54 Geology. (2) {12^}. [2n]. mP2. Take (fig. 35) on OA Oa^ZOa, and on OJ5' Ob\=20b, and corresponding points on the other axes. ab\ is the intersection of (1 2 &) with the plane of the lateral axes. Fig. 35. Since the axes are interchangeable we must also have the face (2 1 n), and the intersection of this face with that plane is afi'. There will be a pair of faces in each of the other compartments determined in the same way. But there will not be 2x12 or 24 faces on the form. For the points afic^ lie on the same straight line, and so do all the corresponding sets of points all round. The base then will again be a Regular Hexagon, and the form an Hexagonal Pyramid. In this case however the axes, instead of passing through the lateral angles, pass through the middle points of the lateral edges. This form is called the Hexagonal Pyramid of the second order ; \n\ being the Pyramid of the first order. Here we should find m'=2, and it is clear from the figure that the intercepts of each face on two of the lateral axes are equal, and double the intercept on the third. Since the forms [n], [2 n\ are identical in shape, it may be asked, Why employ two symbols for them 1 If we had only to deal with single instances of Model Crystals, this would not be necessary ; but if faces belonging to the two forms occur, as they do, on the same crystal, our symbols must tell us the relative positions of the faces, and as the two forms occupy different positions with respect to the axes, we must have a separate symbol for each. The Hemihedral Form of the Hexagonal Pyramid is obtained by omitting every other face, the faces omitted in the upper half of the solid being directly above those retained on the lower half. Mineralogy. 55 In fig. 36 ABCDEF is ,the hexagonal base, and if V and V be the vertices, In the upper half we retain the faces VAB, VCD, VEF. In the lower half we retain the faces V'BC, V'DE, VFA. Fig. 36. If the three faces VFA, V'BC, VAB be prolonged till they meet, then the intersection of these faces will be an angle of the form. Produce FA, CB to meet in H, then V'FA, V'BC intersect in the line VII. Join OH cutting AB in G, then VG is in the same plane as V'H. VG lies in the plane VAB. The point required will evi- dently then be P lt the intersection of VG and V'H. It lies below the base. In the same way the intersection of VAB, VCD, and V'BC will be a point P. 2 in a corresponding position to P lf except that it lies above the base. By proceeding in this way we fix six lateral angles of the form, alternately above and below the base. The form is completed by joining these lateral angles and the vertices. It is evidently bounded by six four-sided equal figures, and it can be proved that they are rhombuses, but the proof is scarcely elementary enough to be introduced here. It is called a Rhombohedron. The vertices and the vertical axis of the Rhombohedron are the same as those of the Hexagonal Pyramid, from which it is derived. The lateral axes join the middle points of opposite lateral edges if that Pyramid is of the first order, or join points corresponding to G, if of the second order. The vertical angles of the Rhombohedron are contained by three 56 Geology. plane angles; of the three angles forming the lateral solid angles only two are equal. This enables us to fix the vertical axis of any given Khombohedron. A second Rhombohedron may be obtained by prolonging those faces which we discarded, and discarding those we retained in the above instance ; it is distinguished by prefixing a negative sign to its symbol. The symbols for these Rhombohedrons are 1{1 1 ?'i}and -1{1 1 h}. This is usually abbreviated into n R and n R. A Rhombohedron obtained from the unit Pyramid, where n = 1, is R or R-, that from the Pyramid {1 1 2} 2 R or - 2 R, and so on. It is impossible to show in a satisfactory way by a figure the relations of the Rhombohedron and the Hexagonal Pyramid ; but models are easily made, and by their help the student will at a glance realize how one form is derived from the other. If the models be made of the following dimensions they will exactly fit one another : Lateral edge of Pyramid . . . . 4*9 Vertical edge of Pyramid . . . . 7*1 Edge of Rhombus ...... 6 -6 Angle of Rhombus . . . . . 96 The Rhombohedron should be halved along its lateral edges, and one half fixed on to the Pyramid by a few drops of gum. The other half may be lifted off or put on at pleasure. It is as well to colour the parts of each face of the Rhombohedron which cover the faces of the Pyramid. We will now take the general form {I m ^}, [m n], m P n, where m is less than 2. It will perhaps make the matter clearer if we give m a definite 3 numerical value, say 9 so that the symbol is {if,,}. Here, as in the last case, since the lateral axes are interchangeable, we must have in each compartment the two planes Take on OA (fig. 37) any length Oa, and 0^ = * Oa; on OB' Ob' 0a, Ob\ = \0b'. Then aft, and ab\ are the intersections of the planes ( 1 ^ n } Mineralogy. 57 ( '^ 1 n\ with the plane of the lateral axes ; and if these lines intersect on P^ aP l} l'l\ are the two lateral edges in the sector AOB'. planes Fig. 37. In the same way we may determine the lateral edges of the in the other sectors. We thus obtain for the base a figure of twelve equal sides, or an Equilateral Do- decagon. The angles at the points P are all equal, as are also the angles at the corners , b', c, etc., intermediate to these points; but the angles at P are not equal to those at a, 6', etc. If 77' be the vertices of the form, the vertical edges VP l} 7P 2 , etc., are all equal, as also are 7, 76', 7c, etc. ; but VP is not equal to Va. The form therefore is bounded by twenty-four equal and similar scalene triangles ; it is called the Dihexagonal Pyramid (fig. 38). m would have been found in this case to be 3 ; and if any face aP : be produced to cut the third lateral axis in c', then it will be found that 0^ = 300 Fig>38> By omitting every alternate pair of faces and producing the other faces till they meet, we get one Hemihedral form of the Dihexagonal Geology. Pyramid ; the faces retained in the lower half are directly below those ^ omitted in the upper half. In fig. 38 the omitted faces are shaded. In the upper half we retain VaP l9 VaP 6 , VcP,, VcP,, VbP 4 , VbP 5 (figs. 37 and 39). In the lower half we retain V'b'P,, V'b'P* V'a'P. A , V'a'Pt, V'c.'P,, V'cP & One angle of the form will be the intersection of the planes VaP l9 VaP 6 , V'b'P l9 V'c'P.. The first two planes intersect in Fa, the second two in Va^. Therefore the intersection of these four planes will be the intersection of these two lines. It will evidently be a point, (>! (fig. 39a), lying below the plane of the lateral axes. The next angle will be the intersection of Vb\ and V'b' ; this will evidently be a point, Q 2 (fig. 396), corresponding in every way to Q lt but lying above the plane of the lateral axes. Fig. 39a. Fig. 396. And thus as we go round, the angles will be alternately above and below the plane of the lateral axes. The vertical edges will therefore be VQ 19 VQ^ VQ 2 , V'Q etc., and the lateral edges will be QiQ* etc. The form will be bounded by twelve equal and similar scalene triangles, and its lateral edges will run up and down in a zigzag fashion ; it is the Hexagonal Scalenohedron (fig. 40). A model of both Dihexagonal Pyramid and Scalenohedron may be constructed from Mr. Jordan's figures. The following are convenient dimensions : Dihexagonal Pyramid. Scalenohedron. Lateral edges Longitudinal edges 2-65 9-0 and 9'1 6-7 8 -5 and 108 The upper and lower halves of the Scalenohedron should be made separately. If one of these halves be slipped over the Pyramid, it can easily be squeezed into shape and its faces fitted on to the faces of the Mineralogy. 59 Pyramid ; they may then be retained by a few drops of gum on each face. The other half of the Scalenohedron should be kept loose, so that it may be slipped on and off at leisure. There is, as in the case of the Rhombo- hedron, a negative as well as a positive Scalenohedron. The Scalenohedron may also be obtained thus. Produce the vertical axis of- a Rhom- bohedron (nR) both ways to an equal extent till it is p times its original length. Join the ends of the line so obtained with the lateral angles of the Rhombohedron. The solid so formed is evidently a Scalenohedron. Its symbol is written riP . A second Hemihedral form of the Dihexa- gonal Pyramid is obtained by omitting every other face, the faces retained in the upper half of the form being vertically above those omitted in the lower half. Reasoning exactly similar to that em- ployed in the case of the Hemihedral form of the Hexagonal Pyramid will show that there are six lateral edges which run up and down in a zigzag way, and that the form is bounded by twelve deltoids. It is called the Hexagonal Trapezo- hedron. Its shape resembles that of the Tetragonal Trapezohedron (fig. 44), only it has twelve instead of eight deltoid faces. The larger h becomes, the more nearly parallel do the vertical edges of the pyramids become ; and when n is infinitely great these edges become parallel and the pyramid passes into a prism. Hence by putting h = in our symbols for the Hexagonal Pyramids we get a series of corresponding Hexagonal Prisms. In Naumann's symbols put m=ao. The symbols for these prisms are thus abbreviated -1 1 1 obj into [ / Fig. 40. oo r \ oo P2 V ooPraj of Naumann. jl 2 00} into [2 oo {1m 00} into [moo These prisms are sometimes bounded above and . below by planes parallel to the base ; sometimes they are terminated by Pyramids, by Rhombohedral faces, or by faces belonging to other forms of the system. Planes parallel to the base are denoted by the symbol for the base itself. This is parallel to the lateral axes, and cuts the vertical axis at a distance = 0, so its symbol may be written (oo GO 0) ; it is generally written, however (110), or [0], or OP. In the crystal of the Quartz on fig. 5 the full symbol would be {1 1 <*>} {1 1 A}, or [/] [n], or oo P, m P. The first denotes the Hexagonal Prism which forms the middle of the crystal, the second the Hexagonal Pyramid which caps its ends. 60 Geology. 3. THE DIMETRIC SYSTEM. There are three axes at right angles to one another, only one of which is a Principal Axis of Symmetry. Usually only two of the Parameters are equal. Exactly as in the case of the Hexagonal System we may show that whatever faces occur in one of the octants, a similar set of faces must occur in each of the remaining seven octants ; that the greatest number of faces in each octant is two ; and that the greatest number of faces a form can have is 2x8 = 16. We must here too agree upon a convention with respect to the Parametral Plane similar to that adopted in the Hexagonal System. We have already agreed to take a face of a Square Octahedron for that plane. Crystals of the same mineral often present several such octahedrons. One of these is taken as the Unit or Standard Octahe- dron, and one of its faces is taken for the Parametral Plane for that mineral. If the intercepts of this face on the lateral and vertical axes are as a : c, its symbol is (a a c) or (l 1 ) or (1 1 i) The symbol for any other octahedron will then be or {1 1 n} As in the Hexagonal System it is necessary, when we use the symbol under the last form, to recollect that n means n times ^ : the accent over the n serves to remind us of this. The holohedral forms of this system are so closely analogous to those of the Hexagonal System that with a few hints the student will be easily able to work them out for himself. Naumann here uses P and m in the same sense as before, and his symbols will explain themselves ; m P n means a form whose faces have intercepts in the ratio of a : n a : m ('. 1 P 1, written shortly P, is the Parametral or Unit Octahedron. The symbol {1 1 n}, [n], m P denotes an Octahedron in which the axes pass through the angles of the base, or of the first order (fig. 18). { 1 oo /i}, [oo n\, m P oo is an Octahedron in which the axes pass through the middle points of the sides of the base, or of the second order. We will now take the general form { 1 m &}, \m n], m P n. In each octant there must be the two faces (\ m A), (m 1 h), and it may be shown, by a construction (fig. 41) exactly the same as that Mineralogy. 6 1 used in the case of the Dihexagonal Pyramid, that the base of the form Fig. 41. is an equilateral Octagon, and that the form is bounded by sixteen equal and similar scalene triangles. The form is called the Ditetragonal Pyramid (fig. 42). The Hemihedral form of the Square Octahedron may be determined in ex- actly the same way as for the Regular Octahedron. If we draw a figure like that on fig. 22, only with the vertical axis of the Octahedron longer than the lateral axes, we can show, by exactly the same reason- ing we applied to fig. 22, that the edges of the Hemihedral form of the Square Octahedron are diagonals of the faces of the circumscribing rectangular prism. The upper and lower faces of this prism are equal squares, and their diagonals are equal : the vertical faces are equal rectangles, and their diagonals are equal to one another, but longer than the diagonals of the square faces. The form is therefore bounded by four isosceles triangles ; it is called a Sphen- oid. The Ditetragonal Pyramid furnishes two Hemihedral forms. One is ob- Fig. 42. 62 Geology. tained by omitting every alternate pair of faces and producing the faces retained till they meet. The faces retained in the upper half are vertically above those omitted in the lower half. In figs. 41 and 42 B is the octagonal base, V and V the vertices. We retain the shaded faces AVD. 2 , D. 2 VB', A'VD, DVB in the upper half, AVD^ DiVS, B'V'D,, V'D^A' in the lower half. This form, which is bounded by eight equal and similar scalene triangles, is the Tetra- gonal Scalenohedron or Dip- loid. It is an extremely difficult form to convey any notion of by words or plane figures, but its relations to the pyramid will be easily realized if a model be constructed. Fig. 43 shows the half of such a model. V and V are the ver- tices of the pyramid. The student will not find it difficult to draw a line BD^A D^B'D. A A 'D 4 round the model in such a way that the different pieces of it shall all lie in a plane perpen- dicular to VV and equidistant from V and V. This is the octagonal base of the pyramid. If VA', VB, V'B, VA, VA, V'B, V A' , VB' be drawn, and the shaded parts be coloured, these are the faces of the pyramid which are prolonged to form the Diploid, P 1 V'P S being the prolongation of V'D^A, P.V'P, of V'D lt PiVPi of VD 4 , and P 2 FP 4 of VD 4 A '. The suppressed faces are of course within the model. P l and P 2 lie on the plane VD^D.^ PZ and P 4 on the plane V'D^D^ The other hemihedral form is obtained by omit- ting every other face in the upper set and every other face in the lower set of octants, the faces retained in the upper set being vertically above those omitted in the lower. We retain VD^A, VD. 2 B\ VD.,A', VDB above, and V D^B, V'D^A, V'DJff, V'D.A below. Exactly as in the case of the Dihexagonal Pyramid and Hexagonal Trapezohedron it may be shown that there are four equal lateral edges run- ning up and down in a zigzag fashion, and the form is bounded by eight deltoids. It is called the Tetragonal Trapezohedron (fig. 44). Fig. 44. Mineralogy. 63 As before, there are positive and negative cases of these Hemihedral Forms. By putting n = oo in our symbols, m = oo in Naumann's, we obtain the symbols for the Dimetric Prisms. Prism of first order {1 1 oo } or [/], oo P. Prism of second order {1 oo oo } or [oo oo ], oo P oo. Ditetragonal Prism {1 m oo } or [m oo ] GO P ft. These prisms are open at top and bottom ; they may be closed by basal planes [0], OP, or terminated by faces belonging to various forms of the system. 4. THE TRIMETRIC SYSTEM. There are three axes at right angles to one another, each of which is an Ordinary Axis of Symmetry. The parameters are usually all unequal. There is nothing geometrically to distinguish one axis from another and one is arbitrarily chosen as the vertical or longitudinal axis. The Parametral Plane is a face of a Rhombic Octahedron, and it is determined by the following convention. Among the crystals of the same mineral several such Octahedrons often occur. One is chosen as the Standard or Unit Octahedron and one of its faces is taken for the Parametral Plane for that mineral. Let the intercepts of this face on the axis be in the ratio of a : b : c, c being proportional to the intercept on the vertical axis, and b being greater than a. Then the symbol for this face is ( 6 =), or (1 6 j). The symbol for a face of any other Octahedron will then be (1, mb_ n c -\ or (1 m n), \ ct? & / where m stands for m times -, n for n times -. a' a Since none of the axes are interchangeable, symmetry does not require a form to have more than one face in each octant. The following are the more important forms. (1) {!*}. zOy is a plane of symmetry, therefore if we have the face (1 m ti) in the first octant, we must have the face (1 mil) in the second. zOx is a plane of symmetry, therefore if we have these faces in the first two octants, we must have the faces (1 m h), (1 m n) in the third and fourth octants. xOy is a plane of symmetry, therefore if we have the above four faces in the upper four octants, we must have in the lower four octants the corresponding faces (1 m h), (1 mti), (1 m n), (1 m n). These are all the faces required by symmetry, and the form is there- fore bounded by eight equal and similar scalene triangles. The base is a rhombus. The form is therefore a Right Rhombic Octahedron. 64 Geology. In fig. 45 Ox and Oy are the lateral axes, OA=a=OA', OB = 01? = b, Ob = Ob' = ml. Ab, bA', A'b', b'A are evidently the intersections of the eight faces of the Octahe- dron with the plane yOx. The general formula {1 m n} includes two cases which it is convenient to distinguish from one another. In the first the intercepts of a face are in the ratio of n : mb : nc. The symbol for this face we write { 1 m n}, or shortly [m n\. In the second the intercepts are iii the ratio of Fig. 45. ma : b : nc, and the symbol in this case will be {m 1 &}, which in accordance with the usual convention we also abbreviate into [m n]. Now we have agreed, when we made b greater than a, that Ox shall coincide with the Brachydiagonal and Oy with the Macrodiagonal of the Rhombic Base of the Octahedron ; and if we stick to our convention that the index placed first belongs to Ox and that placed second to Oy, the full-length symbols explain themselves. But in the abbreviated symbols some additional mark is needed to distinguish between the two classes of Octahedra. The distinction is made thus. The forms whose faces have intercepts in the ratio of fi : ml : nc are denoted by [m n] or m P n. The forms whose faces have intercepts in the ratio of ma : b : nc are denoted by \m n\ or m P n.* If we make m and n= 1 in the symbol {1 m ,}, we get the Unit Octahedron {1 1 1}, or [1], or P. By making n= GO, we get the symbols for the Right Rhombic Prisms corresponding to the two classes of Octahedra, viz. : {1 m oc}, or \m oc]. or GO JPn. \m 1 oo }, or [m GO], or oc Pn. If m l, we have the Unit Prism (1 1 x}, or [/], or x P. As before, the basal planes are denoted by [0] or <>P. (2) {m*/V}. Since zOy is a plane of symmetry, if we have the face (m GO n), we * Breithaupt writes these m I' fi, m P it ; and though this notation has not come into general use, it has certain advantages over Naumann's. Mineralogy. 65 must have also the face (m oo ?k) ; and since xOy is a plane of sym- metry, we must also have the faces (m oo h), (m oo fi). But if these four faces are present, all the requirements of symmetry are satisfied. The form is therefore composed of these four faces alone. It is evi- dently a prism whose axis is yOy' and base a rhombus, and the faces do not enclose a space. Such forms are called "open forms." A similar form is denoted by the symbol {oo m h}, which is a rhombic prism whose axis is Ox. If these two forms are combined, they give a rectangular octahedron, so situated that the axes pass through the middle points of its lateral edges. It not unf requently happens that in a combination the faces belonging to one only of these forms are present. These faces then resemble the sloping sides of the roof of a house, and they intersect, or would inter- sect if produced till they met, in a line like the ridge of a roof. Hence such faces are called " Domes." These forms may conveniently be distinguished thus. When the faces are parallel to the Brachydiagonal of the base of the Unit Octa- hedron they are called " Brachydomes," and the mark ^ is placed over GO ; when parallel to the Macrodia- gonal they are called "Macrodoines," and the mark - is placed over oo. The symbols then are Brachydome {m 06 ?i}, m P oo.* Macrodome {m 66 ?'*.}, m P oo.* In fig. 46 OA = OA'=a, Oa=0a 46 - The faces DEKH, FGHK, DELM, FGML are Macrodomes. If n = oo , the symbol becomes \m oo 00} or more simply {1 oo oo }, or [oo oo] ; for the planes represented by (m oo oo), (1 oo oo) are parallel, and therefore crystallographically identical. This form, it is evident, will consist of two planes parallel to zOy, and intersecting xOx at equal distances on opposite sides of 0. Faces parallel to the Axial Planes are called Pinacoids. Pinacoids parallel to the Brachydiagonal are called Brachypinacoids, and their symbol is {1 oo oo}, or [do oo], or oo Ao ;t if parallel to the Macrodiagonal, they are Macropinacoids, and their symbol is {1 66 00} or [oo oo], or oo P oo.t The two Pinacoids, if combined, would form a right prism on a rectangular base, open at the ends. This prism may be closed by * Possibly these symbols might be better written m P QC , m P 66 . t These symbols might also be written, oo P 06 , oo P 56 . 66 Geology. faces parallel to the plane of the lateral axes, which are called Basal Pinacoids. The student must carefully realize how the inferior degree of sym- metry in the Trimetric System causes its forms to differ in so striking a way from those of the first three systems. In these all the forms were closed forms except the Prisms, and they are open only at their ends. Here there is only one closed form, the Rhombic Octahedron, to correspond to the closed forms of the three previous systems, and only one prism, the Rhombic Prism, to correspond to their prisms. Domes are horizontal prisms open at their ends, and Pinacoids consist merely of a pair of parallel planes. Also the fact that the lateral axes are not interchangeable makes it impossible that any simple forms should exist in this system corre- sponding to the Dihexagonal and Ditetragonal Pyramids and Prisms. By combining Domes with a Rhombic Octahedron we may get a Pyra- mid with sixteen triangular faces, but it will differ essentially from the Ditetragonal Pyramid, in that it is a combination and not a simple form. If we construct a figure similar to that in fig. 22, making OA, OB, OC all unequal in length, we can show, by exactly the same line of reasoning as was used in the case of the Regular Octahedron, that the Hemihedral form of the Rhombic Octahedron is bounded by four similar triangles, and that its edges are diagonals of the faces of the circumscribing right prism. These diagonals will however be all unequal in length, and the faces will be scalene triangles. The form is called a Rhombic Sphenoid. 5. MONOCLINIC SYSTEM. The angles between the axes are two of them right angles and one an oblique angle. The parameters are usually all unequal. The axes are generally so placed that the lateral axes Ox, Oy shall be at right angles to one another, and the vertical axis Oz shall be oblique to Ox. Ox is distinguished as the Clinoaxis, and Oy as the Orthoaxis. The convention with regard to the Parametral Plane is exactly the same as that adopted in the three preceding systems. If its intercepts are in the ratio of a : b : c, the symbol of any face is (1 m ?'i) where m stands for m times JL n for n times C. a a In abbreviated symbols of the form [m n] the index that belongs to the Clinoaxis is distinguished by having a sloping accent placed over it thus, m. Naumann makes the distinction by drawing lines across or over P thus, JP *, or P P. Thus [m n\ or n I* m* are equivalent to [ma, b, nc\, and [m n] or n P m* to \a, mb, nc\. There is only one plane of symmetry, viz. that which contains the * These symbols might also be written n P m, n P m. Mineralogy. 67 two axes which are oblique to one another. The forms however are, as all crystalline forms must be,* symmetrical about a point, and this point we shall take for the intersection of the axes. This degree of symmetry involves the following consequences. Suppose we have in the first octant a face (1 m n) : since zOx is a plane of symmetry, there must be a corresponding face on the other side of this plane, which will lie in the second octant, and whose symbol will be (1 m n). If these two faces are present, all the requirements of symmetry about this plane will be satisfied. But in order that the form may be symmetrical about 0, we must have a face in the seventh octant to correspond to (1 m n\ and one in the eighth octant to correspond to (1 m n). The symbols of these faces will be (1 m n), (I m n). These four faces fulfil all the requirements of symmetry, and the form therefore contains these alone. It is evidently an open form. A second form will contain the four faces (1 m n) 9 (1 m n), (1 m n), (1 m n), lying in the third, fourth, fifth, and sixth octants. A combination of these two forms will give a Monoclinic Octahe- dron. Each of them is called a Hemioctahedron ; the first lies in those octants in which the angle between the vertical and clino-axes is acute, and it is distinguished as the positive Hemioctahedron ; the second is the negative Hemioctahedron. In these Hemioctahedrons the intercept of each face on the Clino- axis is to its intercept on the Orthoaxis as a is to mb ; their symbols are + {1 m n} +\mn\ + m f n (1) -{Ira n} -[mn\ -mfn (2) There will also be a pair of Hemioctahedrons in which the intercept of each face on the Clinoaxis is to the intercept on the Orthoaxis as ma is to b, and their symbols are + imln\ +\mn\ +mPn (3) - {m In}' ~[m n] -m P n (4) The Parametral or Unit Hemioctahedrons are + {1 1 1} or +[1] -{111} or -[1] or +P. or -P. The Prismatic forms, it is easy to see, are not Hemiprisms. For if we have in the first octant the face (1 m oo), we must have in the second the face (1 m GO), because the form is symmetrical about zOx. We must also have the faces (I m oo), (1 m oo), because it is sym- metrical about 0. These four faces bound a Monoclinic Prism open at the ends. It In the preceding systems the forms were necessarily symmetrical about the intersection of the axes. 68 Geology. may be closed by planes parallel to yOx, or by basal planes, the symbols for which are written (o 1 1) or [0] or o P. The symbols then for these prisms, which are called prisms of the first order, are (1 m 00} [moo] oo f m. {m 1 00} [moo] GO 1* m. The Parametral or Unit Prism is {1 1 00} [/] oo P. Pinacoids. The face (oo 1 oo) is a Pinacoid, and as it is parallel to the plane containing the vertical and clino-axes, it is called a Clino- pinacoid. If we have also the face (oo 1 oo), symmetry will be ensured both about zOx and about 0. The Clinopinacoid form therefore consists of the two faces (oo 1 oo), (oo 1 oo ), and its symbol is {oo 1 co} or [cb oo ] or oo P oo. In the same way we may show that the Orthopinacoid consists of the faces (1 oo oo) (1 oo oo), and its symbol is {1 oo 00} or [oo oo] or oo f oo. The Clinopinacoid and Orthopinacoid forms, when combined, enclosed a prism of the second order, open at top and bottom. It may be closed by basal planes [0]. Domes are Clinodomes or Orthodomes according as they are parallel to the Clino- or Ortho-axis. Orthodomes. If we have the face (1 oo n\ it is necessary that there should also be the face (1 oo n) to ensure symmetry about 0. But symmetry requires no other faces. One Orthodome then consists of these two faces only; it is dis- tinguished as the positive Orthodome. Another Orthodome, consisting of the faces (1 oo n), (1 oo n), is the negative Orthodome. The symbols for the Orthodomes then are 4- {1 oo n\ + [co n\ . + {1 oo n} [oo n\ _ + m oo. m P oo. Clinodomes. Reasoning exactly similar to that we have been using will show that in the case of a Clinodome symmetry requires the follow- ing four faces : (oo 1 n), (oo I n), (oo 1 n\ (co 1 n), which will enclose a Rhombic Prism whose axis is Ox, open at the ends. It may be closed by Orthopinacoids. The symbol for Clinodomes will be therefore {oo 1 n} or [ob n\ or m P oo. 6. THE TRTCLINIC SYSTEM. The angles between the axes are all oblique, and the parameters are usually all three unequal. Mineralogy. 69 One axis is arbitrarily chosen for the vertical axis. There is symmetry only about a point, which we shall take for the intersection of the axes. The requirements of this type of symmetry are most simple. If any face is present it is only necessary that there should be a face parallel to it on the opposite side of 0, and equidistant from it. Each form then consists of only two parallel faces. Octahedral Forms. Suppose that we have in the first octant the plane (1 mn); symmetry requires that we should have in the octant diagonally opposite, that is the seventh octant, a parallel plane, whose symbol will be (I m n). But this is all that symmetry requires, and consequently the complete form consists of these two_ planes only. A second form will consist of two planes (1 m n), (I m n) in the second and eighth octants ; a third form of the planes (I m n), (1 m n) in the third and fifth octants ; a fourth form of the planes (I m n), (1 m n) in the fourth and sixth octants. A combination of these four forms will enclose a Triclinic Octa- hedron. Since each form contains one-fourth of the bounding faces of the Octahedron, each is called a Tetarto-octahedron. Of the different Octahedrons met with among the crystals of any mineral one is chosen as the Unit or Standard Octahedron, and one of its faces is taken for the Parametral Plane. If the intercepts of the Parametral Plane on the two lateral and the vertical axes are in the ratio of a : b : c, we will take b to be greater than a, and we will call the longer diagonal (b) of the Rhomboidal Base the Macrodiagonal, and the shorter (a) the Brachydiagonal. The following are some of the methods devised to distinguish the tetarto-octahedral forms. Take first the Unit Octahedron {a b c} or {1 1 1}. 1st Form, the faces (1 1 1), (1 1 l)in the first and seventh octants [l]'or +[!]', orP'or +P. 2nd Form, the faces (1 I 1), (I 1 1) in the second and eighth octants '[1] or +[1], or 'Por + P. 3rd Form, the faces (1 1 1), (1 1 1) in the third and fifth octants [11 or -[!]', orP /0 r -P. 4th Form, the faces (1 1 1), (1 I 1) in the fourth and sixth octants /t 1 ] or - [1] or , p or - p > A similar notation will distinguish the more general forms, whose faces have symbols of the form (1m n): -, -p ( (1 m n) (I m n\ [m n\' or + [m nj or m P'n or + m P'n. ((ml n) (m I n), [m n\' or + [m n]' or m ^'ra, or + m P'n. 2nd Form | (^ mn ) (!*)> '\fnn\ or +[mw] or m'Pn or +mPn. ( (m I 7i) (m 1 w), '[mn] or +[m7i] or m'Pn or +mPn. 3rd Form I (^ m w ) (1 m ^) [ m w ]/ or ~ [^ n ]' or m ^/^ or - m P'n. \ (ml n) (m 1 n) t [m n\ t or - [m n]' or m J 5 ^ or - m P'n. (I m n) (1 m n), y [m ^] or [m TI] or m t P n or m P n. (m In) (ml n), \m n] or - [m n] or mP n or mPn. 70 Geology. To follow these symbols, first bear in mind that in the parentheses ( ), which enclose symbols of faces, - over an index is a negative sign. In the square brackets over an index shows that the intercept is measured along the macrodiagonaL In the abbreviated symbols for forms, enclosed in brackets [ ], and in Naumann's symbols, - and ^ are used exactly as in the Trimetric System. m P n and m P n might possibly be better written m P n, m P ft, but we have adopted the general usage. Further, in the first and third columns the 1st Form is distinguished by an accent above and on the right hand of the symbol ; the 2nd Form by an accent above and on the left ; the 3rd Form by an accent below and on the right ; the 4th by an accent below and on the left. In the second and fourth columns the 1st and 2nd Forms are reckoned positive, and the 1st only is accented ; the 3rd and 4th are reckoned negative, and the 3rd only is accented. Domes will be Hemidomes, and their symbols will be Macrodomes_ (1 oo n) (1 oo ?i), [56 n]' or + [66 n\ or m P'oo or +m P oo. (1 oo n) (I oo n), [oo n]' or - [06 ri\ or m P^cc or -m P oo. Brachydomes (oo 1 n) (oo I n), [oo n\ or + [co n\ or m P'oo or +m P oo. (oo 1 n) (oo 1 n) t [06 n~\ l or - [06 n\ or m P t or - m P oo. The prisms will be Hemiprisms, and their symbols may be written (1 m oo) (1 m oo), \m oo]' or oo P'n. (1) (1 m oo) (I m oo ), '[m oo] or oo'P n. (2) (m 1 oo) (m 1 oo), [ra oo], or oo P t n. (3) (m 1 oo) (m\ oo ), \m oo] or vofn. (4) Fig. 47. Fig. 47 shows the intersections of the faces of the two Hemiprisms (1) and (2) with the plane of the lateral axes. The student should draw a similar figure for (3) and (4). The Macropinacoid is (1 oo oo) (1 GO oo ), [do oo] or oo P oo. The Brachypinacoid (oo 1 oo) (oo I oo), [0600] or oo P GO. The Physical Constitution of Crystals. The preceding pages contain an outline of the more important points in the Geometry of Crystals. Mineralogy. 7 1 But we should be treating the reader very unfairly if we allowed him to go away with a notion that Crystallography was nothing more than a body of neat geometrical artifices for facilitating the study and description of the shapes which crystals assume. Underlying the geometrical facts with which we have become acquainted are others of far deeper moment, for they relate not to mere external shape, but enable us to see into the very heart of crystallized bodies and to realize that the molecules of which they are composed are not thrown together without order, but are built up according to definite plans. Among the facts in question the most striking and suggestive are the effects which are produced on light when it passes through transparent crystals. When a ray of light passes from air into glass or any non- crystallized transparent substance, its path becomes bent or refracted, each incident ray gives rise to a single refracted ray, and the course which the refracted ray takes is determined by a very simple law. It matters not in what direction the incident ray falls on the glass, there will never be more than one refracted ray, and the law which determines its direction will be always the same. This will also be the case when light passes through a crystal belonging to the Monometric System. But light is refracted after a different fashion when it passes through crystals belonging to any of the other systems. A ray on entering these crystals is, unless it travel through the crystal in certain definite directions, split into two rays and is said to be doubly refracted. In some crystals there is only one direction in which the ray can escape double refraction. A line parallel to this direction through the centre of the crystal is called its Optic Axis, and the crystal is said to be Uniaxal. In Uniaxal crystals one ray, called the Ordinary Ray, is refracted after the same law as holds good for glass; the other, or Extraordinary Ray, is refracted after a different and more complicated law ; all rays however which are equally inclined to the Optic Axis are refracted in the same manner. In the case of other crystals, distinguished as Biaxal, there are two Optic Axes, and rays traversing the crystal parallel to either of these do not suffer double refraction. When a ray is doubly refracted by Biaxal crystals, neither of the refracted rays obeys the law of refraction which holds good for glass. Now all these properties are most closely related to the Crystalline Symmetry which the crystals possess. All crystals belonging to the Monometric System are singly re- fracting, those belonging to the other systems are all double refracting. Of doubly refracting crystals those which have only one Principal Axis of Symmetry are Optically Uniaxal, and their Principal Axis of Symmetry is the Optic Axis. The fact that all rays which are equally inclined to the Optic Axis are refracted in a similar manner shows that such crystals are Optically as well as Geometrically symmetrical about their Principal Axis. Again all crystals which have no Principal Axis of Symmetry are Optically Biaxal. In the Trimetric System the Optic Axes lie in one of the Axial Planes, and one Crystallographic Axis bisects the acute angle, and another the obtuse angle, between the Optic Axes. In TI. 73 Geology. Monoclinic crystals the Optic Axes lie either in the plane containing the two Crystallographic Axes which include an oblique angle, or else in planes at right angles thereto. In the Triclinic System the relation between the Crystallographic and Optic Axes has yet to be discovered. In all cases rays of light which traverse a crystal in parallel lines are refracted in exactly the same manner. Again if we investigate the rate at which Heat travels through crystals and their relations to Electricity and Magnetism, or if we determine their cohesion and hardness, we find that these are the same along parallel lines but different in different directions, and the variations are closely related to the crystalline shape. In the case of cohesion an important result follows from this. Planes of Cleavage are planes of minimum cohesion, and since there is a close connection between the degree of cohesion in different directions and the crystal- line shape, there must necessarily be an equally close relation between Cleavage and crystalline shape. This, we have seen, is the case. There is therefore the closest possible connection between the sym- metry and the physical properties of crystallized substances. This and other facts have led to the following hypothesis about the constitution of substances in general. All bodies are made up of particles almost inconceivably small called Molecules.* Even in the densest bodies the molecules are not in contact, there are spaces between them. Also the molecules are in incessant motion, but in solid bodies their excursions are confined within very narrow limits ; they oscillate, rotate, or gyrate about points which are called their mean positions. The spaces between the mole- cules are filled with a substance called Ether, and the sensations of Light and Heat are caused by vibrations of this ether. Electric and Magnetic phenomena are also probably due to the same ether. Now all the physical properties mentioned can be accounted for if we make the following suppositions about the way in which the mole- cules are arranged in crystals. Monometric crystals we must suppose to be what is called Isotropic, that is, in whatever direction you travel through them, the distance between the mean positions of two adjacent molecules is the same. In all other crystals we have only to assume that the distance between the mean position of two adjacent molecules is the same along all parallel lines, but is not the same in directions which are inclined to one another. In a word, the molecules are more thickly packed in certain directions than in others. Such bodies are called Aniso- tropic. It is by an investigation of the optical properties of crystals that we can deter- mine with the greatest certainty to which system they belong. There are cases where their geometrical shape * A molecule is the smallest portion of any substance which can exist separately without losing the properties which belong to and distinguish the said substance. .&&ufnfr V V OF THE ' Mineralogy. |U NI V BIlS IT1 leaves it doubtful to which of two systems certain crystals are to referred. Suppose, for instance, that we have to deal with a righ prism whose cross section is a regular hexagon ABCDEF '(fig. 48). It may belong to the Hexagonal System. But geometrically there is nothing to forbid its being Trimetric. For produce AB, DC to meet in G, and AF, DE to meet in H. Then AHDG is a rhombus. The faces A, CD, DE, FA may be those of the Unit Prism ; the faces BC, FE may be Brachypinacoids. Cases occur in nature of crystals where the angle at A is so nearly 120 that it is difficult to say from a geometrical point of view whether they should be looked upon as Hexagonal or Trimetric. But optical examination settles the point. If they have two optic axes and neither of the doubly refracted rays follows the ordinary law of refraction, they are Trimetric. If they have only one Optic Axis, they are Hexagonal. Trimetric crystals which assume shapes indistinguishable geometrically from those of the Hexagonal system may be called Pseudo-hexagonal. Instances of Combinations in Crystals. We have, in the sketch of Crystallography just given, laid ourselves open to a charge of gross inconsistency. We stated expressly at the outset that the crystallographer cared only for the angles between the faces of crystals, and was not concerned with the size and shape of the faces ; and yet we have gone on page after page determining with the utmost care the shape of the faces of the forms we have been dealing with. It is true that many of the results we have worked out are crystal- lographically of minor importance, but they have their use for all that. The careful mastery of every point connected with the Simple Forms, with which we have been almost exclusively concerned, is most valuable as a preliminary training ; it begets a geometrical habit of thought and gives a facility in dealing with the geometry of solid bodies, which the student will find of the utmost assistance when he comes to work at the complicated combinations of these simple forms which natural crystals usually present. For an account of such combinations the reader must consult works specially devoted to Crystallography. By way of illustration we give here a couple of examples. Fig. 49 represents a crystal of Heavy Spar. On holding it up to the light two sets of parallel cracks can be seen running through it perpendicular to the face 0. These are planes of cleavage. S7~ The faces / are parallel to these ^ "* * cleavage planes. The crystal f also cleaves parallel to the face LP * * The sloping face at the right-hand end is similarly a Macrodome cut- ting the vertical axis at a distance proportional to J c, and its symbol is [co 1]. The crystal then has the faces of the Unit Prism, the Unit Octa- hedron, the Basal faces, Brachypinacoids, Macropinacoids, Brachy- dornes, and Macrodomes. Its symbolical description is [7] [1] [0] [] [ oo] [*J] [doj]. Fig. 50 represents a shape very often assumed by crystals of Ortho- clase. On the faces are placed the symbols we have been using and those of Naumann as well. The two faces marked / or GO P, with the corresponding faces behind, and the faces marked or oP, would if produced enclose a prism on a rhombic base. The long upright edges are per- pendicular to one diagonal of the base and inclined to the other. The prism is therefore Monoclinic. The vertical axis is parallel to the long upright edges : the clinodiagonal is parallel to the line AB or Fig. 50. ED : the orthodiagonal, not shown in the figure, is at right angles to these lines. The faces / belong to the Unit Prism, and are the basal faces. The face cb oo is parallel to the vertical and clino-axes, it is therefore a Clinopinacoid ; two such faces only are present, and the student will recollect that in the Monoclinic System symmetry is satisfied if only Clinopinacoids are present : the Orthopinacoids need not necessarily Mineralogy. 75 be present as well. The faces oo 2 or 2 Poo are parallel to the Ortho- axis ; their intercepts on the vertical axis are twice the vertical para- meter of the mineral ; they are therefore Orthodomes. There are only two of them, because symmetry requires only two. If Clinodomes had occurred, there must have been four of them. The crystal there- fore possesses the lateral and basal faces of the Unit Prism, Clinopina- coids, and Orthodomes. Its symbolical description is or W [o 00 P. OP. [ob oo] [oo 2] O P 00. 2 POO. Twin Crystals. We frequently meet with crystals whose shapes strike us at first sight as having something peculiar about them. They are not any of the simple forms we have become acquainted with, nor can we arrive at similar shapes merely by combining these forms. We can however imitate the shape taken by these crystals in the following way. We take a certain crystal, simple form or com- bination as the case may be, divide it into two parts along a certain plane, turn one of the parts round a certain axis through 180, and then reunite the two parts along the divisional plane. Such crystals are called Twin Crystals. It is not for a moment to be supposed that any division and rotation like that just mentioned has ever taken place ; we merely say that we could make a model of the crystal by this process, not that nature formed it after this fashion. The plane along which the crystal is divided and along which the two parts are united, is called the Composition Plane. The axis about which the imaginary rotation takes place is called the Twin Axis, and a plane perpendicular to it the Twinning Plane. The following are instances of Twin Crystals. Let the student draw on cardboard the "net" in fig. 51 : the figure at the top is a regular hexagon ; the smaller triangles are equilateral Fig. 51. and their sides are equal to a side of the hexagon : the larger triangle is also equilateral and its sides are double those of the hexagon. 76 Geology. Fig. 52. When the net is cut out and pieced together a solid is formed of which one face is a regular hexagon, the opposite face an equilateral triangle, and the remaining faces equila- teral triangles and trapeziums* alternately. Two of these solids should be made. If held with the hexagonal faces in contact they may be so placed as to form a Regular Octahedron, divided into two parts by a plane parallel to one face. Now turn one of the solids round an axis perpendicular to the hexa- gonal face through an angle of 60 ; we get the solid shown in fig. 52 : another turn of 60 gives us the octahedron ; another turn of the same amount, 180 in all, brings us again to the shape in fig. 52. This is a very common Twin Crystal in some Monometric Minerals. In this case the Composition and Twinning Planes are the same, and are parallel to a face of the Octahedron. We obtained this twin by a revolution of 60 only ; but as in most cases a revolution of 180 is required, and as in this case that revolution serves equally to give the twin, for uniformity's sake we consider it pro- duced by the larger turn. As another illustration we will take a Twin Crystal common in Orthoclase. Suppose the crystal in fig. 50 to be cut in two along the plane ABCDEF. Let the hinder half remain fixed ; let the front half be moved parallel to itself towards the spectator ; then let it be turned through 180 round an axis perpendicular to the plane containing the vertical and ortho-axes. If GH be perpendicular to the long upright edges, this axis of revolu- tion is parallel to GH. Now move the half that has been revolved parallel to itself back again and unite it to the hinder half. We get a crystal of the form shown in fig. 53, where A', C', etc., denote the points that coincided with A, C, etc., before the revolution. This is known as the Carlsbad type of Fig. 53. Twinning. Here the composition plane is parallel to the Clinopinacoid ; the Twinning Plane is parallel to the Orthopinacoid. The two planes are therefore not the same as in the last example. * A trapezium is a four-sided figure, two of whose sides are parallel and the other two not parallel. Mineralogy. 77 In fig. 53 the parts of the Composition Plane that are visible are shaded. Fig. 54 is a net by which the student may construct a model of a Unit Prism of Orthoclase with Clinopinacoids and the Hemiorthodome Fig. 54. Ac, jl oo 1|, [oo 1], twinned on the Carlsbad type. The net should be drawn very carefully on a piece of cardboard in the following manner. Construct the rectangle aBCD, making .#(7 = 3, C = 5. Produce Da to A, making Act = 2. Produce AD to F, making DF= 4-7, and BC to E, making CE= 1 -6. From A draw ^#=5-9, making the angle BAH =76; from B draw BG parallel to AH and = 7'7. From B draw J3& = 8'4, making the angle #=64: draw kB' parallel to BC : in Bk take BI= f l'7 : with centre / radius = 1 '6 describe a circle cutting kB' in C' : make C'B' = CB : draw B'k' parallel to Bk and in it take B'M=BI. Then construct the figure B'G'H'A'D'F'E'C' in the same way as BGHADFEC, placing it as in the woodcut. From A' draw An = 8'4: making the angle D'A'n = 6, and through n draw nP parallel to A'D'. In A'n take A'N=5'$, and with centre N radius = 4'7 describe a circle cutting nP in 0. Make OP = A'D' and through P draw PQ parallel and equal to A'N. The figure contained by the strong lines is the net required. The positions of the laps for fastening are also shown. Next with a fine needle prick through the angles on to another bit of cardboard, cut out, and put together. This is one half of the model. Now turn the cardboard pattern over so that the face which was before uppermost is underneath, again prick through the angles on to a bit of cardboard, cut out, and put together. This will give the other half of the model. 7 8 Geology. The two halves can be placed together so as to form a crystal sjich as that represented in fig. 50, only it will have the Orthodome P oo instead of 2 P oo. By rotating one-half in the manner described a model of the twin is obtained. Orthoclase is also twinned, though not so frequently, ^according to a type known as the Baveno type. Here the Clinodome 2 Poo, {oo 1 2(, oo zj is both twinning and composition plane. A model for a Unit ism twinned on this type may be made from the net in fig. 55. The net is thus constructed. Draw the two lines POP\ QOQ', Fig. 56. Fig. 57. making the angle QOP = 70 30', OP = OP= 5, OQ = OQ' = 57. Construct on a separate bit of paper (fig. 56) the rhombus abed, in which Mineralogy. 79 dab = 67 and 5 = 9'1 ; bisect cb, cd in e and/, and/*? in g. Cut out the two figures cfe, abefd. Place cef so that g is on $' and e/ parallel to PP'. Let CE'F' be the position then occupied by cef. Similarly place abefd so that g is on Q and e/ parallel to PP', and let ABEFD be the position it then occupies. Next on another bit of paper con- struct the parallelogram liklg (fig. 57), in which rM = 67, M = 9'l, A/7 = 8'4; bisect Ig, Ik in e and p ; cut out the figures ep, likpeg. Place / coincides with EP, and the longer side el is adja- cent to EB, and place hkpeg so that e^> coincides with jfiTP' and the side e<7 is adjacent to E'C '; in this way we obtain the faces LEP, HGE'P'K of the net. The two faces L'FP, H'G'F'PK' may be ob- tained in the same way from the parallelogram hklk by bisecting two sides containing an obtuse angle, cutting it along the line joining the bisecting points, and fitting the two pieces on as in the woodcut. In making the two halves of the model proceed exactly as in the case of the Carlsbad Twin. Law connecting Crystalline Form and Chemical Com- position, and exceptions to it. We have already given a gen- eral explanation of the way in which a knowledge of the shape they assume when crystallized, enables us to recognise minerals. Subject to certain exceptions to be noticed directly, the following law holds good. A substance which has a definite chemical composition always crystallizes either in the same form, or in forms which belong to the same system and have the same parameters. Very frequently a given substance has so to speak a partiality for certain forms or combinations of the system, and its crystals assume these forms or combinations oftener than the other forms of that system. For instance, there are two minerals, Pyromorphite and Mimetite, that crystallize in almost exactly the same Hexagonal forms. But the hexagonal prisms of Pyromor- phite are usually terminated by basal planes, while those of Mimetite in most cases are capped by hexagonal pyramids. The exceptions to the law just stated are these. 1. Polymorphism. Certain substances, while retaining the same chemical composition, are capable of assuming crystalline shapes be- longing to two or more systems, or of assuming shapes which, though they belong to the same system, have different parameters. When the crystalline shapes are two in number this is called Dimorphism ; when three, Trimorphism ; and generally Polymorphism. Polymorphic forms differ not only in crystalline shape, but usually in hardness and other physical qualities as well. Carbonate of Lime is Dimorphic. It crystallizes in forms belonging to the Hexagonal System, when it has a hardness denoted by 3 * and perfect rhombohedral cleavage. It is then called Calcite. It also crystallizes in Trimetric forms, when it has only very imperfect cleav- age and a hardness of 3*5 to 4. It is then called Aragonite. Titanic Oxide (Ti0 2 ) is Trimorphic and assumes the following shapes. Entile. Dimetric, | = 0'65. Hardness 6 to 6 '5. Anatase. Dimetric, -=1'78. Hardness 5*5 to 6. * See p. 87. 8o Geology. Brookite. Trimetric. Hardness 5 '5 to 6. Silicate of Alumina (SiO 2 Al 2 O 3 ) is also Trimorphic. Its forms are Andalusite. Trimetric. Hardness 7*5. Sillimanite. Monoclinic. Hardness 6 to 7. Cyanite. Triclinic. Hardness 5 to 7 '25. 2. Isomorphism. A second exception to the law connecting crystal- line form and chemical composition arises from the fact that it is possible to remove wholly or in part one of the elements of a body and put certain other elements in the place of the one removed without altering the crystalline shape. Elements which can thus replace one another are said to be isomorphous. A very excellent instance is furnished by the Carbonates of Lime, Magnesia, Iron, Zinc, and Manganese. They all crystallize in Rhom- bohedral forms, and the following table shows that the angle of the rhombohedron is nearly the same for each. Angle of the terminal edges. CaCO 3 105 MgCO, 107 FeC0 3 . . . . . 107 ZnCO 3 107 40' MnCO 3 106 5' It makes practically no difference in the crystalline form whether the base is Lime, Magnesia, Iron, Zinc, or Manganese. These elements are therefore isomorphous. What is more, these bases may all or any number of them be present in the same substance without sensibly altering the crystalline form. Thus the angle of the rhombohedron for the Double Carbonate of Lime and Magnesia (CaCO 3 ,MgCO 3 ) is 106 15'. We may in fact express all these substances by a single chemical formula. Such a formula can be written in two ways. We may write it R.C0 3 , and state that R stands for any or all of the symbols Ca, Mg, Fe, Zn, Mn. Or we may write it (Ca:Mg:Fe:Zn:Mn)CO 3 . The number of 'atoms of each base that is present is variable, but each base enters in the proportion of its atomic weight, and the bases taken altogether are in the right quantity to combine chemically with C0 3 . The minerals Apatite, Pyromorphite, and Mimetite furnish another illustration of Isomorphism. They all crystallize in hexagonal forms, and the values of ^ are for Apatite -735 Pyromorphite . . . . . '736 Mimetite 727 Mineralogy. 8 1 The crystalline forms of the three are therefore practically the same. Apatite is sometimes 3Ca 3 P 2 8 + CaCl 2 , sometimes 3Ca 3 P 2 O 8 + CaF 2 , and sometimes Chlorine and Fluorine are both present. Chlorine and Fluo- rine are therefore isomorphous. Pyromorphite is 3Pb 3 P 2 O 8 + Pb (C1:F) 2 . Pyromorphite therefore differs from Apatite only in containing Lead instead of Calcium. Lead and Calcium are therefore isomorphous. Mimetite is 3Pb 3 As._,O 8 + Pb(Cl:F). 2 ; comparing this with Pyromorphite, we see that Phosphorus and Arsenic are isomorphous. The general formula 3(Ca:Pb) 3 (P:As) 2 O 8 + (Ca:Pb)(Cl:F)., includes all three minerals. The isomorphism is still more strongly shown by the fact that there are varieties intermediate between Apatite, Pyromorphite, and Mimetite, which contain Calcium as well as Lead, and Arsenic as well as Phosphorus. 3. Pseudomorphism, This is the third exception to the general law. It occurs when a crystal has the crystalline form characteristic of one mineral and the chemical composition of another. For instance Calcite crystallizes in rhombohedrons, Quartz in six-sided prisms ; we do find however crystals of Quartz having the exact shape and angles of a rhombohedron of Calcite. Such a crystal is called a Pseudomorph, and in the case mentioned would be described as a Pseudomorph of Quartz after Calcite. Pseudomorphs are arranged in the following classes according to their mode of formation. A. Displacement Pseudomorphs. (A a) By incrustation, when one mineral has coated over a crystal of another mineral. (A b) By replacement, when the substance of one mineral has been removed, and its place taken by another mineral, the substitution having proceeded molecule by molecule, so that the crystalline form and sometimes the cleavage of the first mineral is retained.. B. Alteration Pseudomorphs. (B a) By the removal of constituents. (B b) By the addition of new constituents. (Be) By the taking away of some constituents and their replace- ment by others. Thus there are two minerals, Selenite and Anhydrite, each of which occurs in the crystalline form of the other ; the first is a hydrated, the second an anhydrous Sulphate of Lime. When we find Anhydrite under the form of Selenite, one constituent, the water, has been removed, and the case comes under (B a). Conversely, Selenite in the form of Anhydrite comes under (B b). Again we find Fluor-spar (CaF 2 ) in the crystalline form assumed by Calcite (CaCO 3 ). Here C0 3 has been removed and its place taken by F 2 , and we have an instance of (B c). But the change may have been produced by the gradual removal of the Calcite molecule by molecule, and as each molecule was taken away by a molecule of Fluor-spar being put in its place, and then the Pseudo- morph would be put in the class (A b). 82 Geology. The study of Pseudomorphs, specially those of the last class, often throws great light on the steps through which a rock has passed before it reached its present form. Thus, in many rocks which contain Chlorite, this mineral can be shown not to have been one of the original con- stituents of the rock, but to have been formed by the alteration of Hornblende or Augite. And thus we learn that certain chloritic rocks, though they now differ from others of Hornblendic composition, may have been originally identical with the latter and formed in the same way.* Forms which are sometimes called Pseudomorphs are also produced in this way. An embedded crystal is removed in solution, and the mould thus formed is afterwards filled up with a non-crystallized sub- stance, and so a cast of the crystal is formed. Thus crystals of Common Salt are sometimes formed by evaporation on the margin of a salt lake ; the crystals are afterwards dissolved away, and the hollows produced filled up with mud, and a model of the crystals formed in the latter substance. Mud-casts of the crystals of other salts besides Common Salt have also been noticed, of prismatic crystals of Sulphate of Magnesia for instance (Geol. of Canada, Report to 1863, p. 346). Crystallized and Crystalline States. When minerals have external crystalline form, they are said to be crystallized. It has happened frequently however that a mineral has been prevented by certain circumstances from assuming crystalline shape and yet its mole- cules have arranged themselves in crystalline order ; in such a case the mineral is described as crystalline. That a mineral is crystalline can be detected in various ways. Thus we often meet with rounded cavities in rocks filled in with Calcite. If the nodule of Calcite be extracted from the rock, it has no external crystalline form, but has a rounded contour corresponding to the shape of the cavity. But we can easily show that it is crystalline, for when we break it, it shows the rhombohedral cleavage of Calcite (see p. 33). In other cases the mineral can be seen by a lens, or in thin slices under a microscope, to be made up of a mass of small interlacing or radiating crystals. In all cases the effect of the mineral on the light which passes through it is a most important aid in recognising its crystalline condition. When the crystallization is so minute that it can be detected only by microscopic or optical examination, the mineral is said to be cryptocrystalline. It is convenient to have a term to denote those forms of a mineral which are not obviously either crystallized or crystalline, and to these the word massive is applied. It neither asserts or denies the existence of crystalline structure, but merely expresses the fact that such a structure is not apparent. Amorphous States. When minerals have neither external crystalline form nor internal crystalline structure they are said to be amorphous, i.e. shapeless. The following are the more important of the Amorphous states. On Pseudomorphism see Blum, "Die Pseudomorphosen des Mineralreiches " and its supplements ; Geinitz. Neues Jahrb. 1876, p. 440 ; Mineral. Mittheil. 1880, ii. 489 ; Streng, Neues Jahrb. 1881, i. Refer, p. 5. Mineralogy. 83 The Colloidal States. A substance when in the colloidal state has no power to crystallize ; it is often very largely soluble in water, but it is held in solution very feebly, so that what would seem to be very trifling causes are sufficient to precipitate it. Thus in some cases mere contact with other substances causes a colloid in solution to separate out under the form of a jelly which contains more or less water. In many cases mere lapse of time suffices to produce this separation. This hydrated jelly-like state is known as the gelatinous or pectous condition of the colloid. In many cases the jelly is insoluble in water, and in the course of time or by exposure to heat loses some of its water and solidifies into an opalescent or glassy substance : in other cases the jelly is soluble in water and liquefiable by heat. We thus distinguish two modifications of the Colloidal state, the Soluble when it will dissolve in water, and the Gelatinous or Pectous, which is often insoluble in water. In many cases there is a very strong tendency in Colloids to pass from the Soluble into the Pectous modification. Some bodies, such as Gelatine and Albumen, are known only in the colloidal state. There are other bodies which can exist both in a crystalline and colloidal condition. Of these Silica is the one we are most concerned with. Silica exists in a crystalline form in the mineral Quartz. It may be obtained in the colloidal state in the following way. Pounded Quartz (Si0 2 ) is fused with Sodium Carbonate (Na 2 C0 3 ). Carbon Dioxide (C0 2 ) escapes with effervescence and Sodium Silicate (SiNa 2 3 ) is formed. SiCX, + NaC0 3 = Na 2 Si0 3 + C0. 2 . The fused mass is dissolved in boiling water, and Hydrochloric Acid added to the solution. Silicic Acid (Si0 2 + nH 2 0) partly separates out as a gelatinous mass and partly remains dissolved. If the gelatinous silica be separated by filtration and the clear solution placed in a drum with a bottom of parchment paper, which floats in a vessel of water, the hydrochloric acid and sodium chloride pass gradually through the parchment and a clear solution of pure silicic acid remains in the drum. In this way we obtain Silicic Acid in the soluble modification of the Colloid state. The solution cannot be kept however beyond a few days unless it be very considerably diluted. It grows opalescent and after a time becomes pectous rather rapidly, and a jelly separates which is transparent or slightly opalescent and insoluble in water. This is the pectous modification of Colloidal Silicic Acid. In the course of a few days the jelly contracts and parts with some of its water. The method of separation employed in the above experiment is known as Dialysis; it depends on the fact that bodies in a state capable of crystallizing will under the circumstances described pass through certain membranes like parchment, bodies in a colloid state will not. Dialysis is important from a geological point of view, because there are minerals which assume the form of a hardened jelly, and these 84 Geology. may have been formed by a natural process akin to the artificial dialysis just described. Opal, which is a pectous form of Colloidal Silica, is the most important instance. The Glassy State may be conveniently mentioned here, though it is found rather in rocks than in minerals. At first sight no two things can seem to be so totally distinct as a well-crystallized and a glassy body. The regular geometrical form of the one, and the smooth glistening faces along which it breaks, con- trast in the most marked way with the shapeless lumps and rough uneven fracture of the other. The same substance however is often capable of assuming both states, and experiments lead us to the belief that it is the conditions under which they are formed that decide whether bodies shall be glassy or crystalline. Thus, if a body harden from a state of fusion, it has been observed in many cases to take the shape of a glass if it cools quickly, and to crystallize if it cools slowly.* This fact may be thus explained. We have seen that in a crystal- lized body the molecules are arranged according to a definite law ; they may be compared to the soldiers in an army drawn up in order. In a glass the molecules are not arranged according to any law, they are distributed like the people who compose an ordinary crowd. Now we can readily imagine that time is needed for the molecules to group themselves in the orderly arrangement possessed by a crystal, and that therefore crystalline form is usually assumed only when the body solidifies slowly. If the solid state is produced suddenly, the molecules have not time to range themselves in crystalline ranks, they are huddled together in a more or less confused state and a glass results. This explanation is confirmed by what is known as Devitrification. Glassy bodies under certain circumstances become stony, and when this altered glass is examined under the microscope it is found that the change is produced by the growth of minute crystals within the body. The glass is then said to be Denitrified. Sometimes mere lapse of time is sufficient to produce this change ; it may also be brought about by keeping the glass for some time at the temperature at which it begins to soften. In any case the tendency which the molecules have to range themselves according to fixed laws gradually asserts itself and produces a more or less perfect crystallization. In most cases a substance has a higher specific gravity when crystal- lized than when glassy. In fact the difference in this respect is exactly the same as between a badly-packed and a well-packed port- manteau. With the orderly packing of the molecules in a crystal a larger number go into a given space than when they are distributed anyhow as in a glass. Devitrification is always accompanied by an increase in specific gravity and a decrease in volume. Glass is really only a peculiar modification of the Colloid state, but there is a sufficient difference between the Glassy and the Pectous conditions to make it desirable that they should be distinguished by different names. Glassy bodies can be distinguished from crystalline by their effect * See, for instance, the experiments of Sir James Hall, Trans. Koyal Soc. Edinburgh, v. 43. Mineralogy, 85 on polarized light. Two Nicol's Prisms are placed so that the light polarized by the first will not pass through the second and the field of view is dark. If a glass be now interposed between the prisms no change is produced. But if a transparent slice of a mineral crystal- lized in any system except the Monometric be placed between the prisms, the slice becomes bright and often brilliantly coloured. Allotropism and Isomerism. The power which certain sub- stances have of assuming different physical states is called Allotropism when the substance is a chemical element, Isomerism when it is a com- pound. Silica is a good example of Isomerism. It is dimorphic. One crystallized form, known as Quartz, has a specific gravity of 2 '6 5; the other, Tridymite, has a specific gravity of 2 '31. If fused under the oxyhydrogen blowpipe it solidifies into a glass with the specific gravity of 2 '2. In the pectous modification of its colloidal state its specific gravity is also 2 -2. Carbon furnishes perhaps the most marked case of Allotropism. It occurs as the Diamond, intensely hard, transparent, with brilliant lustre, and crystallizing in Monometric forms ; as Graphite, commonly known as Blacklead, soft enough to be scratched with the nail, black, dull, and opaque, and crystallizing in Hexagonal forms ; and also under the amorphous form of Charcoal. Other shapes taken by Minerals. Minerals also assume what are called IMITATIVE SHAPES. They are said to be Nodular, when there are protuberances over the surface. Globular, when the protuberances are prominent and approximately spherical. Mammillanj, when the rounded masses are partly embedded, and only segments of spheres, which bear some resemblance to the female breast, are seen above the surface of the mass. Botryoidal, when the globules bear a rude resemblance to a bunch of grapes adhering by their surfaces. Reniform, when the masses of mineral are kidney-shaped. In all these shapes the interior of the masses is frequently arranged in concentric coats like those of an onion, and in many cases consists of fine fibres radiating from a centre. These fibres are really long needle-shaped crystals. The concentric and radiating structures often occur together. The imitative shapes just described may be all classed together as Concretionary. We shall have more to say about them further on. Other imitative shapes are Acicular or Needle-shaped, consisting of slender crystals. Filiform or Capillary, very slender and long, like a thread or hair. Dendritic, branching like a tree. Coralloid, with interlacing twisted branches and some resemblance to Coral. The "way in which Minerals break. This is a second pro- perty we are often able to make use of in the identification of minerals. The most important item under this head, Crystalline Cleavage, has been already noticed. The following kinds of cleavage may be use- fully distinguished. If the mineral has a very marked cleavage in one direction so that 86 Geology. it splits into plates, it is said to be Tabular, when the plates are thick (Ex. Barytes); Lamellar, when they are of moderate thickness (Ex. Bronzite); Micaceous, when they are very thin (Exs. Mica, Selenite). Again, if non-crystallized minerals are broken, or if crystallized minerals are broken along a plane which is not a cleavage plane, the different kinds of surface are thus named. Smooth or even, when there are no marked depressions. Splintery, when the mineral breaks into pointed fragments. Hackly, when the surface is covered by sharp wire-like points, as when a stout iron wire is broken by bending it backwards and forwards. Earthy, when the mineral breaks like a bit of dried clay. Conchoidal, when the surface shows curved hollows with concentric ridges like those on the inside of some bivalve shells. A bit of Flint or Cannel Coal shows this fracture very well. Tenacity. The following properties, which are described as dif- ferences in Tenacity, may be placed here. Some minerals are Touc/h, that is, though hard, they feel, when they are struck, as if they allowed the hammer to indent rather than break them. In Brittle minerals pieces fly off when we attempt to cut or scratch them. Friable minerals can be crushed to powder even between the fingers. Sectile minerals can be cut with a knife. Malleable minerals can be flattened with the hammer. Flexible, minerals can be bent. Elastic minerals can be bent, and when released fly back again. Colour and Streak. A perfectly pure mineral would probably be always of the same colour. In nature however a large majority of minerals contain small quantities of foreign substances, and these com- municate to them different tints, and thus the same mineral assumes a great variety of colours. In a very large number of cases then Colour is of little or no use in enabling us to recognise minerals. Some minerals however, most of which are metallic ores, are fairly constant in colour. But though the colour of a mineral is variable, what is known as the Streak is usually constant for the same mineral. The Streak may be obtained, if the mineral is soft, by drawing it over a sheet of white paper and noting the colour and character of the mark it leaves. For harder minerals a bit of unglazed porcelain may be used. In the case of very hard minerals they must be crushed and finely powdered in an agate mortar, or powder may be obtained by drawing a file over them. The colour of the powder then gives the streak. When the mineral can be powdered, we get its streak with least risk of error, by noting the colour of the powder ; and in some cases the characteristic colour comes out only when the mineral has been well rubbed down in an agate mortar. Some crystallized minerals appear differently coloured according to the direction in which light passes through them. This is called Dichroism if there are two colours in two different directions ; Tri- chroism if three colours in three directions. For details and an explana- Mineralogy, 87 tion of the cause of these phenomena, reference must be made to works on Mineralogy. What is called the Lustre of minerals may often be usefully attended to. The more important varieties of lustre are, Metallic or the lustre of metals ; Vitreous or Glassy ; Resinous and Waxy ; Pearly ; Silky. Many minerals are liable to tarnish, and in their case the lustre of a fresh fracture or freshly-formed plane of cleavage is lost or weakened by exposure to the air. Dust also impairs the lustre, and sometimes even dusting and washing does not fully restore it. Hardness. This is a character of the greatest value in the identi- fication of minerals. The hardness of minerals may be compared by trying to scratch them with a knife, or better still by drawing a file over them. The pressure required to make an impression, the depth of the cut, the quantity of powder produced, and the noise made enable us to judge of the relative hardness. For more accurate comparison the following " Scale of Hardness " ib used. 1. Talc. 6. Orthoclase. 2. Selenite. 7. Quartz. 3. Calcite. 8. Topaz. 4. Fluor-spar. 9. Sapphire. 5. Apatite. 10. Diamond. Suppose that we have a mineral which will scratch Apatite and which is scratched by Orthoclase, its hardness is between 5 and 6. We can further approximate to a value thus. If it only just makes a scratch on Apatite with considerable pressure the hardness is only a little over 5, say 5*1 or 5'2 ; if it scratches Apatite easily and there is some difficulty in getting Orthoclase to make an impression on it, the hardness is nearly 6, say 5 '8 or 5 '9 ; if it scratches Apatite with the same ease as Orthoclase scratches it, the hardness is half-way between 5 and 6, or 5*5. Considerable care is necessary to avoid mistakes in determining hardness. The mineral must not be decomposed or tending to decom- position. If the mineral is brittle the student must be careful not to mistake the chipping off of fragments for scratching. Different faces of a crystal have not absolutely the same hardness, and even the hardness of the same face is different in different directions ; usually the differences are so small as to require very refined methods for their detection, but in some few cases they are large enough to be recog- nised by simple scratching. In trying to scratch hard minerals with a knife, the first mineral experimented on may blunt the point or edge, and if the knife be used in this state on a second mineral, the trial is evidently unfair, for we are comparing the effects of a sharp and blunt tool. It is better therefore to use a strong blade with neither point nor edge very sharp. Also be careful not to mistake the mark left on hard minerals by a knife for a scratch. If looked at through a lens this mark appears metallic, and is seen to consist of a thin film of steel worn off the knife. 88 Geology. It is useful to recollect that No. 1 in the scale can be scratched very easily, No. 2 without difficulty by an ordinary finger-nail. Nos. 3, 4, and 5 can be scratched with a knife. No. 6 can be scratched with a good knife, but only with great difficulty ; it scratches glass. No. 7 cannot be touched by the knife and it scratches glass with the utmost ease. The Feel or Touch of some minerals is characteristic. Talc, for instance, is Soapy or Unctuous ; Magnesite is Meagre or dry to the touch ; Actinolite is Harsh or unpleasantly rough. Some minerals adhere to the tonc/ue. Specific Gravity. The accurate determination of the Specific Gravity of a mineral is often a useful aid to its recognition. But in the case of minerals of very high specific gravity mere poising them in the hand often gives valuable hints. For instance Heavy Spar and Calcite, when imperfectly crystallized, are often a good deal alike to look at ; but even a small piece of Heavy Spar gives directly it is taken into the hand the impression of being a very heavy substance, and Calcite does not, and the difference is so marked that it is impos- sible to confound the two minerals. Taste and Smell. A few soluble minerals have characteristic taste. In the case of some minerals friction, breathing on them, striking them, heating them, or treating them with acid produces char- acteristic odours. We have already pointed out how valuable the effects of crystallized minerals on light are in determining their crystalline form. An account of the Optical properties of minerals, and those connected with heat, electricity, and magnetism, must be obtained from works on Mineralogy. We need only mention here that some minerals are magnetic and some polar. If the magnetism is pronounced, it may be recognised by the way in which fragments are attracted by a bar magnet ; if feeble, a delicately-balanced compass needle must be used to detect it. SECTION III. DESCRIPTION OF ROCK-FORMING AND SOME OTHER COMMON MINERALS. In the following pages H. stands for Hardness, G. for Specific Gravity, S. for Streak, B.B. for Before the Blowpipe, O.F. Oxidizing Flame, R.F. Reducing Flame. A. MINERALS COMPOSED OF SILICA. Quartz. SiO 2 . Crystallized Silica. G. 2-5 2-8, of the purest specimens 2 -65. Hexagonal. Commonest form an hexagonal prism terminated by hexagonal pyramids, or a double hexagonal pyramid (figs. 5 and 6). No cleavage. Conchoidal fracture. H. 7, scratches glass with perfect ease. When pure, colourless and transparent ; but stained various colours by impurities. B.B. infusible. Infusible in Microcosmic Salt and Borax. Fused with the proper amount of Carbonate of Soda dissolves with effer- Rock -forming Minerals. 89 vescence and forms a transparent glass. To obtain this reaction mix about equal quantities of powdered mineral and soda, make into a paste with water, dry, and fuse on platinum wire. The bead will probably turn milky as it cools. Add a little more powdered mineral and fuse. The bead is a little clearer. By gradually adding the mineral in small quantities a transparent glass will at last be obtained. Unattacked by acids except Hydrofluoric Acid. Soluble, but only very slowly indeed, when powdered, in a boiling solution of Caustic Alkali ; dissolves more rapidly under pressure. Quartz is one of the commonest and one of the most easily recognised of minerals. When it has no distinct crystalline shape, its hardness, its conchoidal fracture, the absence of cleavage, and its infusibility distinguish it. If crystallized it is still more easily known in most cases, for in the vast majority of instances its crystals are either combinations of the hexagonal prism and pyramid or double hexagonal pyramids. Even when faces belonging to other forms are present, these are usually the preponderating shapes. The non-solubility in Microcosmic Salt, and the formation of a clear glass with Carbonate of Soda are useful confirmatory tests. In microscopic slices Quartz is clear and limpid with unpolarized light : with polarized light, if the plane of the section is inclined at a large angle to the axis and if the section is not too thin, it shows most brilliant changes of colour as the analyzer is rotated ; in very thin sections the tints are pale yellow and neutral tint. When the mineral occurs in crystals, sections nearly at right angles to the axis have an hexagonal outline. The crystals are cracked in many rocks. Another very marked characteristic of Quartz is that it scarcely ever shows any signs of alteration. Crystals of Quartz are most commonly met with on the walls of cavities and fissures in rocks. As a rock-forming mineral it occurs in some crystalline rocks in crystals, but usually fills up the spaces between other crystallized minerals and has no definite shape. Sand consists of grains of Quartz more or less rounded, and this is the form under which Quartz occurs in non-crystalline rocks. Quartz crystals often contain minute hollows or cavities* containing water, saline solutions, or liquid carbonic acid. Sometimes these cavities are so numerous as to make the Quartz opaque and milky ; it is then known as Milk-white Quartz. Under this form it is very common in veins. Of its distinguishing characters the hardness of Quartz is the one which the beginner would perhaps notice first of all, and it will be a useful illustration of the way in which minerals are recognised, if we note here some of the commoner minerals which being about equally hard might by the very inexperienced observer be mistaken for Quartz, and show how they may be distinguished from it. A yellow Quartz is so like Topaz that it is called False Topaz. True Topaz has a hardness of 8, and will therefore scratch Quartz : this alone will distinguish it. Topaz is Trimetric, but the angle of its Unit Prism is 124 17'; its prisms therefore assume pseudo-hexagonal * For more on this subject see p. 310. 90 Geology. shapes.* But it lias perfect basal cleavage, and its prisms are often terminated by a basal plane. The cleavage and hardness alone distin- guish it from Quartz. Emerald or Beryl is hexagonal, and it takes the form of hexagonal prisms : so far like Quartz. But its prisms generally show the basal plane, and it has imperfect basal cleavage. Also it is fusible on edges of thin splinters. Usually it will scratch Quartz : if it fails to do this, its fusibility distinguishes it. Tourmaline. H. 7 7 '5, so that this will not always distinguish it from Quartz. Hexagonal, but the habit of its crystals is totally different from those of Quartz, especially in their terminations. Most varieties are fusible, and all show the presence of Boric Acid by Turner's test.t Staurolite. H. 7 7 '5. Trimetric, but pseudo-hexagonal, and in this respect bears some resemblance to Quartz. But it dissolves in Borax and Microcosmic Salt B.B., and gives a distinct Iron reaction. Epidote, Axinite, and Spodumene have a hardness ranging from 6 to 7 : the crystals of Epidote and Axinite are totally unlike those of Quartz, and Spodumene has a very marked cleavage. More important by way of distinction, all three are fusible. Zircon has a hardness of 7 '5, and in some cases it might not be easy to make it scratch Quartz. But it is Dimetric, its prisms are four- or eight-faced, and the angles between the faces are 90 and 135. In crystallized examples these characters distinguish it from Quartz. Tridymite. Si(X. A second form of crystallized silica. G. 2 '3. The crystalline system is still somewhat uncertain. Von Rath, who discovered it, considered it Hexagonal ; + reasons have since been given for thinking that it is Triclinic, the angle of the unit prism being nearly 120, and the vertical axis being nearly perpendicular to the base. It occurs in very minute six-sided plates, and each of these is considered to be a twin- aggregate of three individuals. H. 7. B.B. infusible : with Carbonate of Soda fuses with effervescence to a clear glass. The plates of Tridymite are seldom more than '02 mm. across, and it then requires a power of at least 120 to show its characteristic structure distinctly. This is then so very marked that it is scarcely pos- . sible to mistake the mineral. It forms collections of thin pellucid scales dovetailing into one another and overlapping one another like tiles on a roof, as in fig. 58. The scales sometimes show hexagonal outlines, but often these grow more and more indistinct, till they pass into rounded contours. Tridymite is at present known only in a comparatively few cases of acid crystalline rocks of late geological age. * Seep. 125. t Landauer's Blowpipe Analysis, p. 84. J Pogg. Ann. 1868, 135; 1874, 152. Schuster, Min. mid Petrog. Mittheil. i. 74 ; Von Lasaulx, Zeit. fiir Krystall. ii. 253. Rock-forming Minerals. 9 1 Tridymite has the same specific gravity as Opal, and is constantly associated with it ; it is possible that it may be a de vitrified form of that mineral. The question arises whether the absence of Tridymite from the older rocks may not be due to the fact that in them a further stage of alteration has transformed it into Quartz. Opal. SiO 2 + Water, the water varying from 1 to 21 per cent. Pectous Silica. G. 1-92-3 ; of the purest examples 2 '3. H. 5 -5 6 -5. B.B. infusible. Dissolves, when powdered, in hot alkaline solutions : it may require several hours' boiling. In the closed tube gives water. Opal in its purest form can only be looked upon as a rare accessory constituent of rocks. Impure varieties, Half -Opal, containing Alumina, Ferric Oxide, Lime, Magnesia, and Alkalies, occur rather oftener, filling in cracks and cavities, or in nodules or thin bands and layers. It has been produced by the decomposition of silicates, and has been separated by a natural process analogous to that by which gelatinous silica is artificially obtained. Cpal is under ordinary circumstances isotropic, and sections of it are dark throughout an entire revolution between crossed Nicols. There is however a variety of Opal, which occurs in small globular masses, consisting of concentric coats, which shows a black cross, like that exhibited by uniaxal crystals when viewed by convergent polarized light between crossed Nicols. This peculiarity seems due to the fact that the globules have undergone contraction ; the result of contraction on a sphere is that the molecules are less thickly packed along any radius than perpendicular to that radius. Hence if we cut a plate from such a sphere, the plate will have to a certain degree the mole- cular structure of a uniaxal crystal, the radius of the sphere perpendi- cular to the plate corresponding to the axis of the crystal. But the parallel is not complete, for in a crystal any line parallel to the crystal- lographic axis is an Optic Axis, but in the case of the plate it is only the one particular radius of the sphere that has the properties of an Optic Axis. It sometimes happens that the exterior coats have been unable to bear the strain put on them, and have relieved themselves by cracking. They then again become homogeneous and isotropic, and the globule is surrounded by a dark ring. This form of Opal is sometimes distinguished as Hyalite. The concentric structure is probably due to curved cracks produced by contraction, and is similar to what is known as Perlitic structure (see p. 305). Chalcedony. Si0 2 . G. and H. about the same as Quartz. Crypto- crystalline, perhaps sometimes a mixture of pectous and cryptocrystal- line silica. Mammillary, botryoidal, or stalactitic, never crystallized. Agates are nodules consisting of concentric bands of various colours, some of which are chalcedony and some quartz. Jasper is an impure chalcedony of various colours, red from ferric oxide being the commonest. Flint (Feuerstein). Amorphous or cryptocrystalline silica, or a mix- ture of both, with various impurities. Brittle. Breaks with a very marked conchoidal fracture and sharp cutting edge. 92 Geology. C/iert (Hornstein) differs but little from Flint, but it is tough instead of brittle, and it breaks with a splintery in place of a conchoidal fracture. Both Flint and Chert occur chiefly in the form of nodules in lime- stone rocks. Chalcedony occurs in some rocks in small globular masses, with concentric and strongly radiated structure. Under the microscope with polarized light the fibres often show very brilliant colours, and between crossed Nicols there is sometimes a black cross. B. MINERALS COMPOSED OF SILICATES. The following are useful tests for recognising Silicates. Fused in Microcosmic Salt the bases are dissolved and the silica floats about on the bead, forming what is called a "siliceous skeleton." The skeleton is best observed while the bead is in the flame ; when cold, the skeleton is apt to settle down into a corner out of sight. Many Silicates are decomposed by boiling them in strong Hydro- chloric Acid. The silica sometimes separates as an impalpable powder ; it may then be recognised by its being insoluble B.B. in Microcosmic Salt, by its forming a clear glass with Carbonate of Soda, and by its dissolving readily in a strong boiling solution of Caustic Soda or Potash. Some- times the silica dissolves in the acid : if the solution be then evaporated down, the silica separates out as a jelly and is said to gelatinize. Gela- tinization sometimes takes place without evaporation ; usually the evaporation must be carried almost to dryness. This gelatinous silica is now insoluble, and if it be separated by filtration, the filtrate may be tested for the bases. To ensure the complete separation of the silica from the bases the acid containing the pulverulent or gelatinous silica should be evaporated to dryness and the residue heated with stirring at a. temperature above 100 C. till no more acid fumes escape. The residue is then moistened with strong Hydrochloric Acid, water is added, and the whole well boiled. The silica separates and may be removed by filtration, and the filtrate tested for the bases. In dealing with Silicates the student must not suppose because he finds in works on Mineralogy the curt statements, " Soluble in Acids," " Gelatinize with Acids," that these tests are always easy of application. We are seldom told how long it takes to decompose the Silicate. Some decomposable Silicates dissolve largely and readily at moderate temperatures even in dilute Hydrochloric Acid ; some dissolve slowly in dilute Hydrochloric Acid at temperatures below the boiling-point, and several hours' digestion is required to decompose even a small quantity ; others require long boiling in strong acid to produce even partial de- composition ; and some scarcely any amount of boiling will decompose completely. To ascertain whether a silicate is decomposable, powder it very finely and digest the powder in strong Hydrochloric Acid at a temperature near the boiling-point for some hours. Separate the insoluble part by filtration, and add ammonia to the filtrate till it ceases to redden blue Rock-forming Minerals. 93 litmus-paper. If any decomposition has been produced there will be, if the Silicate contains alumina, a greater or less precipitate of alumina : this may be separated by filtration and the solution further tested with Ammonium Carbonate, Sodium Phosphate, and other reagents. If none of the reagents produce any precipitate the Silicate is not attacked by Hydrochloric Acid. Aqua regia may then be tried, and after that a mixture of three parts of concentrated Sulphuric Acid with one part of water. If after treatment with these the filtrate gives no precipitate, the Silicate may be pronounced to be unaffected by acids. The pulverulent silica separated by the action of acids can usually be easily distinguished from undecomposed mineral, however finely the Silicate may have been powdered. The most finely-powdered Silicate is gritty to the feel if it is rubbed with a glass rod against the bottom of the beaker, or when it is dragged by the blade of a knife over a plate of glass, and in the case of most Silicates will scratch the glass. Pre- cipitated silica is far more finely grained than the finest artificial powder, has no gritty feel, and cannot be made to scratch the glass. But in case of any uncertainty as to whether the residue is precipitated silica or undecomposed mineral, or a mixture of the two, the student may boil it in a strong solution of Caustic Soda or Potash. Freshly-pre- cipitated silica, which has not been ignited, dissolves readily : most Silicates are insoluble, and those that do dissolve, dissolve much more slowly than newly-precipitated silica. A sufficient amount of decomposition to be of great value in identi- fying Silicates may often be produced thus. Place some of the coarsely- powdered mineral in a small glass bottle with a wide mouth, and pour over it enough strong Hydrochloric or Nitric Acid to cover the powder, and let the whole stand for at least twenty-four hours. The acid will frequently show the characteristic colours of Sodium, Potassium, Cal- cium, and Lithium in the Bunsen flame, or will give precipitates when tested with the usual reagents. We will speak of this treatment as "slow cold digestion." The student must also guard against the idea that gelatinization is a distinctive test of a Silicate. The same mineral, even the same speci- men, will often yield at one time pulverulent and at another time gelatinous silica, according to the way in which it is treated. As a rule rapid decomposition, such as is produced by boiling in strong- acids, seems to tend to a separation of silica in a pulverulent state ; slow disintegration by digestion in dilute acid, at a temperature below the boiling-point, seems to favour the production of gelatinous silica. Some insoluble Silicates dissolve and gelatinize after having been exposed to a high temperature; but a white heat of several hours' duration may be required. Silicates insoluble in acids are fused with four parts of Carbonate of Soda, or a mixture of Carbonates of Soda and Potash. The fused mass is dissolved in dilute Hydrochloric Acid and the solution evaporated to complete dryness. The residue is moistened with strong Hydro- chloric Acid, water added, and the mixture boiled. Silica separates, and when it is removed by filtration, the filtrate is tested for the bases. For most Silicates this method is far preferable to treatment with 94 Geology. acids, which is usually tedious- and uncertain. But for Silicates that decompose easily, Leucite for instance, treatment with acids is handy. In the case of some of the Silicates we shall have to study, I have found that when other methods do not suffice for their determination, wet tests may be applied with success, even by one who has no great experience of chemical methods, and in such cases these tests will be mentioned. But unless he has the run of a well-equipped chemical laboratory and has considerable skill in chemical manipulation, the geologist will do well to leave even the qualitative analysis of the majority of Silicates alone, and to hand over to the chemist any Silicate which he cannot determine by physical characters and blowpipe tests. (B 1) THE FELSPAH GROUP. Among the minerals composed of Silicates by far the most important and widely diffused are the Felspars. The Felspars are silicates of Alumina and Potash, Soda, or Lime,* or of Alumina and two or more of these three bases. Their com- position is further complicated by the fact that these three bases are apt to replace one another ; thus Orthoclase, which is normally a silicate of Alumina and Potash, is scarcely ever free from Soda which replaces part of the Potash. Their specific gravity lies between 2 '5 and 2 '8 Their hardness, when they are not decomposed, is about 6, so that they can be scratched with a good knife, but only with con- siderable difficulty ; and they will scratch glass, but not so easily as Quartz. Of the species we shall be concerned with all are Triclinic except Orthoclase which is Monoclinic. The angles in the different species are however so near to one another that they may be all looked upon as isomorphous and capable of replacing one another. All the Felspars have two cleavages. One is parallel to the base of the Unit Prism in all of them. In Orthoclase the other is parallel to the clinopinacoid. In the Triclinic Felspars the second cleavage is parallel to the brachypinacoid. In Orthoclase then the two cleavages are at right angles to one another, whence its name. In the Triclinic Felspars the two cleavage planes are not at right angles, and hence they are called Plagioclastic. The difference from a right angle is however too small to be detected by the eye. Twinning is common in all the Felspars. In Orthoclase the com- monest case is that already described (p. 76) as the Carlsbad type. Two individuals only are usually present in each crystal. The twofold composition can often be recognised by the eye by a difference of lustre in the two portions ; if a thin section be cut along a plane making a large angle with the composition plane, and examined in polarized light, the simple twinning becomes very apparent, for the crystal is then seen to consist of two portions of different colours. Twinning on the Baveno type is less common. The geologist has usually to deal with twinned crystals in thin microscopic sections, and in this case the two types of twinning may be thus distinguished. In a Carlsbad Twin the composition plane cuts the section along a line * To these may be added Baryta, which occurs in Hyalophane ; but this is a comparatively rare mineral. Rock-forming Minerals. 95 which is parallel to two of the sides of the section of the crystal : in a Baveno Twin the composition plane is represented by a line which crosses the section of the crystal in a diagonal manner. In the Triclinic Felspars twinning is many times repeated in the same crystal, so that a single crystal contains a large number of thin parallel lamellae. Twinning many times repeated in this fashion is spoken of as Poll/synthetic. In the commonest type, known as the Albite type of twinning, twinning and composition planes coincide and are parallel to the brachypinacoid. In this way the base becomes striped across by a number of very fine parallel grooves, so shallow and narrow that they look like striations. The diagram in figs. 59-61 will explain how this is brought about. Fig. 59 is a section perpen- Fig. 61. dicular to the twinning plane of two thin lamellae A BCD, BCFE, before revolution has taken place ; if now the left-hand lamina remain fixed, and the right-hand lamina be rotated through two right angles round an axis GH perpendicular to BC, so that it comes into the position CBEF in fig. 60, we shall obtain a crystal whose section is shown in fig. 60, with a groove running across the base at the upper end, and a ridge at the lower end. A repetition of this produces a shape such as is shown in fig. 61, and when the laminae are very thin, this gives rise to the striated appearance described. In a thin section cut across the twinning plane, each alternate lamina when examined in polarized light is differently coloured, and the section seems to be composed of a number of narrow parallel bands variously tinted. Another less com- mon type of twinning is known as the Perikline type. The base is the composition plane, the twinning axis is parallel to the base, and perpendicular to the brachydiagonal. This causes the brachypinacoid to be striated parallel to its intersection with the base. All the Felspars are fusible B.B., but some with more difficulty than others. 9 6 Geology. Some of the chief characters of the principal Felspars are tabulated in the following list. The last column shows the ratio of the oxygen of the silica, of the sesquioxide alumina, and of the protoxide bases. In the case of the two last the ratio is the same for all, but there is a gradual decrease as we go down the column in the proportion of oxygen belonging to the silica. On this ground the Felspars may be divided into the Highly Silicated or Acid, and the Poorly Silicated or Basic. Orthoclase, Microcline, and Albite form the first class ; Ande- site, Labradorite, and Anorthite the second ; Oligoclase occupies an intermediate place. There is considerable difference of opinion among mineralogists as to how many of the Felspars are really distinct species. Some only allow of three independent species, Orthoclase, Albite, and Anorthite, and look upon all the rest as mixtures in different proportions of these three through contemporaneous crystallization. It is in favour of this view that a felspar Perthite is known which certainly consists of alternate laminae of Orthoclase and Albite.* On this head see Sterry Hunt, Chemical and Geological Essays, 2nd ed. (Salem, 1878), p. 443 ; Tschermak, Sitzungsberichte d. Kais. Akad. d. Wissen. Bd. I. Abth. i. 571 ; Rammelsberg, Mineralchemie, ii. 561 ; Dana's System of Mineralogy, 5th ed. p. 366 ; Both, Allgemeine und Chemische Geologic, p. 13 ; Fouque and Levy, Mineralogie Micrographique, 204. Angle between the two planes of Cleavage. Normal Chemical Composition. 1 Oxygen Ratio. Moiioclinic \ Or t ho ?kse and / bamdme /Microcline . Albite . . . J90 90 16' 93' 36' 6Si0 2 .AU) 3 .K 2 > 12 : 3 : 1 J Olisroclase . Triclinic ( * \ Andesite . . 93 50' 87 to 88 20Si0 2 .4Al 2 3 . 3Na 2 O.CaO 8SiO 2 . 2 A1 2 3 . Na 2 O. CaO 10 : 3 : 1 8:3:1 1 Labradorite 93 20' 12Si0 2 .4Al 2 O 3 .Na 2 0. 3CaO 6:3:1 ^ Anorthite . . 94 10' 2Si0 2 .Al 2 3 .CaO 4:3:1 Orthoclase, Potash Felspar. Orthoclase frequently contains some soda. This is in some cases owing to an admixture of Albite. In other cases, according to Des Cloizeaux, no Albite can be detected under the microscope in Orthoclase containing soda up to 7 per cent.t One of the commonest crystallized shapes is shown in fig. 50. G. 2-4 2-6. H. 66-5. The edges and points of splinters can be fused B.B.. but not easily. The edges of a splinter 1 mm. square in cross section can be rounded. The beginner in blowpipe-work must be cautious how he pronounces a * Des Cloizeaux, Ann. Chem. et Phys. [5] ix. (1876), 465. t Ibid. [5] ix. (1876), 465-475. Rock-forming Minerals. 97 mineral infusible. He will very likely fail to fuse Orthoclase at the first trial. The flame must be of a pure blue inside, and very steady. The ability to round the edges of a splinter of Orthoclase is a good test of the student's power to produce such a flame. Orthoclase con- taining much soda is more easily fusible. Not acted on by acids. Sanidine or Glassy Felspar is a variety of Orthoclase with a glassy brightness and more or less transparent. The crystals tend to be tabular in shape, their broad faces being clinopinacoids : they are often traversed by numerous cracks. Clear glassy forms of Orthoclase, free from cracks, are sometimes distinguished as Adularia. Microcline is a felspar with the composition of Orthoclase, but Triclinic.* In a very large number of cases Microcline has been found to contain included bands and portions of Orthoclase and Albite. Albite, Soda Felspar. G. 2-59 2*65. H. 67. Fuses more easily than Orthoclase ; for instance, in an irregular fragment, base 5x8 mm., height 5 mm., tapering to an edge, the edges were rounded easily : the ends of a fragment 2 mm. long, 1 mm. square in cross section, could be fused into a bead. Colours the flame deep yellow. Unaffected by acids. Oligodase.G. 2-56 2-72. H. 67. Fuses in moderate-sized fragments. A piece 2 mm. square in cross section runs down easily. Scarcely attacked by acids. Andes it e. Found more commonly in a granular form than well crystallized. G. 2 -6 12-74. H. 56. Fuses in thin splinters. Imperfectly soluble in acids. Labrador ite. Good crystals rare. G. 2 -67 2-76. H. 6. Frequently shows a beautiful play of colour on the brachypinacoid cleavage planes. But this is not always present, nor is it confined to Labradorite. Labradorite is more fusible than Orthoclase. In chips 3 mm. thick and 1 mm. broad, the difference is scarcely perceptible. But if we take splinters 1 mm. square in cross section, we find that in the case of Orthoclase we can do no more than round the edges ; with Labradorite we can fuse the end into a bead. The comparison was made between clear transparent Adularia and the chatoyant Labra- dorite of Labrador. Decomposed only imperfectly and very slowly by strong Hydro- chloric Acid ; after four hours' boiling there was still a large amount of undecomposed mineral; the liquid, when the Alumina had been precipitated by Ammonia, gave a distinct precipitate with Oxalate of Ammonia. Slow cold digestion has very little effect on Labradorite. After a week's standing no Calcium colour was obtained from the acid in the Bunsen flame. Some amount of decomposition is however produced, for both Ammonia and Ammonium Oxalate gave a precipitate. Anorthite, Indianite. G. 2-662-78. H. 67. Brittle. About the same fusibility as Orthoclase. Decidedly more easily decomposed than Labradorite. By boiling glassy Anorthite from Vesuvius in strong Hydrochloric Acid, the silica was separated in * Ann. Chem. et Phys. 433. Q 98 Geology. a pulverulent state. The solution gives a distinct Calcium colour in the Bunsen flame, and large precipitates with Ammonia and Ammonium Oxalate. Slow digestion in cold acid produces the same results. The Felspars may be distinguished from Quartz (1) By their cleav- age, even small grains show when broken bright cleavage faces, while Quartz breaks, like glass, with an uneven or conchoidal fracture ; (2) by their fusibility, if the student is sure of his power to produce a steady hot name ; (3) generally by their inferior hardness. Transparent, unaltered Felspar, in which the cleavage is not distinctly shown, cannot always be distinguished with certainty from Quartz on microscopic slices. The outline of the crystals then is a point to be attended to : usually too in some part or other of the slide the Felspar will show more or less of decomposition. The colours given by Felspar with polarized light also are not by any means so brilliant as those of Quartz under favourable circumstances. A mineral that polarizes in very brilliant colours can hardly be a Felspar. But to tell one Felspar from another is by no means an easy matter. The more massive cleavable varieties of Orthoclase have a character- istic look, and though mere appearance is a very dangerous test to trust to in determining a mineral, a fairly experienced eye can often be pretty sure of such forms of Orthoclase by their look alone. Again a Felspar is known to belong to the Tri clinic group if it shows the characteristic striation either to the naked eye or by the aid of a pocket-lens. Whenever this striation is visible, we may be sure that the Felspar is not Orthoclase;* the absence of striation, however, does not prove that the Felspar is not Triclinic. To detect the striation the crystal should be held so that a good light falls on the basal plane, and turned backwards and forwards till the light falls at the right angle to show the marking distinctly. When the sections are examined by polarized light under the microscope, the striation is very strongly brought out, unless the section happen to be cut parallel or nearly parallel to the brachypinacoid. If the section be very thin, the laminae are of various shades of neutral tint ; in thicker sections they are variously coloured, and the colours change as the analyzer is rotated. We can say then often enough that a Felspar is certainly Triclinic, but in many cases nothing short of quantitative analyses, or at least a determination of the alkalies, will tell us which of the Triclinic Felspars it is. But sometimes the following considerations will help us to an approximate determination. Labradorite and Anorthite can both be sufficiently decomposed by acids to give a precipitate with Ammonium Oxalate after the solution has been neutralized by Ammonia. They may be distinguished from one another by the fact that Anorthite is more readily decomposed than Labradorite. A very moderate amount of boiling in strong acid completely decomposes Anorthite ; with Labradorite, even after many hours' boiling, there is some undecomposed mineral remaining. Slow cold digestion produces a far more marked decomposition in Anorthite than in Labradorite. Labradorite also fuses more easily than Anorthite. * Notes on some Peculiarities in the Microscopic Structure of Felspar. F. Rutley. Quart. Journ. GeoL Soc. xxxi. 479. Rock-forming Minerals. 99 Of the remaining three Triclinic Felspars noticed above, Andesite may be distinguished from Albite and Oligoclase by its more difficult fusibility. Methods of distinguishing the Felspars have been devised by Professor Szabo", and Professor Boriclrf, which in skilled hands give good results ; * and Des Cloizeaux has adopted a plan which depends on their optical properties.! For the geologist who has usually to deal with the Triclinic Felspars in thin rock sections, .the following method will often give useful hints as to the species. When the two Nicols Prisms of a polarizing microscope are placed so that the corresponding diagonals of the rhombic faces are perpendicular to one another, no light can pass through the second Nicol and the field of view is rendered dark. Now if a slice of a mineral crystallizing in any system except the Monometric be placed between the Nicols, the light becomes modified in such a way that for most positions of the section it can pass through the second Mcol. The field of view becomes light and is usually coloured. This is very incorrectly spoken of sometimes as Depolarization. But there are in most sections two lines at right angles to one another which have this property ; if either of them is parallel to a diagonal of the faces of the Nicols, the field of view remains dark. These are called the Directions of Extinc- tion, and when we say that the mineral "extinguishes along these lines," it is only a short way of expressing the fact that the section is dark between crossed Nicols where one of these lines coincides with a diagonal of the face of the Nicols. In any given section the position of these lines is the same approximately for the same mineral, and usually different for different minerals. Suppose now we have a section of a Felspar cut in a known direc- tion, and that we know the angles which the directions of extinc- tion in that section make with a certain line, the trace of a cleavage plane for instance, in the different Felspars. We first make this trace coincide with the diagonal of one of the Nicols ; we then rotate the section till the field of view becomes dark, and measure the angle through which we have turned the slice. Our Felspar belongs to the species which extinguishes at this angle. Des Cloizeaux has deter- mined the directions of extinction for the Triclinic Felspars in sections parallel to the base and brachypinacoid,J and MM. Levy and Fouque have investigated formulae for determining these directions in any section of any crystal. * Ueber eine neue Methode die Feldspathe auch in Gesteinen zu bestimmen. Dr. Joseph Szabo: Buda Pest, 1876. Elemente einer neuen chemisch- mikroskopisclieii Mineral- und Gesteins- analyse. Dr. E. Boricky, Prag. 1877. t Comptes Rendus, Feb. 8, 1875, April 17, 1876; Ann. Chem. et Phys. [51 iv. (1875). J Ann. Chem. et Phys. [5] ix. (1876), 475. There is a summary of the results in Dana's Text-Book of Mineralogy, p. 298. Mineralogie Micrographique Roches Eruptives Francises, p. 64; Annales des Mines [7] Memoires, xii. (1877), 392. See also Max Schuster, Sitz. Wien. Akad. Ixxx. 1 (noticed in Neues Jahrb. 1880, ii. Ref. p. 8); Min. und Petrog. Mittheil. iii. 117 (noticed in Neues Jahrb. 1880, ii. Ref. p. 343, and in the Mineralog. Mag. iv. 194); and Bauerman, Systematic Mineralogy, i. 258. 100 Geology. In order to apply this method to the determination of Felspars on rock sections, it is necessary to know the direction in which the section is cut, and this we have seldom any means of ascertaining exactly. It is however of considerable assistance to bear in mind some such points as the following. In Orthoclase sections parallel to the base are often elongated parallel to the clinodiagonal ; they show one set of cleavages parallel to this direction, and extinguish parallel to the cleavage. Sections parallel to the clinodiagonal show one set of cleavages, and the direction of extinction makes an angle which varies from to 5 with the cleavage. Sections parallel to the orthodiagonal, with the exception of the base, show two sets of cleavages and extinguish along the cleavages. Sections parallel to the faces of the Unit Prism show two sets of cleavages making an angle of 63 53' with one another ; the directions of extinction vary very much in different individuals. The following table gives some of the more important constants relative to the directions of extinction in the Triclinic Felspars. Angle between the Directions of Extinction of two adjacent lamellaj twinned on the Albite Type. Section parallel to Brachypina- coid. Angle Section parallel to Base. Section paral- lel to Braehy- diagonal. Section perpendi- cular to Brachy- pinacoid. Direction of Extinction and Brachydiagonal Microcline 30 to 32 to 31 to 36 5 to 7 Albite . 7 to 8 to 12 to 31 30' 18 to 20 Oligoclase 2 to 3 to 3 to 37 1 to 3 Labradorite . 10 to 14 30' to 18 to 62 30' j 26 to 28 Anorthite 57 to 74 to 40 1 \ to more ) 010 \ than 74 42' j i The table must be read thus. In a basal section of Microcline the angle between the directions of extinction of two adjacent lamellae lies between 30 and 32. In the various sections that may be cut parallel to the brachydiagonal this angle is 31 on an average in the basal plane, and diminishes as the inclination of the section to the basal plane increases till it comes down to 0. In a section parallel to the brachypinacoid the direction of extinction makes an angle of from 5 to 7 with the brachydiagonal. If the rock we are dealing with contains crystals large enough to allow of their being detached, we may sometimes obtain, by mere cleavage, slices parallel to the base and brachypinacoid thin enough to be transparent, and then the first and fourth columns will enable us to determine the species. When we have only rock-slices, the following considerations allow us to employ the second and third columns. In a large number of rocks the felspars present themselves in long slender crystals elongated parallel to the brachydiagonal. If there are a large number of such crystals visible, and if in each of them all its twin lamella? extinguish nearly at the same time, the felspar is probably Oligoclase : if the angle between the directions of extinction of two adjacent lamellae does not exceed. 1 2, the felspar is probably Albite, and so on. Rock-forming Minerals. 101 Where we have to deal with broad irregularly-bounded sections of crystals, we may endeavour to pick out those perpendicular to the brachypinacoid. These are recognised by the fact that in them the directions of extinction in two adjacent lamellae are equally inclined in opposite directions to the trace of the composition plane. Having selected such sections we determine the angle between the directions of extinction of two adjacent lamellae for as many crystals as possible. If this angle does not exceed 37, the felspar is probably either Albite, Oligoclase, or Microcline ; it is not possible to measure the angle with .sufficient exactness to say which. If this angle exceeds 37 in all the crystals but falls short of 62 30', the felspar is probably Labradorite ; if it exceeds 62 30', it must be Anorthite. It is not uncommon to find the Albite and Pericline types of twin- ning combined in the same crystal in the Triclinic Felspars, and this gives rise in some sections to two series of striations running nearly at right angles to one another. In some cases the lamellae of one series abut against those of the other but do not cross them : in others tlie lamellae of one series run over those of the other and produce a very peculiar kind of cross-hatched pattern. This is very characteristic of Microcline, but it is by no means certain that it is confined to that species. (B 2) MINERALS ALLIED TO THE FELSPARS IN COMPOSITION BUT DIFFERING FROM THEM IN CRYSTALLINE FORM. Nepheline. A silicate of Alumina, Soda, and Potash. No single formula has been devised which will include all the variations which have been observed in the composition of this mineral* S. 2 -5 2-65. Hexagonal. Commonest forms six-faced or twelve-faced prisms ter- minated by basal planes ; occasionally the terminations show faces of various hexagonal pyramids as well. Cleavage too imperfect to serve as a means of recognition. H. 5 % 5 6. Frequently glassy and transparent, but also of various colours. B.B. easily fusible. A cubical fragment of glassy Nepheline from Mt. Somma about 2 mm. in the edges, fused on the points and edges to a clear glass ; thinner splinters fused completely. When boiled in strong Hydrochloric Acid is easily decomposed and the silica separates in a pulverulent state. If gradually heated with dilute acid till the liquid just boils, the silica gelatinizes. The solution gives a strong Sodium colour in' the Bunsen flame, but the Potash is often present to so small an extent that it does not show even through Cobalt glass, t * See Rammelsberg, Mineralchemie, ii. 477. t In using Cobalt glass to detect Potash in the presence of Soda it is essential that the glass be thick enough to cut off the yellow Sodium rays completely. The glasses often supplied are quite useless. To test the glass, take up a little Carbonate of Soda on a platinum wire and hold it on one side of a Bun sen flame so as to colour only half the flame. Look at the flame through the glass ; if the coloured part of the flame is brighter than the uncoloured side, the glass is not thick enough. IO2 Geology. When in crystals Nepheline can scarcely be mistaken. Of common minerals crystallizing in hexagonal forms it is not so hard as Quartz, and its crystals usually show the basal plane ; it is far harder than Calcite which occurs occasionally in hexagonal prisms terminated by basal planes, and it is harder than Apatite which occurs frequently in this shape. Even the coloured varieties of Nepheline could scarcely be confounded with Tourmaline or Beryl, and its inferior hardness will distinguish it from either, should external appearance fail to distinguish it from them. When massive there may be a little difficulty in being sure of Nephe- line, without having recourse to wet tests. These show it to be a silicate, and it may be distinguished from the only silicates with which it is likely to be confounded thus. Among the Felspars the only ones that are as easily decomposed by acids as Nepheline are Labradorite and Anorthite ; the solutions of both react for Lime, which is not present in Nepheline. Nepheline is also more fusible than any felspar except perhaps Oligoclase. From Leucite Nepheline is distinguished by its fusibility. Elceolite is a variety of Nepheline occurring in large coarse crystals or massive in granite or metamorphic rocks. It has an oily or greasy lustre, whence its name. Since the crystals of Nepheline are usually bounded by the basal and prismatic faces of the unit prism, they show generally in thin microscopic sections hexagonal outlines if they are cut nearly perpen- dicular, and rectangular outlines if they are cut parallel to the axis ; the rectangular sections are generally not far from square because the crystals are most commonly short prisms. Triangular and penta- gonal sections occur when the slice is inclined to the axis. Sections accurately perpendicular to the axis are dark between crossed Nicols during an entire revolution : sections inclined to the axis are dark in four positions only, and between these are coloured, the brightness of the tint increasing as the section becomes more nearly parallel to the axis. But the colours are never so vivid as in Quartz. Nepheline may be distinguished from Apatite by remember- ing that Apatite tends to form long needle-shaped crystals, and that the prisms of Nepheline tend to be short and stout. Long needle- shaped crystals or " microliths," and fine dusty matter are often found in Nepheline, and they are frequently arranged parallel to the faces of the crystals, so that in a section perpendicular to the axis the needles make angles of 120 with one another and in a section parallel to the axis cross one another at right angles. Nepheline when it is not decomposed is clear and translucent, but by decomposition a fibrous pale-yellowish substance is often developed at first on the exterior part of the crystals and gradually extends itself inwards. Leucitej Amphigene. 4SiO 2 ,Al 2 O 3 ,KoO. Silicate of Alumina and Potash. G. 2-442-56. Crystallizes in twenty-four-faced trapezohedrons which cannot be distinguished by the eye from those of the Monometric System. Von Rath came to the conclusion that they are Dimetric, and that the faces are those of the unit pyramid and of the pyramid (4 P 2) or (2 4). Rock-forming Minerals. 103 Weisbach has since endeavoured to show that the angles agree better with those of a Trimetric crystal, the angle between the faces of the unit prism of which is 92 1' (Jahrbuch fiir Mineral., 1880, i. 143). Mallard concluded that the mineral is Monoclinic, while Levy and Fouque suggest that his observations rather tend to indicate that it is Triclinic. But however this may be, the crystals differ but very slightly, if at all, from Dimetric forms. Cleavage none or very imperfect. H. 5'5 6. Brittle. Conchoidal fracture. White or smoky or ashy grey. Infusible. Decomposes easily when boiled in strong Hydrochloric Acid with separation of pulverulent silica. Solution gives a strong Potash colour in Bunsen flame when viewed through Cobalt glass ; usually there is enough Soda present to mask this colour when seen with the naked eye. The crystalline shape and ashy-grey colour of Leucite generally enable us to recognise it with fair certainty. When crystallized the only common mineral it can be mistaken for is Analcime ; but Anal- cime is fusible and gives water in the closed tube. Crystalline grains of Leucite can be distinguished from Felspars by their infusibility, the absence of cleavage, and the ready decomposition by acid : from Nepheline by the strong Potash colour which the solu- tion gives in the Bunsen flame. In rock-slices crystals of Leucite are generally polygonal in outline with a strong tendency to be eight-sided ; the angles how r ever are frequently rounded, and sometimes this is carried so far that the section becomes almost circular. Sometimes several crystals are closely packed together, forming what looks like a single crystal traversed by a polygonal network of cracks. The crystals are colour- less and usually clear and transparent, but they very commonly contain small enclosures ; these appear sometimes as granules and sometimes as microliths, and in many cases they are arranged in concentric bands parallel to the external contour, but radial and other kinds of grouping do occur. The crystals are occasionally completely surrounded by a border of microliths of Augite lying parallel to the outline. The difference between the two indices of refraction is very small in Leucite, hence it never gives brilliant colours with polarized light ; with crossed Nicols the tint is dark bluish grey, and this alters so slightly when the section is rotated that the mineral might well be believed to be singly refracting : by this property, when other tests fail, Leucite can be distinguished from Nepheline. Polysynthetic twinning is common in Leucite, the composition plane being parallel to (2 P ) ; sections of twinned crystals are seen in polarized light to be traversed by fine parallel striations, or narrow bands of various tints of dark grey : in sections cut in the right direction four series of such bands are seen. The twinning is com- monest in the larger and is often absent in small crystals. Even where only one set of twin bands is visible, the dull grey colour and feeble contrast between the tints of adjacent bands distinguishes Leucite from the Triclinic Felspars, which show either brilliant colours, IO4 Geology. or strongly-contrasted white and neutral tint. The twin banding is sometimes best brought out by rotating the section between crossed Nicols. (B 3) SODALITE GROUP. This group contains three minerals whose composition may be repre- sented bv the following formulae. Sodalite. 3(NaoOSiO.,Al.,O 3 Si0 2 )2NaCl. Hauyne and Nosean. 2((Ca:Na 2 )OSiO 2 .Al 2 O 3 SiO 2 }(Ca:Na 2 )SO 4 . They are all silicates of Alumina and Soda, with the addition in the case of Sodalite of Sodium Chloride, and in the case of Nosean and Hauyne of Sodium and Calcium Sulphate. Some Sodalites contain a little Lime, but the above formula repre- sents the normal composition. In many Noseans the Lime is so small in amount that their com- position may be written 2(Na 2 OSiO 2 Al 2 3 Si0 2 )Na 2 S0 4 , and they may be looked upon as silicates of Alumina and Soda with Sodium Sul- phate. Hauyne is richer in Lime than Nosean. In fact Hauyne and Nosean may be looked upon as varieties of the same mineral : Hauyne containing a preponderance of Calcium Sulphate and often scarcely any Sodium Sulphate, and Nosean a preponderance of Sodium Sulphate and often scarcely any Calcium Sulphate. All three of the minerals usually contain a little Potash. All three are Monometric, have a specific gravity of about 2 '25, and a hardness of about 5 '5, and are decomposed by acids. Sodalite when crystallized takes dodecahedral and other monometric forms. It has a more or less perfect dodecahedral cleavage. It is common in a massive imperfectly-crystallized state. Very variable in colour. A massive specimen from Julianhaal, Greenland, bluish green mottled with white, after several hours' boiling in strong Nitric or Hydrochloric Acid was decomposed and the silica separated in a pulverulent form. The above methods of treatment will not detect the Chlorine in Sodalite, because that element is driven off by heat. But if the coarsely-powdered mineral be placed in a small glass bottle and strong Nitric Acid be poured on it and the whole allowed to stand for at least twenty-four hours, partial decomposition is produced. A part of the acid is drawn off and a solution of Nitrate of Silver added : the liquid becomes at once turbid, and in a few hours' time a white precipitate, which gradually blackens when exposed to the light, settles down. Fusible B.B. ; in an irregular fragment, whose extreme dimensions were 2x4x7 mm., it was not difficult to fuse the projecting points with some intumescence to a clear glass. When the powder was made into a paste and dried, a bit about 2x2x1 mm. fused altogether into a glass. Hauyne crystallizes in dodecahedrons, octahedrons, and other mono- metric forms. It has dodecahedral cleavage. In rocks it occurs very frequently in rounded glassy grains, looking like crystals which have been partially fused; or it coats cavities looking as if it had been run in in a fused state. Rock-forming Minerals. Usually bright blue, but also bluish green and green, slightly bluish to white. By boiling in strong Hydrochloric Acid it is decomposed with separa- tion of pulverulent silica ; by long digestion in dilute Hydrochloric Acid without boiling it is partly decomposed and the solution gelatinizes. If the Alumina be precipitated from the solution by Ammonia and separated by filtration, a portion of the filtrate treated with Barium Chloride becomes instantly cloudy and shortly a white precipitate insoluble in Hydrochloric Acid settles down, indicating the presence of Sulphuric Acid ; another portion treated with Ammonium Oxalate gives after a time a white precipitate, indicating the presence of Lime. Fusible : a rudely spherical fragment of blue Hauyne from Vesuvius, about 1 mm. in diameter fused readily to a colourless blebby glass. Nosean is generally darker in colour than Hauyne, sometimes black. B.B. and with acids it behaves similarly to Hauyne, but in the more normal varieties the precipitate with Ammonium Oxalate will be either wanting or very small. Of the three minerals in this group Hauyne is the easiest to recognise on account of its marked colour. Of the minerals which are usually of a bright blue, Allophane is much softer ; Lapis Lazuli, if it be really a distinct species, differs mainly from Hauyne in con- taining Sodium Sulphide, and hence when decomposed by Hydrochloric Acid it gives off Sulphuretted Hydrogen ; Lazulite, a hydrated phos- phate of Alumina and Magnesia, when moistened with Sulphuric Acid and heated B.B., gives the bluish green of Phosphoric Acid; it is also unaffected by acids ; Azurite, blue Carbonate of Copper, is softer than Hauyne and effervesces with warm acid. In the case of a blue mineral which has the physical characters of Hauyne, the tests described will detect the presence of Sulphuric Acid and Lime, if it be that mineral, and completely identify it. The dark colour of Nosean will usually distinguish it from Hauyne, and it generally gives a feeble reaction or none at all for Lime. In the case of a mineral which has the physical characters and look of Sodalite the test for Chlorine should be applied. If the mineral be an anhydrous silicate containing Chlorine it must be either Sodalite or Scapolite,* and Scapolite contains lime. It seems scarcely possible to fix on any characters by which Hauyne and Nosean can be distinguished in microscopic sections. The sections of their crystals are usually four- or six-sided : since they crystallize in the Monometric System they as a rule remain dark during a whole revolution under crossed Nicols, when they are unaltered. Cases have been observed where they show a certain amount of double refraction, but this is probably due to the crystals being in a state of strain. They both are very generally filled more or less with dark dusty matter, which consists of very small granules, minute glass enclosures, and microliths. This dust forms a dark-brown border to the crystals, or is distributed in bands parallel to their outline, or occurs in irregular patches, or is arranged in very fine parallel lines * F. D. Adams has shown that Scapolite contains up to 2 '4 per cent, of Chlorine: Journ. Chem. Soc. xxxvi. 697. IO6 Geology. giving the appearance of dark striae. This dusty matter often spreads out beyond the boundary of the crystal and forms a dark encircling band, which grows gradually less and less opaque outwards. These enclosures of dusty matter seem to be never absent from Nosean, while Hauyne is occasionally free from them. Sodalite differs frequently from both Hauyne and Nosean in being perfectly free from dusty matter, but this is by no means always the case, and when it occurs in crystals it cannot be distinguished with certainty under the microscope from these minerals. It is met with sometimes in a compact form moulded on the surrounding minerals, and this has never been observed to be the case with Hauyne and Nosean, which are always in crystals, fragments of crystals, or crystal- line granules. There seems to be such good grounds for looking upon Hauyne and Nosean as varieties of the same mineral, that their distinction is perhaps of no great importance, and Sodalite can in many cases be discriminated from them only by chemical tests. A very convenient method is to place a little of the powdered mineral on a glass slide, add a few drops of dilute Hydrochloric Acid, and allow the whole to stand till the acid is evaporated. In the case of Hauyne long needle- shaped crystals of Gypsum are seen in the residue when the slide is examined with a low power : they occur singly, or arranged in star- shaped clusters, or shoot out from the grains of unclecomposed mineral. In the case of Sodalite cubical crystals of Common Salt are scattered through the residue. (B 4) THE PYROXENE AND AMPHIBOLE GROUPS. The minerals of these groups are silicates of various Protoxide bases, and their chemical composition can be represented by the general formula ROSiO 2 . The commonest Protoxides (RO) are Magnesia (MgO), Lime (CaO), Ferrous Oxide (FeO) : Manganous Oxide (MnO), Zinc Oxide (ZnO), Soda (Na 2 O), Potash (K,O), Lithia (LLO), and Water (H 2 0), also occur in small quantity in some varieties. Sesquioxides replace a portion of the silica in some members of the group : Alumina (A1 2 O 3 ) is the most important of these, but Ferric Oxide (Fe 2 O 3 ) and Manganic Oxide (Mn 2 O 3 ) occur as well. The varieties containing Sesquioxides are known as the Aluminous, while those containing only Protoxides are distinguished as the Non- aluminous members. The most important difference between the two groups in composition is that in all the Pyroxenes Lime is an important constituent, while it is wanting or occurs only in very small quantity in some Amphiboles. Both Pyroxenes and Amphiboles are Monoclinic ; but in the Pyroxenes the angles between the faces of the unit prism are 87 5', 92 5', while in the Amphiboles these angles are 124 30' and 55 30'. H. in both groups 5 6. The fusibility of both groups is very variable : as a rule the varieties containing the largest percentage of Iron are most fusible. Most varieties are not acted on by acids. Rock-forming Minerals. 107 The following are the commonest members. 1. Non-aluminous AmpkiboJ.es. Tremolite, Grammatite.(CsiO'MgO)SiO^ Lime-magnesia Amphi- bole. G. 2-9 3*1. Commonest shapes, long-bladed crystals, or groups of thin fibrous crystals radiating from a centre : sometimes compact. H. 5 6 '5. Generally white or greyish. Nephrite or Jade has the same general composition as Tremolite ; but it does not occur crystallized. It is tough, breaks with a splintery fracture and glistening lustre, and has a greenish or bluish tinge. H. 66-5. Actinolite, Strahlstein. (CaO:MgO:FeO)SiO. 2 . Lime-magnesia-iron Amphibole. G. 3 3 "2. Generally in long thin crystals, sometimes arranged in radiated forms. Also granular. Colour green. When in long bright-green translucent crystals it is called Glassy Actinolite. Tremolite, Actinolite, and other Non-aluminousAmphiboles frequently put on forms so fibrous that they can be separated by the fingers into thin, soft, cotton-like or silky threads. These very fibrous varieties are called Asbestus or Amianthus. Sometimes the fibres are matted together in such a way as to form a felt-like substance, which is known as Mountain Leather. 2. Aluminous Amphiboles. Hornblende. Aluminous Lime-magnesia-iron Amphibole. G. 3 '05 3-47. Very distinct cleavage parallel to the lateral faces of the unit prism. Tough. H. 56. Various shades of green up to black. The green and bluish-green varieties are sometimes distinguished as Pa.rgasite. Smaragdite. Aluminous Lime-magnesia Amphibole. G. 3. H. 5. Foliated ; light grass-green colour. 3. Non-aluminous Pyroxenes. Diallage. (CaO:FeO:MgO)SiO 2 . Lime-magnesia-iron Pyroxene. G. 3-23-25. Splits up into thin plates bounded by planes parallel to the ortho- pinacoid. Whether this structure is true cleavage will be discussed presently. H. 4. Various shades of green. The ease with which Diallage splits along the orthopinacoid gives it a foliated structure, which is characteristic. Sahlite. Lime-magnesia-iron Pyroxene. G. 3*25 3 -4. Crystals usually with rounded edges and angles. Cleaves parallel to the base. Various shades of green, but sometimes grey or white. The absence of foliation distinguishes it from Diallage. Dwp&ide, Malacolite. (CaO:MgO)SiO2. Lime-magnesia Pyroxene. G. 3-23-38. Cleaves parallel to the base and orthopinacoid. io8 Geology. White or grey to pale green. The Non-aluminous Pyroxenes put on fibrous forms and give rise to varieties corresponding to Asbestus. But the asbestiform varieties are far less common than in Amphibole. 4. Aluminous Pyroxene. Augite. Aluminous Lime-magnesia-iron Pyroxene. G. 3 '23 3 '5. Cleavage parallel to lateral faces of unit prism sometimes distinct, also sometimes parallel to the base. In other cases no good cleavage. H. 56. Brittle. Various shades of green, through greenish black up to black. The clear green varieties are sometimes distinguished as Fassaite. Note on Diallage and Augite. We have provisionally retained Diallage as an independent species, but it is a question whether the platy or laminated structure, which is characteristic of Diallage and is supposed to distinguish it from Augite, is due to true cleavage. Rosenbusch * is of opinion that the tendency to divide parallel to the orthopinacoid is caused by the juxtaposition of a number of twin lamellae whose composition planes are parallel to that face. In short, that division takes place not along cleavage planes, but along planes of twin composition. It is in favour of this view that thin films of Calcite and other foreign substances are interposed in some cases between adjoining lamellae. The minerals of this group which have been just described are none of them rare, but by far the commonest and the most important as constituents of rocks are Hornblende and Augite. There is nothing in chemical composition or blowpipe reactions which will serve to distinguish these species, and we have to rely mainly on their crystalline form and habit and their optical characters to tell one from the other. The unit prisms of the two species are widely different. In Augite the faces of this prism are inclined at angles which cannot be distin- guished by the eye from right angles : in Hornblende the corresponding angles differ widely from a right angle. In natural crystals pinacoids are usually present in addition to the prismatic faces. In Hornblende very frequently only the clinopinacoid occurs and the cross section of the crystal is six-sided. Fig. 62 shows such a crys- tal and fig. 62 a cross section of it perpendicular to the vertical axis. Here the angles at A and D are evi- dently far from being right angles. Even if these angles had been cut off by orthopinacoids, it would be easy to Fig. 62. see that the prismatic faces if produced would not include anything like a right angle. * Mikros. Physiog. i. 303 ; ii. 462. Rock-forming Minerals. 109 The crystal in fig. 62 is terminated at each end by three planes. The two at the lower end, marked + P, and the corresponding planes at the upper end, are the faces of the positive Unit Hemioctahedron. The plane at the top, marked oP, and the parallel plane at the bottom, are basal faces. In Augite it very often happens that both orthopinacoids and clinopinacoids are present, so that the cross section of the prism is eight-sided. Fig. 63 allows such a crystal, and fig. 63 a cross section perpendicular to the vertical axis. a; 2- ^V D Fig. 62a. Here it is clear enough that the faces AH, EG are as near as may be at right angles to one Fig. 63. Fig. 63a. another, and each of the angles of the octagon is very nearly 135. Such a cross section cannot possibly be confounded with that on fig. 62. We must not forget however that the obtuse angle of the Horn- blende section will grow smaller and the angles of the Augite section will also change in size as the inclina- tion of the plane of section to the vertical axis decreases, and before we apply this test we must be sure that the section is nearly perpendicular to this axis. The crystal in fig. 63 is terminated by basal planes and faces of the Unit Hemioctahedron. Fig. 64 shows a similar crystal but without the basal planes. Both are very common forms in Augite. Fig. 65 is a twin derived from the crystal in fig. 64. The composition plane is parallel to the orthopinacoid, and the vertical axis is the twinning axis. If the crystal in fig. 64 were divided by a plane parallel to oo P GO, each of the faces P will be cut in two. After the revolution the part which originally sloped down to the right slopes down to the left, and we get the re-entering angle at the summit. Fig. G4. no Geology. As a general rule Hornblende crystals tend to be long and columnar ; and in the Amphibole group generally needle-shaped, fibrous, and radi- ated shapes prevail. In the crystals of Augite on the other hand short, stout forms are more common. Hornblende and Augite also differ in their cleavage. Hornblende has perfect cleavage parallel to the faces of the unit prism. Augite also cleaves sometimes parallel to the faces of its unit prism ; but not uncommonly its cleavage is so indistinct as to be scarcely recognisable. But even when the cleavage of Augite is well marked, it is easily distinguished from that of Hornblende, for the angle between the two directions of cleavage is nearly 90, while the corresponding angles in Hornblende are 124 30' and 55 30'. The perfect cleavage of Horn- blende often gives rise to a platy structure, which is somewhat charac- teristic. When thin sections of rocks are examined under the microscope, the difference in their cleavage often enables Hornblende and Augite to be easily distinguished, provided the section is perpendicular or nearly perpendicular to the vertical axis. In microscopic sections the dichroism of Hornblende furnishes another means of discriminating between it and Augite. If the analyzer be removed and the polarizer rotated, the colour of Hornblende changes distinctly as the polarizer turns round ; with Augite no change or very little occurs. There are exceptions to this rule, but in most cases it is a trustworthy test. In Hornblende sections parallel to the vertical axis are generally elongated parallel to that axis and show one set of cleavages parallel to the longer edges. They extinguish when the cleavages are parallel or nearly parallel to a diagonal of the Nicols. This is shortly expressed by saying they extinguish parallel to the cleavages. Sections parallel to the orthodiagonal are either lozenge-shaped if they are bounded by the faces oo P, or six-sided if by GO P and oo P oo, or eight-sided if by oo P, 00 P oo, oo P oo : they show two sets of cleavages and extinguish parallel to the lines bisecting the angle between these cleavages. In Augite sections parallel to the vertical axis are often elongated parallel to that axis and show one set of cleavages parallel to the longer edges. AVhen the section coincides with the clinopinacoid the direc- tion of extinction makes an angle of 39 with the cleavages ; as the plane of the section approaches the orthopinacoid this angle decreases, and in an orthopinacoid section the directions of extinction coincide with the cleavage. Sections parallel to the orthodiagonal, if they show any definite outline, are generally eight-sided : they may show two sets of cleav- ages, one and sometimes both of which are very irregular, and they extinguish parallel to the lines bisecting the angles between the cleavages. Augite gives very vivid colours in polarized light, and in some cases the surface of its sections has very much the roughened look of those of Olivine. There is then some risk of confounding the two minerals. To avoid this error we may bear in mind that when Augite shows cleavage, there are two sets, unless the section be cut parallel to the Rock-forming Minerals. 1 1 1 vertical axis, and that they run with some degree of irregularity, but yet preserve on the whole a general tendency to be perpendicular to one another. Olivine, where its cleavage is visible, shows only one set of cleavage lines : frequently it shows no cleavage, and it is usually traversed by numerous irregular cracks. We may also bear in mind that when the minerals occur in elongated crystals, the pointed termin- ations tend to be more acute in Olivine than in Augite. Uralite is a mineral crystallizing in the same form as Augite, but with the cleavage of Hornblende. That some Uralite is altered Augite is shown by the presence of a kernel of unchanged Augite in the middle of the crystals. (B 5) THE ENSTATITE OR TRIMETEIC PYROXENE GROUP. Three minerals, Hyperstliene, Enstatite, and Bronzite, which have been looked upon as distinct species, constitute this group. The chemical composition of all three may be represented by the formula SiO 2 (MgO:FeO), that is they are silicates of Magnesia and Ferrous Oxide. Hypersthene contains the largest percentage of Iron, Bronzite not so much, while in Enstatite there is very little, perhaps sometimes none, of that element. So that if we take Enstatite as the type and look upon it as a silicate of Magnesia, the other two are, as far as composition goes, varieties of Enstatite in which part of the Magnesia is replaced by Ferrous Oxide. These minerals have then the same general composition as the Pyroxenes, but they crystallize in Trimetric forms, and hence they are sometimes called the Trimetric or Rhombic Pyroxenes. The angles between the faces of the unit prism are practically the same for all three, 93 and 87 on an average, but they vary slightly with the percentage of Iron. Differences in cleavage have been supposed to distinguish the three. All three cleave parallel to both the faces and the pinacoids of the unit prism, but the statements of different authors as to the relative importance of these cleavages in the three minerals are very conflicting. The truth seems to be that in Hypersthene and Bronzite there are generally present a large number of minute foreign bodies of platy shape, ranged in layers with their flat faces parallel to the brachy- pinacoid, and that the presence of these scaly films makes the brachy- pinacoid cleavage more pronounced than the prismatic or clinopinacoid cleavages. In what has been known as Enstatite these enclosures are absent, and hence in it the prismatic cleavage is the most strongly marked. The minerals do differ in fusibility. Enstatite and Bronzite are fusible only on the edges of the thinnest splinters : Hypersthene fuses to a black enamel. But this is obviously owing to the large percentage of Iron in Hypersthene and is not a distinction of specific value. It is for the same reason that while the non-ferriferous forms are not attacked by Hydrochloric Acid, Hypersthene is in some cases partially decomposed. Again the optical properties have been supposed to furnish a ground for separating Hypersthene from Enstatite and Bronzite. In all three 1 1 2 Geology. the optic axes lie in the plane passing through the brachy diagonal * and the vertical axes, but in Enstatite they make acute angles with the vertical, in Hypersthene with the brachydiagonal, axis. The angle however between the optic axes varies very largely, and it seems likely that their inclination to the vertical axis increases with the percentage of Iron, and that a gradual passage can be traced from these varieties in which the optic axes lie nearer to the vertical than to the brachydiagonal axis, into those in which the reverse is the case ; in short that the inclination of the optic axes to the vertical axes depends solely on the extent to which Magnesia is replaced by Ferrous Oxide. Further Des Cloiseaux states that in Hypersthene the angle between the optic axes is greater for red than for violet light, while in Enstatite and Bronzite the reverse is the case. It can be easily understood that this will be true when we compare a typical Hypersthene rich in Iron with a typical Enstatite almost free from that metal, but it remains to be shown whether there is not in this particular also a passage from one extreme to the other. It must also be borne in mind that the variations in the size of the angle between the optic axes for different rays are very small. One point however, which is often very useful for purposes of identi- fication, does seem to be ascertained. Hypersthene is strongly dichroic, almost as strongly as Hornblende ; + such dichroism as Enstatite and Bronzite show is at the best but feeble and quite inappreciable in thin sections. All the facts just enumerated, with perhaps the exception of the last, lead to the conclusion that Hypersthene, Bronzite, and Enstatite are merely varieties of one mineral ; but the varietal differences between these are sufficient to justify us in retaining all three names. These differences we will now specify. Hypersthene, Paulite. G. 3 '39. H. 5 6. Generally dark in colour, greenish or blackish to black. The specimens of this variety that have been most carefully examined are from the Island of St. Paul, Labrador, and from some Norwegian localities. They all con- tain numerous minute brown scales of some foreign substance, ranged in planes nearly parallel to the brachypinacoid and lying with their flat surfaces parallel to that plane. Owing to these interpositions the brachypinacoid cleavage is intensified and the mineral has a platy structure. They also have the effect of producing a somewhat metallic- lustre on the faces of these cleavage planes. Hypersthene also fre- quently contains disseminated grains of Magnetite. Bronzite. G. 3'1 3'3. H. 5'5. This name may be taken to include these varieties which are less rich in Iron than Hypersthene, but like it contain platy inclusions ranged parallel to the brachy- pinacoid and therefore cleave most conspicuously parallel to that face. * The brachydiagonal is the axis of greatest, the vertical the axis of least, elasticity. t So say all the books. I must add that the only two sections of the typical Hypersthene (Paulite) of St. Paul's Island which I have examined did not show the faintest trace of dichroism. Rock-forming Minerals. \ \ 3 The interposed layers are often very numerous and close together, so that the mineral has a micaceous or foliated structure, and they pro- duce on the cleavage planes a bronze lustre. Enstatite. G. 3'1 3'13. H. 5'5. This includes the varieties very poor in Iron and comparatively free from foreign included bodies. Consequently the prismatic is more marked than the pinacoid cleavages. Generally paler in colour than Hypersthene and Bronzite, whitish, yellowish, greyish, or greenish white. Hypersthene is a good deal like the darker varieties of Hornblende and Augite. If it occurs in well-formed crystals we can sometimes see that the prismatic faces and the pinacoids are all perpendicular to the basal face, and so distinguish it from the Monoclinic forms of these minerals. Its strong brachydiagonal cleavage also distinguishes it from these minerals, both of which cleave only imperfectly in this direction. In rocks Hypersthene usually occurs not in well-defined crystals but in crystalline grains or masses with, irregular outlines and moulded on the adjoining minerals. Here its strong dichroism at once dis- tinguishes it from Augite, from Diallage which it resembles in its platy structure, and also from Enstatite and Bronzite. From Hornblende it may be discriminated by the fact that it shows either only one cleavage, or three cleavages one of which is much more strongly marked than the other two. Hornblende generally shows two cleavages of equal importance. When the prismatic cleavages are visible the angle between them at once distinguishes it from Hornblende. These tests would fail if the section happened to be cut nearly parallel to the brachypinacoid, and in that case special optical methods have to be resorted to. Enstatite and Bronzite are distinguished from Hypersthene, Augite, and Hornblende by their difficult fusibility. Its marked foliation and the brassy lustre of its cleavage planes are the main characteristics of Bronzite, and will serve to distinguish it from Enstatite. In rocks these two minerals usually occur in irregularly outlined crystalline masses. Their feeble dichroism distinguishes them from Hornblende. The strong brachydiagonal cleavage of Bronzite will distinguish it from Augite, and even Enstatite often has this cleavage well enough marked to discriminate it from this mineral. But it is not so easy to tell them from Diallage which is just as well foliated. They may frequently be distinguished thus. If a section of Enstatite or Bronzite cut nearly perpendicular to the brachypinacoid be rotated between crossed Nicols, it will extinguish when the diagonals of either of the Nicols is parallel or perpendicular to the edges of the foliar. Diallage in certain sections also extinguishes parallel to the edges of the lamella, but in the majority of sections its directions of extinction make angles lying between 35 and 39 with these edges. (B 6) THE OLIVINE GROUP. Only one of the minerals of this group is common enough to require notice here. Olivine, Peridote, Chrysolite. SiO 2 .2(MgO:FeO). The amount of Iron varies very much, some varieties contain none or scarcely any ; H 1 14 Geology. those very rich in Iron are sometimes distinguished as Fayalite. G. 3-33-5. Trimetric. Cleaves very imperfectly parallel to the macropinacoid, rather more distinctly parallel to the brachypinacoid ; but in many cases the cleavage is not strongly marked in either direction. Con- choidal fracture. H. 6 7. Very generally of an olive-green colour, but sometimes brownish or yellowish. Glassy lustre. Varieties poor in Iron are infusible alone, but fuse with Borax to a green glass : highly ferruginous varieties fuse alone to a black magnetic globule. Decomposed by Hydrochloric Acid, the ferruginous varieties most easily. When boiled in strong acid, the silica separates in the pulveru- lent form. Gently heated for several hours in dilute acid, the' mineral is largely decomposed, and when the solution is evaporated, silica separates in gelatinous clots. Olivine usually occurs in rocks in glassy blebs without crystalline shape, sometimes in small crystals. Some lavas contain rounded nodules of this mineral, which occasionally reach the size of a man's head. Blebs of Olivine are very like broken bits of green bottle-glass when the mineral has its characteristic olive-green colour, and in such cases the look of the mineral is so marked that it may be recognised with very little risk of error by its appearance alone. When the typical colour is exchanged for some other tint and the mineral is not distinctly crystallized, Olivine is a good deal like Quartz in hardness and fracture, and might on a hasty examination be mistaken for it. If we are dealing with one of the fusible varieties this alone distinguishes it from Quartz, and in any case the solubility of Olivine in acids makes it impossible to confound the two. Three characters combined will generally enable us to identify Olivine under the microscope. The surface of a slice never attains a perfect polish, but has somewhat the look of ground glass : this is seen by ordinary light. The crystals or grains are traversed by numerous irregular cracks. With polarized light no mineral gives such gorgeous colours : those of Quartz approach them, but the surface of a slice of Quartz takes a perfect polish. Another very marked characteristic is the alteration into serpentinous products which so frequently happens in this mineral. They occur as bands edging the cracks, or replacing parts of the crystals. Hexagonal sections of Augite bounded by prismatic faces and clino- pinacoids and elongated parallel to the clinodiagonal, are sometimes hard to tell from the corresponding sections of Olivine (the brachy- diagonal corresponds to the clinodiagonal). But they may often be distinguished by noting that the angle between the prismatic faces is more acute in Olivine than in Augite. (B 7) MICA GROUP. The composition of the minerals of this group is complicated and variable and can hardly be represented by a chemical formula.* * Tschermak has endeavoured to reduce the chemical composition of the Micas to a system; Zeit. fur Krystallog. iii. (1879) 122. Rock-forming Minerals. 1 1 5 They all contain Silica, Alumina, and Potash. In many cases Fluorine is present, probably as Silicon Fluoride which replaces part of the Silica. Some varieties also contain Titanic Acid. The Alu- mina is also in some varieties partly replaced by Ferric, Manganic, or Chromic Oxides. According to their other constituents the Micas may be grouped as follows. 1. Alkali Micas. In these Magnesia and usually Ferrous Oxide are present only to a small extent, and the Alkalies become consequently the characteristic constituents. They also sometimes contain small quantities of Lithium, Caesium, and Rubidium. 2. Magnesia and Magnesia-iron Micas. Containing, besides Alkalies and occasionally Lithia, Magnesia and Ferrous Oxide in considerable quantity. The proportions of Magnesia and Ferrous Oxide vary very much ; the latter is almost absent from some varieties. 3. Lime and Baryta Micas, in which in addition to Alkalies, Mag- nesia, and Ferrous Oxide, Lime and Baryta are important ingredients. These are comparatively rare minerals. There is some uncertainty as to the crystalline systems to which the Micas should be referred. All crystallize in prismatic forms the cross section of which has the angles of a regular hexagon. Some are apparently uniaxal, and these have generally been looked upon as hexagonal. There are grounds however for thinking that these Micas are not really uniaxal, but that they are biaxal minerals in which the inclination of the optic axes to one another is so small that very refined and delicate methods of observation are needed for its detec- tion. Other Micas are distinctly biaxal, and these have been referred to Trimetric shapes in which the angles of the unit prism are 120 and 60, and which therefore show a pseudo-hexagonal cross section, when the bounding faces of the crystal are those of the unit prism and brachypinacoids (see p. 73). Tschermak has endeavoured to show that all are really Monoclinic, but that the inclination of the vertical axis to the plane of the lateral axes differs so little from a right angle, that their Monoclinic character might be easily overlooked. The angles between the faces of the unit prism he makes 120 and 60.* But whatever be the true view on these points, the following facts are unquestionable and of great value in enabling us to recognise Micas. They all have a very perfect basal cleavage, so that they split readily into very thin laminae that are both flexible and elastic. Their hardness does not exceed 4, and all the common species are soft enough to be scratched by the finger-nail. The following are the most important species. Potash Mica, Muscovite. An Alkali Mica, Potash predominating over Soda. G. 2'75 3-1. Distinctly biaxal. Frequently in rhombic or hexagonal plates; sometimes in irregularly-shaped scales. H. 2 2 -5. Mostly silvery white, but occasionally dark coloured. * Zeit. fur Krystallog. ii. (1878) 14; see also Bauer, Miner. Mittheil. i. 1. 1 1 6 Geology. B.B. whitens and fuses on thin edges to a grey or yellow glass. Scarcely, if at all, attacked by acids. Muscovite is found in Russia in plates large enough to allow of its being used for the panes of lanterns and windows, and other similar purposes ; it is then known as Muscovy Glass. Maryarodite, Gilbertite, Damourite, and Sericite are hydrated Musco- vite, probably produced by the alteration of that mineral. Fuschite is a Muscovite containing Chromium. Lepidolite, Lithionite, Lithia Mica, Lithionglimmer, Zinnwaldit. An Alkali Mica in which part of the Potash is replaced by Lithia. G. 2-853. Usually in scaly or granular aggregates rather than in crystals. Distinctly biaxal. H. 2*54. Often rose-red or violet, but variable in colour. B.B. fuses very easily with some intumescence ; when the percentage of Lithia is large it gives a very intense purplish carmine colour to the flame ; in Lepidolites not so rich in Lithia the colouration is fainter and appears only at the moment of fusion ; in such cases it requires to be looked for carefully. Imperfectly decomposed by acids. After having been fused is readily dissolved in dilute Hydrochloric Acid, and it is stated in most works on Mineralogy that the silica gelatinizes. In the only two ex- periments I have made it separated in a pulverulent form ; in one the acid was diluted with eight parts of water and was never allowed to boil. Most Lepidolites contain Fluorine. In some cases this is given off, by merely heating in a closed tube, in sufficient quantity to turn moist Brazil-wood paper yellow, but in some cases I have found it neces- sary to fuse the mineral with Bisulphate of Potash in a closed tube in order to get this reaction. Cryophyllite is a variety of Lepidolite which is green by transmitted light. Pldogopite. A Magnesia Mica containing little iron. G. 2 '78 2-85. It occurs not unfrequently in six-sided prisms with a cross section approximating to a regular hexagon and irregular lateral faces ; also in small plates. Distinctly biaxal. H. 2 '5 3. The thin laminae often show a stellate figure when looked at through a candle. Very variable in colour, frequently brown or copper-red. B. B. whitens and fuses on the thin edges. Attacked by boiling dilute Sulphuric and Hydrochloric Acids, but very long boiling is required f in- complete decomposition. Biotite. A Magnesia-iron Mica. G. 2 '7 3'1 Most frequently disseminated in scales. H. 2 '5 3. Colour usually black or dark green. B.B. whitens and fuses on the thin edges. Completely decomposed by dilute Hydrochloric or Sulphuric Acid, leaving a residue of glistening scales of silica. Lepidomdane. This mineral holds in a measure an intermediate Rock-forming Minerals. 1 1 7 place between the Alkali- and the Magnesia-iron Micas. It contains up to 10 per cent, of Potash, up to 12 per cent, of Ferrous Oxide, and very little Magnesia. The Alumina is very largely replaced by Ferric Oxide, some varieties containing 35 per cent, of Ferric Oxide and only 9 per cent, of Alumina, and on an average the percentage of Ferric Oxide is larger than that of Alumina. G. 3'0. In six-sided tables or scales. H. 3. Only slightly elastic. Black. Streak, greyish green. B.B. turns brown at a red heat and fuses to a black magnetic globule. Easily decomposed by Hydrochloric Acid, leaving scales of silica. That a mineral is a Mica is usually a point that can be settled with- out much difficulty. Their marked basal cleavage causes the Micas to split into very thin plates, and these plates are flexible and elastic. The only other two common minerals that split up to the same extent are Talc and Selenite. The laminae of Talc are flexible, the laminae of Selenite are imperfectly flexible, but neither are elastic. The soft- ness too of the Micas is characteristic ; the common species can be scratched by the finger-nail. But the species of a Mica can in many cases be determined only by quantitative analysis. The following considerations however often help us to an approximate determination. Lepidolite and Lepidomelane are easily fusible, the other common Micas are fusible only on thin edges. Lepidolite gives a very distinct Lithia colour to the flame, and this distinguishes it from Lepido- melane : further Lepidomelane is easily, Lepidolite only very slowly, decomposed by Hydrochloric Acid. Of the three other species described above, Biotite is completely decomposed by acids ; Phlogopite is attacked very slowly and Mus- covite scarcely at all. Biotite may also be distinguished from Phlogo- pite, and often from Muscovite also, by noting that Biotite gives a strong Iron reaction with Borax or Microcosmic Salt, Phlogopite gives this reaction very feebly or not at all ; this will also be very generally the case with Muscovite. Colour, though to some degree useful, is an uncertain guide. Mus- covite is usually silvery ; the Magnesia-iron or Iron Micas dark in colour ; Lepidolite is often of a violet or reddish hue, and Phlogopite brown or copper-coloured. But there are frequent exceptions in the case of each. In rock sections Micas seldom occur in well-defined crystals, but have generally ragged irregular outlines ; occasionally if the section is cut nearly perpendicular to the vertical axis they present hexagonal contours. In sections cut across the cleavage, the minute lamination is very marked, the mineral being striped across by a large number of parallel lines very close to one another. In sections perpendicular to the vertical axis they are practically dark between crossed Nicols dur- ing a whole revolution ; in any other section they extinguish when the edges of the cleavage planes are parallel to a diagonal of the Nicols. 1 1 8 Geology. Muscovite gives tolerably strong colours with polarized light in sec- tions inclined at large angles to the base. In sections parallel to the base no Mica shows any sensible dichroism. In sections across the cleavage Biotite and Phlogopite show the strong- est dichroism, generally an absolute change of colour; while Muscovite and Lepidolite are either altogether devoid of dichroism, or at best show only a change in depth of the same tint. There are only three minerals whose dichroism is at all comparable with that of the Magnesian Micas, Hornblende, Tourmaline, and Epi- dote. From these it may be thus distinguished. Sections of Horn- blende parallel to the base show two sets of cleavages, and this at once distinguishes them from Mica. Sections however parallel to the ver- tical axis will show only one set of cleavages, and so bear a consider- able resemblance to Mica. But the cleavage planes of Hornblende are very seldom as close together as those of Mica; also Mica has usually a much more ragged outline than the crystals of Hornblende, and the ends of its laminae have a very characteristic frayed-out look ; finally, unless the section be cut exactly parallel to the orthopinacoid, Hornblende does not extinguish when the edges of the cleavage planes are parallel to a diagonal of the Nicols. Tourmaline is limpid and has no cleavage. Epidote has usually a characteristic yellowish-green colour ; its cleavage planes are not so close together as those of Mica ; and its dichroism is not so strong. Also it occurs usually in fan-shaped aggregates of needles or fibres along the sides of fissures or on the walls of cavities. Chlorite can generally be distinguished from Mica by its green col- our, the dark Micas being usually brown by transmitted light : it is not so strongly dichroic as Mica ; and under high powers it can gene- rally be seen to be fibrous in structure which is never the case with an unaltered Mica. Petrographical microscopes are now generally fitted with an arrange- ment for observing the effects produced by viewing crystals with con- vergent polarized light, and by it the so-called " Uniaxal " Micas may be distinguished from th_e Biaxal. The best plan is to detach a small plate of Mica and view it placed between two strips of thin glass which are kept closely pressed to- gether. The " Uniaxal " Micas show a series of concentric circular coloured rings with a black cross whose arms pass through the centre of the rings and are parallel to the diagonals of the Nicols. The cross will not undergo any perceptible change when the stage is rotated. In the Biaxal Micas there are two sets of coloured rings arranged around two centres or eyes. When the line joining these eyes is parallel to a diagonal of the Nicols, there is a black cross : as the stage is rotated this cross breaks up, and, when the line joining the eyes makes an angle of 45 with a diagonal of the Nicols, it takes the form of two curved black bands or " brushes," one passing through each of the eyes. The whole of these figures is not always visible ; when the plate is very thin or the field of view very small, only portions of them can be seen. Rock-forming Minerals. 119 The following table shows the main points in the chemical com- position and physical properties of the common Micas. O c > Sos I o 1 fc IjH II- o,i & 03 pp s s 03 fl IS 020 ffb 03 i P-T^S j3 O -M ^O ^ 03 BOB > o s O OJ -M 1 a ri fc O 1 20 Geology. (E 8) HYDRATED MAGNESIAN SILICATES. Talc. A hydrated Silicate of Magnesia, probably of not very definite composition. Water varies from a mere trace up to 7 per cent. Con- tains sometimes small quantities of Ferrous, Manganous, and Nickelous Oxides, and Alumina. G. 2-5652-8. Crystals rare, hexagonal prisms and plates ; it is doubtful whether they are Trimetric or Monoclinic. Very perfect basal cleavage which causes it to split in the same way as Mica into thin plates, which are flexible but not elastic. H.I 1'5. Very sectile. Soapy or greasy to the touch. Silvery white to various shades of green, with pearly lustre. B.B. gives off water in closed tube only at a very high temperature. In forceps turns white and exfoliates and fuses only on thin edges, and then with difficulty. The paler varieties show the flesh-colour characteristic of Magnesia when moistened with Nitrate of Cobalt and ignited, but the colouration is often feeble and shows only on the edges of the assay. Unaffected by acids both before and after ignition. In its foliation Talc resembles both Mica and Selenite. Its laminae are not elastic and this distinguishes it from Mica. Except in foliation Talc and Selenite are so unlike that there is little chance of their being mistaken for one another, but in the case of any uncertainty the diffi- cult fusibility of Talc will distinguish it ; Selenite fuses quite easily. The extreme softness of Talc is also a valuable distinguishing char- acter. Its very greasy feel is also in most cases distinctive, but not always, for Professor Heddle has shown that Margarodite, a hydrated Mica, possesses this property in so high a degree that it has been mis- taken for Talc.* But he points out that "when rubbed on the lingers, however greasy ' pellicles ' of Margarodite may appear, friction merely reduces them to thinner and thinner pellicles : while Talc, when similarly treated, rubs down to an impalpable coating, the individual particles of which are not recognisable from their minuteness." t Steatite, Soapstone, Speckstein, has the same chemical composition and the same physical properties as Talc except its foliation. Its hardness sometimes exceeds that of Talc, ranging up to 2 '5. Ignited B.B. it becomes hard enough to scratch glass. French Chalk, used by tailors to mark cloth, is a finely-grained Steatite. Meerschaum or Sepiolite is another hydrated Silicate of Magnesia. G. 1 '5. It is earthy in texture, gives off water in the closed tube more readily than Talc, and gelatinizes with Hydrochloric Acid. H. 2 2'5. Absorbs water greedily. Talc and its allies belong very often, perhaps always, to the class of what are called alteration products. That is to say where we now find Talc there was originally some other mineral. This mineral has undergone a certain amount of decomposition, by which some of its original ingredients have been removed, and at the same time new ingredients, such as water, have been taken up. What remained of the original mineral has combined with the newly-acquired ingredients to form a fresh substance, Talc. * Mineralog. Mag. ii. 174. t Ibid. iii. 30. Rock-forming Minerals. Alteration of this character has taken place to a greater or less degree most extensively among the minerals of which rocks are composed. That a mineral is an alteration product can be established in several ways. Sometimes only a part of the original mineral has been changed, and a passage can be traced from the unaltered into the altered form. Sometimes where the alteration has extended throughout the entire mineral, the alteration product still retains the crystalline form and occasionally the cleavage also of the mineral from which it was derived. In many cases the alteration product has assumed a crystalline form different from that of the mineral from which it was derived. Even where the change has been as complete as this however, we can infer with a high degree of probability that many minerals have had a secondary origin, if it can be shown that there are numerous cases elsewhere in which they are unquestionably the result of alteration, and if they occur in rocks which have evidently been subjected to change. When we consider the way in which they were formed, it will be no matter for surprise that the composition of alteration products is very variable. Hornblende and Augite are two minerals which have frequently been altered into Talc. In microscopic slices Talc is colourless or pale green by ordinary light. It gives brilliant iridescent colours with polarized light, but is not clichroic. It occurs in aggregates of small plates, which sometimes show an hexagonal outline, but are often irregular in shape and have frayed outlines. The plates can sometimes be seen under high powers to be made up of fibres, arranged either in a radiating fashion or lying parallel to one another. The radiated structure distinguishes Talc from the non-dichroic Micas ; its want of dichroism from Chlorite which has often a radiated fibrous structure. Serpentine. A hydrated Silicate of Magnesia containing frequently Ferrous Oxide up to 10 per cent., and in small quantities Nickelous and Manganous Oxides and Lime, Alumina, Ferric, and Chromic Oxides. G. 2'5 2'65. Earely crystallized and then in pseudomorphs. It is always an alteration product, and this accounts for its very variable composition and pseudomorphism. H. 2 '5 4. Sectile. Sometimes rather greasy to the touch, but not so decidedly as Talc. Of various shades of green, red, and yellow, usually mottled : often veined with Steatite. In the closed tube yields water. Fuses on the edges with difficulty. Decomposed partly by Hydrochloric, often completely by strong Sulphuric Acid. The. effect of acids however on Serpentine varies very much in different examples ; it depends probably on the nature of the mineral from which the Serpentine was derived and on the extent of the alteration which has taken place. Noble Serpentine is light coloured and translucent even in large pieces. The commoner varieties are opaque. Its softness, sectibility, and general bright and mottled colouring are the most obvious characteristics of Serpentine. Chrysotile, Serpeniinasbest, is fibrous Serpentine, and occurs in veins traversing rocks composed wholly or largely of Serpentine. 122 Geology. It is very like Asbestos in look, but may be distinguished by its giving water in the closed tube and usually by its being decomposed by Sulphuric Acid. The silica separates in fine fibres. Professor Heddle states that in all the Scotch cases he has seen the fibres in veins of Chrysotile are perpendicular, in veins of Asbestos almost always parallel to the walls of the vein.* This however is not uni- versally true. Hornblende, Augite, and Olivine are among the commonest minerals from which Serpentine has been derived. Olivine partly altered into Serpentine is of common occurrence. The Olivine is traversed by numerous cracks, and each crack can be seen in microscopic sections to be bordered by a band of Serpentine, t The appearances which Serpentine presents in thin slices under the microscope are somewhat varied and depend in some measure on the mineral from which it has been derived and the extent to which the alteration has been carried. When the alteration has been complete the serpentinous product has either no action on polarized light or gives only pale-bluish or neutral tints. It has a curious gummy look and is very likely a colloidal substance. Through this there run some- times veins of a substance which has a decided action on polarized light, and which under high powers is seen to consist of fibres arranged at right angles to the walls of the veins. This is very probably ser- pentinous matter which has assumed the crystalline form of Chrysotile. When, as is so often the case, Serpentine has arisen from the alteration of Olivine, the various steps in the process may often be traced in the same slide. The change begins along the cracks which traverse the Olivine ; on either side of a crack a band of Serpentine appears and a sort of net is formed : the meshes being unaltered Olivine still retaining its brilliant polarizing power, and the web being formed of irregularly-branching threads of Serpentine of various breadths. As the process goes on the Olivine kernels grow less and less and the Serpentine bands grow broader and broader, and we at last arrive at a stage where we have a broad spread of Serpentine giving only neutral- tint colours with polarized light, thickly dotted over with tiny brilliant points of Olivine still unaltered. These at last disappear, but even in this final stage dark ramifying veins often mark the position of the great cracks in which the alteration took its rise. Serpentine which has arisen from the alteration of other minerals has similar optical properties but presents a different pattern. Thus where Enstatite has been altered into Serpentine it has first passed through a stage to which the name Bastite has been applied, and in the final product the network of cracks is more regular and rectangular in its pattern, because it represents the planes of cleavage of the original mineral. (B 9) CHLORITE GROUP. The minerals of this group are hydrated Silicates of Alumina, * Mineralog. Mag. ii. 131. t Bonney, Quart. Journ. Geol. Soc. xxxiii. (1877) 915 ; Allport, ibid. xxx. (1874) 539 ; Rutley, Study of Rocks ; Zirkel, Mikro. Beschaffenheit, 99, 216, 312. Rock-form ing Minerals. 1 2 3 Magnesia, and Ferrous Oxide. Their specific gravity ranges from 2-62-9. They have all one very perfect cleavage which splits them up into thin laminaB that are flexible but not elastic. Their hardness is low, ranging from 1 to 2 '5, seldom reaching 3. Their fusibility depends largely on the percentage of Iron. Where this is small, only the edges of thin plates can be fused B.B. Varieties rich in Iron are more fusible. They give water, but only when strongly heated, in the glass tube. They are many of them decomposed by hot Sulphuric Acid, in some cases more readily after ignition. Various shades of green are the prevailing colours. All the members of this group were once included under the single term Chlorite, but the minerals formerly grouped together under this name have been found to crystallize under different systems and therefore comprise several species. The distinctions between these species can however frequently be recognised only by careful optical examination. The following are the best established species. Chlorite, Ripidolite. Hexagonal. Crystals in small six-sided plates bounded by basal planes and faces of the hexagonal pyramid. Very common in minute scales scattered over the cleavage planes of certain rocks. Often as a thin coating intrusting other minerals. An amor- phous form is called "Peach" in Cornwall. H. 1 2. Clinochlore (Ripidolite, Dana). Monoclinic. H. 2 2 '5. Penninite. Rhombohedral. H. 2 2 '5, sometimes 3. Several other chloritic minerals have received distinguishing names ; whether they are entitled to the rank of distinct species is perhaps doubtful. The Chlorites are at once distinguished from Mica by the fact that their laminae are non-elastic. They usually differ from Talc in being decomposed to a greater or less extent by Sulphuric Acid ; they also are seldom so greasy to the touch as Talc. In thin microscopic sections of rocks patches of green translucent matter are often seen in the form either of little scales or of fibrous aggregates. Its chemical composition and mineralogical affinities are unknown, and it is provisionally designated Viridite. It is a decom- position product, allied perhaps in some cases to Chlorite and in others to Serpentine. The Chlorites have been hitherto very generally looked upon as decomposition products. Professor Heddle however is of opinion that the minerals we have placed under this head are original constituents of the rocks in which they occur. He places certain other minerals, which have previously been looked upon as belonging to the Chlorite group, in a separate class under the head of Saponites, and these he believes to be decomposition products. See Trans. Royal Soc. Edinburgh, xxix., Chapters on the Mineralogy of Scotland, chap. vi. Chloritic Minerals. As a constituent of rocks Chlorite generally occurs in irregularly- shaped scaly plates which do not lie exactly parallel to one another like the plates of Mica, but overlap somewhat after the fashion of the 1 24 Geology. scales of an onion. Sometimes detached plates are met with which have the angles of a regular hexagon. Each plate, when sufficiently magnified, is seen to be fibrous in structure, the fibres being entangled, or radiating from numerous centres. With ordinary light the usual colour is palish green ; with polarized light many Chlorites show only tints changing from blue or neutral tint to white, others polarize in brighter colours. Sections cut across the cleavage show feeble dichroism, usually only changing from one strength of green to another. (B 10) SILICATES OF ALUMINA. Andalusite. SiO 2 Al 2 O 3 . Silicate of Alumina. The Alumina is often partly replaced by Ferric Oxide and occasionally by Manganic Oxide. Lime and Magnesia are occasionally present in small quantity. G. 3-053-35. Trimetric. Angle of the unit prism 90 48'. Very frequently in prisms, the cross section of which can scarcely be distinguished by the eye from a square. Large crystals frequently coated and penetrated by Mica. H. 7 '5. Infusible, unattacked by acids. Gives readily a bright-blue colour with Nitrate of Cobalt. Its apparently square prisms, its hardness, and its infusibility enable us to recognise Andalusite easily when it is crystallized. The crystals of Andalusite are usually elongated parallel to the vertical axis, and in thin slices they show cleavage parallel to the faces of the unit prism. It polarizes in brilliant colours, and usually shows strong dichroism ; but in some thin slices which are colourless by ordinary light the dichroism is scarcely perceptible. Chiastolite (Holdspath) is a variety of Andalusite in which dark- coloured impurities are regularly distributed through the crystals in such a way that the cross section of the prisms shows a cruciform or tesselated pattern. The design takes various shapes, but in all there is a tendency to an arrangement of the foreign matter in the shape of a cross. On account of the large admixture of impurities and in some cases from partial decomposition the hardness of Chiastolite is some- times as low as 3. The presence of its peculiar pattern makes Chiastolite very easy to recognise both in large crystals and under the microscope. Cyanite, Disthem. SiO 2 Al 2 3 . Silicate of Alumina. The Alumina may be replaced to 2 per cent, by Ferric Oxide. G. 3'45 3'7. Trimetric. Usually in long, broad, blade-shaped crystals. Cleaves parallel to the macropinacoid. H. 5 7'25, least on the lateral faces. White shot with pale blue is the commonest colour. Infusible B.B. ; unaffected by acids. In many cases the peculiar blue colour of Cyanite would alone almost suffice to identify it. Where it assumes other tints, the bladed crystals are characteristic. Hornblendic minerals assume similar shapes, but they are fusible and their hardness is less than the maximum hardness of Cyanite. Rock-forming Minerals. 125 Sillimanitc, Fibrolite, is Silicate of Alumina crystallized in the Monoclinic system. G. 3-2 3'3. Crystals are generally long and slender and have an easy and brilliant cleavage parallel to the ortho- diagonal. H. G 7. Like the two minerals last described, it is infusible and unattacked by acids ; its cleavage and the habit of its crystals distinguishes it from them. Staurolite. 6Al.,O 3 SiO 2 .3(Fe:Mg)O.H.,O. Silicate of Alumina with Magnesia, Ferrous Oxide, and Water. G. 3*4 3 '8. Trimetric. Angle of unit prism 120 21'; common shape, pseudo- hexagonal prisms bounded by faces of unit prism and pinacoids. Four individuals are very often united by twinning so as to form a cross. H. 7 7 '5. B.B. infusible except the magnesian varieties which fuse to a black magnetic glass. Dissolves in Borax and Microcosmic Salt, giving reactions for Iron. Not usually decomposed by acids, though in some cases Sulphuric Acid seems to attack it slightly. Staurolite is not known except in crystals. When in twinned cruciform crystals, this shape and its great hard- ness will distinguish Staurolite. In untwinned crystals their shape will distinguish it from Andalusite and Cyanite. Its hardness, the absence of pyramidal terminations in its crystals, and its solubility B.B. in Microcosmic Salt distinguish it from dark- coloured Quartz. Staurolite is generally brown or yellow in thin slices by ordinary light : it polarizes in brilliant colours and shows distinct but not very strong dichroism. Its twinned structure is very often useful in enabling us to recognise it. Topaz. 5Al,O 3 SiO 2 .Al 2 F 6 SiF 4 . Silicate of Alumina with the Oxygen partly replaced by Fluorine. G. 3 '4 3 '65. Trimetric. Angle of the unit prism 124 17'. Usually in prisms which are bounded by the faces of the unit prism and by pinacoids, and cannot be distinguished by the eye from regular hexagonal prisms. Basal face generally present. Very perfect basal cleavage. H. 8. Infusible ; unattacked or only slightly attacked by acids. If heated with Microcosmic Salt in the open tube will sometimes give off enough Hydrofluoric Acid to turn moist Brazil-wood paper yellow. Von K obeli says (Tafeln, Einleitung xvii.) that Topaz will not give off enough Hydrofluoric Acid to etch glass when heated with strong Sulphuric Acid, and I found this to be the case in the only instance in which I made a trial. He gives another method of determining the presence of Fluorine. Topaz scratches Quartz ; this and its highly perfect basal cleavage are sufficient to distinguish it when in crystals. If imperfectly crystal- lized, the hardness and fluorine reaction will suffice to identify it. In microscopic sections Topaz is limpid like Quartz, but the colours it gives with polarized light are far more vivid than those of that mineral, and are comparable in brilliancy with the colours of Olivine. Unless the section is cut accurately parallel to the base its marked basal cleavage will be visible and will distinguish it from Quartz. From Olivine, which it resembles in the brilliancy of its colours, 1 26 Geology. it is distinguished by the absence of the irregular cracks almost always present in that mineral, and also by the absence of altera- tion. Topaz is feebly dichroic, but its dichroism is inappreciable in thin sections. Beryl. 6SiO 2 .3BeO.Al 2 O 3 . Silicate of Beryllia (or Glucina) and Alumina. G. 2'63 2 -6 7. Hexagonal ; commonest form six-sided prisms, often with termina- tions showing faces of various hexagonal pyramids and the basal plane. Basal cleavage, but imperfect. H. 7-58. Brittle. Various tints of green, light blue, and yellow to white ; often a tendency to be blotchy. Streak white. B.B. just fusible on thin edges. Unaffected by acids. The only mineral with which Beryl is likely to be confounded is Topaz, and the perfect basal cleavage of Topaz alone distinguishes them. Again Topaz is absolutely infusible, and Beryl gives no Fluorine reaction. Beryl by the very inexperienced might be con- founded with Tourmaline. The crystals however of the two are very different in general shape and specially in their terminations. Beryl also is harder than Tourmaline and gives no reaction for Boric Acid. Beryl is far too hard to be mistaken for Apatite. Emeralds are green transparent Beryls, the colour probably being due to Chromium. Beryl has feeble double refraction and does not give brilliant colours with polarized light. In microscopic sections cleavages parallel to the base and the faces of the unit prism are indicated by irregular cracks along which the mineral is kaolinized and otherwise altered. (B 11) GARNET GROUP. The composition of the Garnets may be represented by the formula, 6RO. 3SiO 2 . 2R' 2 O 3 . 3SiO 2 = R 3 R' 2 Si 3 O 12 , where R stands for Ca, Mg, Fe, or Mn, and R' for Al, Fe, or Cr. G. 3-154-3. The Garnets are Monometric : the commonest forms are the Rhombic Dodecahedron and a combination of it and the Twenty-four- faced Trapezohedron, but other forms and combinations occur. Also twins, the twinning and composition plane being parallel to a face of the octahedron. Polysynthetic twinning after this type occurs, and as a result the faces of the crystals are traversed by numerous fine parallel striae. Cleavage dodecahedral, sometimes good, often scarcely recognisable. H. 6-57-5. Of various colours, but red tints very common. All are fusible, most varieties easily, but fusibility and blowpipe reactions vary according to composition. For a description of the various species reference must be made to works on Mineralogy. Garnets usually occur embedded in crystalline rocks, most plenti- fully in those of the Metamorphic class. The crystals are sometimes Rock-forming Minerals. 1 27 well defined, but not unfrequently their angles and edges are more or less rounded. The student will get to know by a little practice the general look of Garnets well enough to enable him pretty safely to refer to that group detached embedded crystals which have that appearance and the requisite hardness. A useful additional test is to chip off a splinter thin enough to be translucent and determine by viewing it with polarized light whether it is singly refracting. The determination of the exact species often requires more or less of analysis. Cases may arise in which analysis will be necessary before a mineral can be with certainty pronounced even to be a Garnet. In rock sections Garnet crystals are polygonal in outline, but the angles are frequently rounded off and the edges serrated. Cleavage is scarcely ever visible, but the crystals are traversed by numerous irregular cracks which are often lined by decomposition products. The surface is rough like ground glass, and the crystals stand up in bold relief. In the majority of cases the crystals are coloured by ordinary transmitted light, the colour varying with the species, but they are found colourless. Sometimes the crystals are free from enclosures, at others they are thickly crowded with crystals and frag- ments of other minerals. A dark Iron-lime Garnet, Melanite, is very common, associated with Leucite, Nepheline, Nosean, and Sanidine, in certain rocks. It has well-defined crystalline outlines, is translucent and brown in colour, and frequently shows concentric bands of various strengths of brown running parallel to the boundaries of the crystal. Since they crystallize in the Monometric system Garnets are singly refracting. To detect Garnets in a rock-slice, cross the Nicols and pick out those minerals which remain dark during a whole revolution of the stage ; then examine each with ordinary light. Hauyne and Nosean are two not uncommon minerals which behave in this way, but their very characteristic dusty enclosures distinguish them, and they have neither the colours nor the rough surface of Garnet. Spinels have sharp well-defined crystalline outlines usually more or less rectangular. Leucite has enough double refraction to distinguish it, even if its char- acteristically-arranged enclosures and its twinning are absent. (B 12) EPIDOTE GROUP. The general formula is H,0. 4CaO. 3( Al 2 :Fe,)0 3 . 6SiO,, or H 2 Ca 4 (Al 2 :Fe 2 ) 3 Si 6 26 . The Calcium is sometimes replaced to a small extent by Iron and Magnesium. There are two principal species whose composition is represented by the above formula. Iron Epidote or Pistazite. Proportion of Ferric Oxide above 6 '3 3 per cent. G. 3-25 3 '5. Monoclinic. Crystals usually long prisms, bounded laterally by orthopinacoids and orthodomes ; they are very frequently longer in the direction parallel to the orthodiagonal than in the direction perpendi- 128 Geology. cular to it, so that they look somewhat oblong in cross sections ; often in radiating groups. Cleaves parallel to the orthopinacoid ; imperfectly parallel to the base.* H. 67. Brittle. Colour various, but very frequently of the peculiar yellowish green known as pistachio green. Unaffected by acids, and decomposes only very slowly after being fused B.B. The two following experiments t however show that pro- longed ignition renders it soluble. After having been heated to incipient whiteness in a platinum crucible for thirty minutes and boiled for five minutes in dilute Hydrochloric Acid (1:2) the solution suddenly gela- tinized. An estimation of the silica showed that the mineral was completely decomposed. After similar heating for fifteen minutes and boiling for some time in dilute Hydrochloric Acid, silica separated in a bulky granular form. The mineral was only partially decomposed. Fusible. The peculiar colour and the prismatic habit of the crystals are the most useful characteristics for the recognition of Iron Epidote : its brittleness is also noticeable. Lime Epidote or Zoisite. Ferric Oxide not above 6*33 per cent. G. 3-113-38. Trimetric. Usually in long prismatic crystals that often have the faces of several rhombic prisms in addition to those of the unit prism, so that their cross section is many sided. Faces often longitudinally striated. Cleaves parallel to brachypinacoid. H. 6 6 -5. Colour usually grey to pale brown. Fusible. After prolonged heating at a high temperature gelatinizes with acids. There are besides Piedmontite, a Manganese Epidote in which the Ferric Oxide is largely replaced by Manganic Oxide, and Allanite, a cognate mineral containing Cerium. In thin sections Epidote often shows its characteristic green or yellowish-green colour by ordinary light ; but it is sometimes brown, and occasionally colourless. The crystals are often elongated parallel to the orthodiagonal, and in sections parallel to this axis they extin- guish parallel to the length of the crystals, or what is the same thing, the edges of the planes of easy cleavage. The crystals stand out in fair relief. With polarized light the colours are brilliant and limpid ; coloured sections usually show recognisable dichroism, but hardly so strongly as Hornblende ; in faintly-coloured sections dichroism is often scarcely perceptible. Epidote frequently occurs in fan-shaped aggregates of radiating crystals on the walls of fissures and cavities in rocks ; its peculiar colour and its dichroism will generally enable us to recognise it without much risk of error in such cases. It also occurs in larger irregular crystalline masses, and it is then not always distinguishable readily from Horn- * Many crystallographers read the crystals in a different way, and consider the plane of easy cleavage the base, and the plane of difficult cleavage the ortho- pinacoid. Practically there is only one good cleavage. t For which I am indebted to Mr. C. H. Bothamley, Assistant Lecturer in Chemistry at the Yorkshire College. Rock-forming Minerals. 129 blende cut in such a way as to show only one set of cleavage planes. If its dichroism is feeble, it may be confounded with Augite. The rocks in which Epidote is known under this form however are not many, and the mineral will generally be easily identified in them microscopically. There is also considerable resemblance occasionally between Epidote and Chlorite : the hexagonal form of the plates of Chlorite when cut parallel to the base, the absence of dichroism in such plates, and the difference in colour help to distinguish the two minerals. (B 13) ZEOLITE GROUP. The Zeolites are hydrated Silicates of Alumina, Lime, Potash, and Soda : some authors also include under this head certain silicates which contain no alumina, but otherwise resemble the aluminous Zeolites. They occur filling up cracks or hollow spaces or coating the walls of cavities in rocks, in which they have been deposited from mineral solu- tions that percolated through the rock ; or they occur as alteration products. It is often easy to say that a mineral is a Zeolite on the strength of its mode of occurrence and general look, but the species are in many- cases very difficult to distinguish, and can often be identified with certainty by nothing short of quantitative analysis. The following are a few of the more easily recognised species. Analcime. Generally in twenty-four-faced trapezohedrons, resem- bling Leucite in shape. But it fuses easily and gives water in the closed tube. Chabazite. Very often in rhombohedrons the angle of which is so nearly 90 that they can scarcely be distinguished from cubes. It is then very like Fluor-spar, and it also shows the interpenetrating twins so common in that mineral. But it has not the octahedral cleavage of Fluor and it will not give the reaction for Fluorine. It also gives water in the closed tube. ffatrolite.QitQii occurs in beautiful silky hair-like crystals spread- ing over the walls of cavities and shooting out into their interior. Its easy fusibility is characteristic, one of the hairs, when held in the flame of a candle, melts at the end into a globule. Stilbite or Desmine may sometimes be recognised by its sheaf-like aggregations of thin prismatic crystals, Heulandite by the tabular habit of its crystals. Indications of this character will sometimes enable us to make a likely guess at a Zeolite, but these are seldom sufficient for certain identification. For a description of the species the student must refer to works on Mineralogy. (B 14) SUNDRY SILICATES NOT COMING UNDER ANY OF THE PRE- CEDING GROUPS. Tourmaline, Schorl. Of very variable and complicated composition, which may however be represented by the formula or 130 Geology. where R stands for H, K, Na, Li, or F ; R' for Mg, Fe, Mn, or Ca ; R" for Al, Fe, or B. All contain Boric Acid. G. 2 -94 3 "3. Hexagonal (rhombohedral). The crystals are usually long prisms, and owing to the unequal development of the prismatic faces, or to every alternate face being suppressed by hemihedrism, their cross sec- tion is very often an equilateral triangle. Nine-faced prismatic com- binations are produced when the faces of the hexagonal prism of the second order are added to these three faces. The prismatic faces are often finely striated parallel to their length. The terminations are very commonly formed by three rhombohedral faces. The faces of opposite extremities are often unlike. Other more complicated combina- tions occur, and the terminations are often composed of numerous faces. No distinct cleavage. H. 7 7 '5. Of various colours, dark green and black the commonest. Fusibility varies much according to composition ; some varieties fuse easily, some with difficulty, some are infusible. Blowpipe reactions vary in the same way, but Boric Acid may always be recognised by Turner's test. Unaffected by acids. Under its commonest forms the peculiar crystalline habit and the hardness of Tourmaline alone enable one to recognise it with fair cer- tainty. It might by a hasty observer be confounded with Hornblende, but the absence of cleavage and its greater hardness distinguish it. The strong tendency of its prisms to be three-faced, and their dis- similar rhombohedral terminations are also totally different from any- thing seen in Hornblende. In the case of any uncertainty it should be tested for Boric Acid. The name Schorl is given to varieties in fine, needle-shaped, black crystals, which are often grouped together in radiating clusters. Radiated Schorl has occasionally somewhat the look of the similar form of Pyrolusite, but Pyrolusite has only a hardness of 2 to 2*5. In microscopic sections Tourmaline is not unfrequently blue by ordinary light, but it presents other colours. It polarizes very brilliantly and is strongly dichroic. Its dichroism will distinguish it from all minerals except Hornblende and the Magnesian Micas : it differs from these in the absence of any distinct and regular cleavage. Axinite. R 2 O.6R'O.3R" 2 3 .8SiO 2 where R,O is Potash or Water ; R'O is Magnesia, Lime, Ferrous Oxide, or Manganous Oxide ; R" 2 O 3 is Alumina or Boric Acid. Always contains Boric Acid. G. 3 '271. Triclinic. Crystals usually with sharp knife-like edges. H. 6 '5 7. Colour various, clove-brown very common. Fuses easily to a glass and gives the green of Boric Acid to the flame. Boric Acid may also be recognised by Turner's test. Unaffected by acids. A fusible mineral, in bladed crystals, with a hardness between 6 and 7, and which reacts for Boric Acid may be pronounced to be Axinite. Scapolite, Wernerite. (Ca:K 2 :Na 2 )O.Al 2 O 3 .2Si0 2 . Of most vari- able composition owing to variations in the ratio of the Protoxides, Sesquioxides, and Silica. As has been already mentioned (p. 105) Scapolite contains Chlorine. G. 2 '63 2 - 8. Rock-forming Minerals. 1 3 1 Dimetric. Prisms bounded by faces of the unit prism or by them and pinacoids and terminated by octahedral faces, are the commonest forms. Cleavage parallel to faces of unit prism and pinacoids, but inter- rupted ; the pinacoid cleavage the more perfect. Brittle. H. 56. Colours various, usually light. Fuses easily, with intumescence. The end of a crystal 2 mm. square in cross section can be fused into a bead. Only partly decomposed by acids, but to a variable amount according to composition. When crystallized the square form of its prisms and its indistinct rectangular cleavage are useful aids to its recognition. Crystalline, it might be mistaken for some Felspars ; but there is a characteristic fibrous appearance on its cleavage planes and a resinous lustre on its fracture surfaces which serve to distinguish it. Sheaf- like aggregates of prismatic crystals of Scapolite are not unlike the paler varieties of Zoisite : in anything like fairly well-crystallized specimens the difference between the crystals alone distinguishes the two minerals. Zoisite too has only one cleavage, and that remarkably perfect. Scapolite occurs most frequently in altered crystalline limestone, and occurrence in this rock often furnishes a hint that a mineral is Scapolite. Scapolite puts on very different appearances in different rocks. In some metamorphic limestones, that of Pouzac for instance, it yields crystals with well-defined rectangular or eight-sided outlines, sometimes showing also pointed terminations formed by octahedral faces ; it is clear and homogeneous and free from inclusions. In a Wernerite rock from Odegaard in Norway it occurs in aggregates of irregularly-shaped granules with a tendency to rectangular or eight-sided outlines crossed by one or two sets of cleavage cracks ; here again it is free from inclu- sions. In some schistose rocks, the Dipyrschiefer of Engoumer for instance, it occurs in crystals more or less rounded, but many of which show a tendency to an eight-sided contour, but it is crowded with inclusions, some of which are fragments of Quartz and other minerals identical with similar fragments which occur in the matrix, and others are small prismatic crystals possibly of a talcose character. Scapolite gives colours with polarized light comparable to those of Quartz. Sphene, Titanite. CaO.2Si02.CaO.2TiO,, or CaSiTiO 5 . Silicate and Titanate of Lime. G. 3 '3 3 '7. Titanite includes black and brown varieties in which the Calcium is partly replaced by Iron and a little Manganese. The varieties of light shades, yellow, greenish, are known as Sphene. There is also a rose-coloured or red variety, Greenovite, which owes its colour to Man- ganese. Powder of lighter coloured varieties white. Monoclinic. Crystals usually thin and platy with sharp wedge-like edges, but sometimes very complicated. Twins common in which the orthopinacoid is both twinning and composition plane : they have frequently the shape of a thin table with a deep narrow groove along 132 Geology. one face, the edges of the groove being sharp and knife-like. The groove is produced as in so many other cases of twinning, that of Felspar for instance (figs. 59, 60, 61), and it is deep and narrow on account of the sharp knife-like edge of the untwinned crystal. Cleaves sometimes parallel to the faces of the unit prism. H. 5 5 '5. Rather brittle. Resinous lustre. Most varieties are fusible B.B., but the fusibility varies : some are fusible only on thin edges. B.B. dissolves somewhat slowly in Microcosmic Salt. The varieties containing little Iron give the violet colour of Titanic Oxide in R.F. If much Iron is present the bead must be treated with tin on charcoal. Completely decomposed by boiling in strong Sulphuric Acid ; a quarter of an hour's boiling will usually be enough. An insoluble residue is left which contains Silica and Titanic Acid : the solution contains Sulphate of Lime and some Titanic Acid. If the Sulphuric Acid is not allowed to boil during decomposition, most of the Titanic Acid will be retained in solution. If then the solution be evaporated to dryness and the residue dissolved in Hydrochloric Acid and Zinc added, the solution turns a violet colour, which changes to rose-red when diluted with water, indicating the presence of Titanium. A better plan is to fuse the mineral with Bisulphate of Potash and triturate the fused mass with cold water. A white insoluble residue remains which is mainly Silica : the solution contains Sulphate of Lime and Titanic Acid. Filter, dilute the filtrate largely with water, add a few drops of Nitric Acid, and boil. Titanic Acid comes down as a white powder. If this be separated by filtration and dried, it may be tested in a Microcosmic Salt bead B.B. ; Ammonium Oxalate will show the presence of Lime in the filtrate. The peculiar wedge-shaped form of its crystals, its strong resinous lustre, and its hardness often enable us to identify Sphene without any further tests. We may confirm by detecting Titanium in a Micro- cosmic Salt bead B.B. If further confirmation is desired, fusion with Bisulphate of Potash enables us to detect Silica, Titanic Acid, and Lime. In rock sections the outlines of the crystals of Sphene have very commonly the shape of an elongated rhombus or a rhombus with its acute angles cut off lines perpendicular to the longer diameter. The cleavages parallel to the faces of the unit prism are sometimes repre- sented by two sets of somewhat thick irregular cracks. The surface has a rough, ground-glass-like look. The crystals are often bordered by a strong black edging, which is owing to the total internal reflection produced by the high refractive power of the mineral. The colours with ordinary light are yellow or reddish to dark brown ; with polarized light there is no brilliant colouring but only variations in tint as the analyzer is rotated. Rhomboidal sections generally extin- guish when the longer diagonal of the rhombus is parallel to a diagonal of the Nicols. Sphene shows sensible but not strong dichroism. In- clusions are only exceptionally present. Where Sphene shows a marked acute lozenge-shaped outline, it can be identified without difficulty. It occurs sometimes in larger masses Rock-forming Minerals. 1 3 3 without definite shape, and then its rough surface and the irregular cracks representing the cleavage give it some resemblance to divine. But it never polarizes with the brilliant colours of Olivine ; Olivine is not sensibly dichroic in thin sections ; Olivine, if it shows cleavage at all, shows only one set of cleavages, while in Sphene two are often visible. Zircon, Hyacinth', Jargon. ZrO,Si(X Silicate of Zirconium. Contains sometimes Ferrous and Manganous Oxides and Magnesia. G. 4-05475. Dimetric. Square prisms terminated by octahedrons, eight-sided prisms bounded by faces of unit prism and pinacoids and similarly terminated, and more complicated forms. Cleavage very indistinct, fracture conchoidal. H. 7 '5. Yellow, brown, and reddish brown. Infusible. Unaffected by Hydrochloric and Nitric Acid, attacked by strong Sulphuric Acid. The less complicated crystals are readily seen to be square prisms, or modified square prisms, and their shape together with the other physical properties and general look of the mineral usually suffices to identify it.* In thin sections the usual colour of Zircon by ordinary light is light brown. It has strong double refraction and gives with polarized light the most brilliant colours accompanied by a certain iridescence ; the colouration is so vivid as to catch the eye even when the mineral occurs only in very small crystals. The crystals stand out in bold relief and are surrounded by a black margin. The dichroism varies very much in different specimens, it is sometimes strong, sometimes scarcely perceptible. (B 15) HYDROUS SILICATES OF ALUMINA. Kaolin, China Clay. Al 2 3 -2SiO 3 .2H 2 O. G. 2 -4 2 -63. Occurs usually in a clayey or mealy state, but can be sometimes observed under the microscope to contain minute six-sided plates, which are trimetric crystals. H. 1 2'5. Rather unctuous. Plastic. B.B. infusible ; gives a blue colour with Nitrate of Cobalt ; yields water in the closed tube. Kaolin has been produced by the alteration of silicates, specially of Orthoclase. It forms an important constituent of many clayey rocks, and is the material out of which the best porcelain is made.t We also meet with clayey substances composed of very finely com- minuted but undecom posed or only partially decomposed Orthoclase. These may conveniently be distinguished from Kaolin by calling them Felspathic mud. If a little of this mud be made into a paste with water, pressed into the shape of a chip, and dried, it may be held in the forceps in the blowpipe flame, and it will be found that the edges * For a method of detecting Zircon when finely disseminated in small quantity in rocks see Exploration of the Fortieth Parallel, ii. 397. t For further particulars see American Journal of Science and Arts, [2] xliii. (1867) 351 ; Percy's Metallurgy, vol. on Fuel, p. 92. 1 34 Geology. can be fused. Pure Kaolin treated in the same way is absolutely infusible. Jtock-soap, Bergseife, are sometimes clayey substances allied to Kaolin, sometimes soft magnesian silicates. Lithomarge, Steinmar/c, is a term often used somewhat loosely. The typical form is an indurated Kaolin. H. 2 2 -5. It is compact, yields to the finger-nail affording a shining streak, feels greasy, and adheres strongly to the tongue. It is often coloured in a mottled way by Ferric Hydrates. Infusible. JBole, Eisensteinmark, is a similar substance, containing a large amount of Iron, which renders it fusible. It falls to pieces in water with a crackling noise. Bole may have arisen from the decomposition of ferruginous silicates like Hornblende and Augite, while Kaolin has come more largely from alkaline silicates. Fuller's Earth, Walkererde, Walkthon. A hydrated silicate of alumina, soft, yielding a shiny streak to the nail, and adhering only slightly to the tongue. When placed in water it falls to pieces with a hissing or puffing sound, and forms a pulpy mass which is not in the least plastic. It absorbs grease largely, and is employed in con- sequence for fulling cloth. Allopkane. ALA.SiO 2 .5H,0. G. 1 '851 -89. Occurs in stalagmitic crusts on the walls of cracks in chalk, clays, and other rocks. H. 3. Very brittle. Glassy lustre externally, waxy inside. Often pale sky-blue, also yellow, green, brown, to white. B.B. crumbles, but is infusible : yields much water : white varieties turn blue with Nitrate of Cobalt. Is decomposed by dilute Hydro- chloric Acid ; an hour's treatment at a temperature below the boiling- point, though it did not thoroughly decompose the mineral, produced a considerable amount of decomposition. Green Earth, Grilnererd. A hydrous silicate of Alumina and Ferric Oxide, with small quantities of Ferrous Oxide, Lime, Magnesia, Soda, and Potash. It is a green earthy substance and has been largely produced by the decomposition of Augite. Glauconite is a substance very similar to Green Earth both in appear- ance and composition. It is now in process of formation at various spots on the ocean-bed, and fills up in some cases the empty shells of Foraminifera. Casts of such shells in Glauconite are also found in rocks, known as " Greensands." According to Professor Heddle some Glauconite is an original rock constituent and should be placed in his Chlorite group (see p. 123). C. CARBONATES OF LIME, MAGNESIA, AND IRON. The general composition of the minerals to be placed in this group may be represented by the formula {Ca:Mg:Fe:Mn:Zn}CO a . Rock-forming Minerals. 135 Lime, Magnesia, and Ferrous Oxide are the commonest bases, the oxides of Manganese and Zinc entering only occasionally and in small quantity. The bases replace one another isomorphously, so that some of the species vary much in composition and tend to shade off into one another. All the members of the group with the exception of Aragonite are rhombohedral in crystallization, and the differences between the angles of the Unit Rhombohedron in the various species is very small (p. 80). All except Aragonite have good rhombohedral cleavage. All dissolve with effervescence in acids. Calcite, Gale-spar, Kalkspath. CaC0 3 . Calcium Carbonate or Car- bonate of Lime. G. 2-72. H. 3. Hexagonal. Several hundred different forms and combinations have been observed. Most perfect cleavage parallel to the faces of a Rhombohedron (see p. 33). This Rhombohedron is taken for the Unit or Standard form : more than forty other Rhombohedrons occur. Scaleuohedrons are also common, more than eighty being known. We also find Hexagonal Prisms, and less often Hexagonal Pyramids. The Unit Rhombohedron (R, fig. 66) is not very common : a Rhom- bohedron, the ratio between the vertical and lateral axes of which is one-half the corresponding ratio in the Unit Rhombohedron. and which is therefore denoted by J R (fig. 67), is very common in combinations : J R (fig. 68) very often occurs, it is so very flat that it is difficult on looking at it to realize that it is a Rhombohedron and not a thin lamina. Cardboard models of these Rhombohedrons are easily constructed by means of the net in Mr. Jordan's book. If they are made of the dimensions given below, the lateral axes will be the same for all three and the relation between the vertical axes will be evident. R. iR. JR. Length of edge 6'4 5 9 5-8 Angle of rhombic face . 78 654 6H The very commonest form of all is a Scalenohedron, the ratio between the vertical and lateral axes of which is three times the corresponding 136 Geology. ratio in the Unit Rhombohedron, and which is therefore denoted by R 3 . It is called Dogtooth Spar and is shown in fig. 69. Hexagonal Prisms terminated by basal planes (fig. 70) are not rare. c P' ccP ccP Fig. 69. Fig. 70. The following are the commonest combinations. Hexagonal Prisms terminated by the faces of - J R : if the prismatic faces be very short, as in fig. 71, we have the common shape known as Nailhead Spar. Hexagonal Prisms terminated by the faces of \ R (fig. 72) are also met with. Very elongated Scalenohedrons ter- minated by Rhombohedral faces are also very common, as in fig. 73, where jffi 6 occurs combined with - | R. of twinning, the commonest of which is Fig. 71. There are several type, shown in fig. 74. The plane of twinning and composition is parallel Fig. 72. Fig. 73. to a face of - \ It, or what comes to the same thing, it is parallel to a plane such as DFD'F' in fig. 66 passing through the parallel diagonals of two opposite faces of R. The twinning is often repeated many times over, producing a succession of grooves and ridges on the faces of the crystal. If the twin laminae are very thin, these grooves become reduced to striations. In fig. 74 four twin laminae occur : one of the composition planes is shaded, the others are omitted to avoid con- fusion, but the student will have no difficulty in introducing them. The letters have the same meaning as in fig. 66, though the crystals are shown in different positions in the two figures. Fig. 75 is a net from which a cardboard model of a crystal once twinned may be con- structed. A BCD, AGEF are equal and similar rhombuses, the THE TJ HI VISIT'S Rock-forming Minerals. larger angle of each of which is 102: FGH, AEB are of an equal and similar rhombus divided by its longe? EB or FH : BEKL is a rectangle, EK being equal to EA. This figure must be drawn carefully on a piece of cardboard and transferred to another piece by pricking through the angles with a fine needle : Fig. 74. Fig. 75. from this net one half of the twinned crystal will be constructed. Now turn over the pattern, laying that side undermost which was before uppermost, and again transfer the net to a piece of cardboard : the model thus constructed will be the other half. The two placed together make up a model of R : if one hal 'be rotated through two right angles, keeping the twinning faces BEKL in contact, the twin will be obtained, with a groove on one face and a ridge on the opposite face. Very strong double refraction, easily observed on the transparent varieties called Iceland Spar. B.B. infusible, but is reduced to a white mass of quicklime which glows with an intense white light. Moistened with Hydrochloric Acid gives the Calcium colour to the flame. With soda on charcoal fuses, then the soda is absorbed by the coal, and a glowing mass of quicklime left. A lump touched with a drop of cold dilute Hydrochloric Acid effervesces briskly. Fragments or powder dissolve readily even in cold acid, and a platinum wire moistened with the solution gives the Calcium colour to the Bunsen flame. Aragonite. CaCO. ? . A second form of crystallized Calcium Car- bonate. G. 2-931. Trimetric. In simple but more frequently in complex crystals with many-faced terminations. Twinning is very common in Aragonite. In one case which frequently occurs the cross section of the twinned crystal is cruciform in shape. But the most remarkable effect is pro- duced by repeated twinning ; three individuals are so united as to form a prism the cross section of which is apparently a regular hexagon. Cleaves parallel to the brachypinacoid. H. 3 '5 4. B.B., or in a closed tube, falls instantly to powder. This arises from the fact that Aragonite is a very unstable form, and the heat converts it into Calcite ; each of the minute fragments is a rhombohedron of that mineral. 138 Geology. Aragonite is harder than Calcite, but this test cannot be applied to the imperfectly crystallized varieties. The absence of the marked cleavage of Calcite is a good distinguishing test. But nothing is so characteristic as the way in which it falls into powder before the blow- pipe. Aragonite often occurs in delicate branching and interlacing fibrous forms which look something like coral. When this form is associated with Iron Ores it has often a rusty colour and it is then called Flos Ferri. Siderite, Chalybite, Spathic Iron Ore, Eisenspath, Spatheisenstein, Sphcerosiderite. FeCO 3 . Ferrous Carbonate. G. 3*7 3 - 9. Faces of crystals very often much curved. Very fiat knife-edged crystals common. H. 3 '5 4*5. B.B. blackens, becomes magnetic, and fuses, but sometimes only with difficulty. Scarcely acted on in lump, and effervesces only feebly in powder with cold acid. Dissolves with brisk effervescence in warm acid. Oxidizes and turns rusty outside by exposure to the air. Siderite containing much Manganese turns deep black by exposure. Clay Ironstone is a mixture of clay and Ferrous Carbonate. It occurs in nodules or thin bands in clayey or shaly rocks. When it contains a large proportion of dark-coloured organic matter it is called Blackband. Magnesite, Talkspath. MgCO 3 . Magnesium Carbonate. G. 3 3-08. H. 3-54-5. The white varieties give the flesh-colour characteristic of Magnesium with Nitrate of Cobalt. B.B. infusible. Only slightly acted on by cold acid. Bitter Spar, Dolomite,* Bitter spatli, Pearl Spar, Rautenspath (Rhombspar). (Ca:Mg)C0 3 . Double Calcium and Magnesium Car- bonate. The proportions of the two Carbonates vary very much ; Rammelsberg instances the following. (1) CaCO,.MgCO 3 . (2) 3CaC0 3 .2MgC0 3 . (3) 2CaC0 3 .MgCO 3 . (4) CaC0 3 .3MgCO,. G. 2-82-9. H. 3-54. Faces of crystals often curved. B.B. infusible. Only slightly acted on by cold acid. Ankerite, Brown Spar. (Ca:Mg:Fe)C0 3 . Lime, Magnesia, and Iron Carbonate. G. 2 -95 3'1. Often contains Manganese. H. 3 -5 4. B.B. infusible ; darkens and becomes magnetic on heating. Usually only slightly acted on by cold acid. Turns rusty outside by exposure to the air. Mesitite. 2MgC0 3 .FeCO,. G. 3'33 3-36. H. 44-5. Pistomesitite. MgC0 3 .FeC0 3 . G. 3'41. H. 3'5 4. Carbonates of Iron and Magnesia; cannot be distinguished from one another except by quantitative analysis. B. B. infusible ; darken and become magnetic. Only slightly acted on by cold acid. * Dolomite is often used as a name for this mineral. It would be better to restrict the term to the rock composed largely of Bitter Spar. Rock-forming Minerals. 1 39 Of the rhombohedral carbonates described in tins section Calcite is by far the most important, because it is the main constituent of the very common rock limestone. Calcite when crystallized can scarcely be mistaken for any other mineral ; the only minerals with which it can possibly be confounded are those placed with it in this section. From these it may be thus distinguished. First there is its pre-eminently perfect cleavage and the brilliant lustre of its cleavage planes : its fellow-carbonates do cleave parallel to the faces of their Unit Bhornbohedron, but seldom in so striking a manner as Calcite. Secondly comes the fact that the faces of the crystals of Calcite are very rarely indeed, if ever, curved : curvature of the faces is the rule in the case of all the other minerals of this group with perhaps the exception of Magnesite. Thirdly, Cal- cite effervesces briskly even in lump with cold acid ; the other rhombohedral carbonates are scarcely affected in lump, and usually effervesce only feebly in powder, with cold acid, though they effervesce briskly when the acid is warmed. A mineral then soft enough to be easily scratched with the knife, with conspicuous rhombohedral cleavage, which effervesces briskly when a drop of cold acid is placed on it, is almost certainly Calcite. In the same way a rock which can be scratched with the knife and which effervesces when it is touched with cold acid, contains a good deal of Calcite. A mineral on the other hand with rhombohedral cleavage and crystallizing in rhombohedral forms, the faces of . whose crystals are curved, and which effervesces only slightly with cold acid, is almost certainly not Calcite. Which of the rhombohedral carbonates it is, is not always easy to determine. If the mineral becomes strongly mag- netic on heating it is probably Siderite, or may be Ankerite, or Mesitite. If it gives a flesh-colour with Nitrate of Cobalt, it indicates Magnesite. Frequently however nothing short of analysis will decide the species with certainty, but fortunately a very easy qualitative analysis suffices for approximate determination. The powdered mineral is dissolved in dilute Hydrochloric Acid and any insoluble matter separated by nitra- tion. The filtrate is warmed with a few drops of strong Nitric Acid, and Ammonia added till it becomes alkaline. This will precipitate the Iron and any Alumina that may be present. Filter and add Ammonium Chloride and Ammonium Carbonate ; this brings down the Lime as Calcium Carbonate and any Barium or Strontium. The two latter if in any quantity may be detected in the precipitate by taking up a little on a platinum wire, moistening it with Hydrochloric Acid, arid holding it in the Bunsen flame. The Calcium colour first flashes through the flame; very soon the Barium produces a steady yellowish-green colour ; Strontium must be separated by the spectro- scope. After filtering off the precipitate by Ammonium Carbonate add Sodium Phosphate to the filtrate. A white precipitate indicates Magnesia. If the Magnesia is small in amount, some hours may be necessary to produce precipitation. If the precipitate of Ferric Oxide be very large and the precipitates by Ammonium Carbonate and Sodium Phosphate very small, the mineral is Siderite. If there is a 14 Geology. large precipitate by all three reagents the mineral is Ankerite. If there is scarcely any precipitate of Calcium Carbonate, but the two other precipitates are considerable, the mineral is either .Mesitite or Pistomesitite ; which of the two, quantitative analysis alone can decide. If Sodium Phosphate alone gives a large precipitate, the mineral is Magnesite. Manganese is best detected by fusing with Carbonate of Soda and Nitre, when it gives a green mass of Sodium Manganate. In microscopic sections the cleavage of Calcite is usually very distinct, the edges of the cleavage planes being shown by numerous fine and very regular lines : if the section be cut parallel to the base of the hexagonal prisrn all three cleavages are visible making angles of 120 with one another. The mineral also shows what can hardly be called dichroism for it is colourless, but it possesses that property of absorbing unequally rays whose planes of vibration differ in position which produces dichroism in coloured minerals. If the analyzer be removed and the polarizer rotated, the illumination changes sensibly, and when the mineral is twinned, the twin lamellae extinguish in succession as the polarizer goes round, and occasionally show faint colours for certain positions of the polarizer which disappear when that position is changed. This result is more clearly shown if yellow light is used, when the depth of tint shows a marked variation. Twinning, especially that parallel to a face of - J- R, is common. Some crystalline limestones are made up of irregularly-outlined inter- locking grains of Calcite, each one of which is twinned according to this law, but the directions of the twinning planes are different in the different grains. Aragonite differs from Calcite in showing only one set of cleavages, and these are far less sharply marked and regular than the cleavages of Calcite. D. COMPOUNDS OF LIME OTHER THAN THOSE OF THE LAST SECTION. Gypsum, Gyps, Selenite. CaSO 4 .2H.,O. Hydrated Calcium Sul- phate. G. 2-312-33. Monoclinic with very perfect cleavage parallel to the clinopinacoid, which causes it to split easily into very thin laminae which are slightly flexible, but not elastic. There are also less perfect cleavages parallel to the faces and the base of the unit prism ; these are often indicated by fissures crossing the laminae, or they appear in the shape of inter- rupted cracks when the laminae are bent. The term Selenite is confined to the crystallized varieties. They occur in broad platy masses, in complicated crystals, in arrow-head- shaped twin crystals, and in star-shaped clusters. The finely-grained cryptocrystalline varieties are called Gypsum. When very finely grained and mottled by coloured impurities, so as to be available for ornamental purposes, the mineral is called Alabaster. Intermediate between the largely crystalline and the cryptocrystalline forms are fibrous varieties, which when the fibres have a silky lustre are called tiatin Spar. Rock-forming Minerals. 141 H. 1'5 2, easily scratched by the finger-nail. B.B. fuses easily and gives the Calcium colour to the flame. Yields water in the closed tube. Very frequently readily and largely soluble in dilute Hydrochloric Acid : some varieties however dissolve to such a small amount that they are practically insoluble. Selenite can seldom be mistaken. Its foliation is most pronounced, and the laminae are neither elastic like those of Mica, nor greasy and difficult of fusion like those of Talc. The mere look of Gypsum taken in conjunction with its softness usually enables us to recognise it with certainty. But both Selenite and Gypsum are occasionally met with under forms which it is not so easy to be sure about. In such cases if the minerals will dissolve in acid, their recognition is still easy. The solution will show the Calcium colour when a platinum wire moistened with it is held in the Bunsen flame ; if Barium Chloride be added, a white precipitate is formed which is insoluble in Hydrochloric Acid, indicating the presence of Sulphuric Acid. Lastly water is given off in the closed tube. If the mineral refuses to dissolve in acid, we may proceed thus. Sulphate of Lime does not very readily show the Calcium colour in the Bunsen flame, but with care it may be made to afford this reaction. The secret consists in getting the right quantity, neither too much nor too little, on the wire. Bend a fine platinum wire into a ring 1 mm. in diameter, moisten it in dilute Hydrochloric Acid and touch it lightly on to the powdered mineral so as to take up a small quantity on the loop, and hold it in the Bunsen flame 5 or 6 mms. from the base. If the colour does not show itself, moisten again, and again introduce the wire into the flame, and repeat the process if necessary some dozen times. Each moistening washes off a little powder and reduces a portion of what is left to Calcium Chloride ; the right stage is at last arrived at, and flashes of the Calcium colour shoot through the flame. The experi- ment often fails of success till the amount of powder left on the wire is so small that the eye can scarcely detect it. Again the mineral will give the Sulphur reaction with Carbonate of Soda. As there is no natural Sulphide or Sulphite of Lime known, these two reactions prove the presence of Sulphate of Lime, and if water is obtained in the closed tube, the mineral is Gypsum. Anhydrite. CaSO 4 . Anhydrous Calcium Sulphate. G. 2 -9 2 '98. Trimetric. Cleaves parallel to the base, the brachypinacoid, and the macropinacoid of the unit prism. The three cleavages are thus at right angles to one another; in well-crystallized specimens this is a very characteristic property. H. 3 3 '5. B.B. fusible. Frequently readily soluble in dilute Hydrochloric Acid, but some varieties are practically insoluble. Cryptocrystalline Anhydrite is often very like Gypsum to look at, but it is harder and it gives no water in the closed tube. Crystallized Anhydrite cannot possibly be mistaken for Selenite. Anhydrite does not split into thin laminae, and Selenite does not cleave in three directions at right angles to one another. In doubtful cases we must proceed as directed under Gypsum. The 142 Geology. reactions both for the soluble and insoluble modifications will be the same as for that mineral, only there will be no water in the closed tube. Fluor-spar, Flusspath.CaiF. 2 . Calcium Fluoride. G. 3 '01 3 -25. Monometric. Common in cubes. Perfect octahedral cleavage (see p. 35). H. 4. Brittle. B.B. fusible. If the powdered mineral be placed in a small platinum crucible, or in a depression made in a bit of sheet lead, moistened with Sulphuric Acid, and gently heated, Hydrofluoric Acid is given off. This turns Brazil-wood paper yellow, or if the crucible be covered by a bit of glass, it will corrode the glass. The glass must be washed and dried, and it will then be seen to be rough like ground glass. The solution shows the Calcium colour in the Bunsen flame. Crystallized Fluor-spar is easily recognised by its form, which is much oftener a cube than anything else, and which it is generally easy to see is monometric, and by its octahedral cleavage. In crypto- crystalline or obscurely-crystallized specimens, the tests described render its determination a very simple matter. Blue John is a beautiful variety which occurs in concretions at Castleton in Derbyshire. The nodules are radiated and made up of concentric coats of various shades of blue. Apatite. 3Ca 3 P. 2 O 8 .Ca(Cl:F) 2 . Calcium Phosphate with Calcium Chloride and Fluoride ; the proportion of Chlorine to Fluorine is variable ; the varieties containing little or no Fluorine are called Chlor- apatite, and those with little or no Chlorine Fluorapatite. G. 2 '92 3-25. Hexagonal. In hexagonal prisms of the first order, or in dodecahedral prisms bounded by faces of the hexagonal prisms of the first or second order, terminated by pyramidal faces, and often capped by the basal plane. Very imperfect cleavages parallel to the base and faces of the unit prism. H.5. Brittle. B.B. fusible with some difficulty on the edges. Treated as directed under Gypsum the powder will show the Calcium colour in the Bunsen flame. Moistened with Sulphuric Acid it shows the bluish green of Phosphoric Acid, faintly in the Bunsen flame, distinctly but momen- tarily in the blowpipe flame. Easily soluble in dilute Hydrochloric and Nitric Acids ; a moderately large quantity of Ammonium Molybdate added to a small quantity of the solution, and warmed, gives a yellow precipitate, indicative of Phosphoric Acid. Fused with Microcosmic Salt in an open tube the varieties containing Fluorine give off enough Hydrofluoric Acid to turn Brazil-wood paper yellow, if not to corrode the glass. The same result may be obtained by gently heating the powdered mineral with a small quantity of strong Sulphuric Acid in a small test tube ; the acid must not boil. The nitric acid solution of the varieties containing Chlorine gives with Nitrate of Silver a white curdy precipitate which turns black or dark grey by exposure to the light. When in crystals Apatite may be often identified by its shape. Its Rock-forming Minerals. 143 inferior hardness prevents its being mistaken for Beryl, which assumes very similar shapes and has often the same general appearance. Pyromorphite, with which it might possibly be confounded, is much heavier and softer, and gives lead with soda B.B. Hexagonal crystals of Calcite are softer, differently terminated, and effervesce with acids. The compact or cryptocrystalline forms of Apatite are called Phos- phorite, and are valuable for the manufacture of artificial manures. The tests given above enable us to recognise them. As a constituent of rocks Apatite generally occurs in long slender hexagonal prisms terminated by basal planes. The boundaries of the crystals are always sharply denned, and the crystals, though they are often abundant in some parts of a section and nearly absent from others, usually occur isolated from one another. In longitudinal sections the basal cleavage is generally shown by irregular cracks perpendicular to the length of the crystal. The colours with polarized light are feeble, and there is no dichroism in thin sections. Longitudinal sections generally extinguish when their axis is parallel to a diagonal of either polarizer or analyzer. Apatite scarcely ever shows any alteration ; even where the surrounding minerals are greatly altered it remains unchanged. Sections inclined at a large angle to the axis are hexagonal in outline ; they differ from the corresponding sections of Quartz in the sharp definition of their contours, and they do not occur in overlapping aggregates nor are their angles rounded as is the case with Tridyrnite. It is not always easy to distinguish between the hexagonal sections of Apatite and ]N T epheline, but crystals of Nepheline have a tendency to be more numerous and more closely crowded together than those of Apatite. When the hexagonal sections are oblique enough to the axis to allow of portions of the prismatic faces being seen, the latter are crossed by the planes of basal cleavage, and the presence of these dis- tinguishes Apatite from Quartz, Tridymite, or Nepheline. Longitudinal sections of Apatite are needle-shaped : those of Nepheline are usually about as long as broad. It is often desirable to confirm by wet tests the indications of Apatite given by the microscope. A tolerably large sample of the pounded rock is treated with Hydrochloric Acid, any insoluble matter is separated by filtration, the filtrate is evaporated to dryness, and the dried residue is moistened with strong Hydrochloric Acid which renders the silica insoluble; Nitric Acid is then added, the mixture is warmed and the silica filtered off. In a portion of the filtrate Phosphoric Acid is detected by Ammonium Molybdate. The other portion is neutralized by Ammonia, and then tested for Lime with Ammonium Oxalate. E. COMPOUNDS OF BARIUM. Barium gives a yellowish-green colour to the Bunsen flame. When a small quantity of the powder of a mineral containing Barium is taken up on a small ring at the end of a fine platinum wire, moistened with Hydrochloric Acid, and held in the Bunsen flame, the Calcium colour, if any Calcium is present, appears first, shooting in flashes 144 Geology. through the fjame. Then usually follows a strong Sodium colouration. As the Sodium burns off, the Barium colour comes into view and soon produces a broad sheet of green if Barium is present even to a moderate amount. Xo substance, except Sodium, gives so strong a colouration, and one so easy to obtain. Indeed it is no easy matter to get a wire clean that has been once used with a Barium compound. Such a wire must never be trusted till after having been repeatedly dipped in acid and held in the flame it ceases to show any Barium colour. Heavy Spar, Schiver^ath, Barytes, Cawk. BaS0 4 . Barium Sul- phate. G. 4-34-72. Trimetric. The crystals tend very much to a tabular form bounded above and below by basal planes, the lateral planes being those of the unit prism, domes, or octahedral faces, as in fig. 49. Fig. 76 shows another common shape. Clusters of flat knife-edged crystals common. Cleaves Fig. 7(3. parallel to the base and lateral faces of the unit prism, the basal cleavage being the more perfect. The cleavability varies, sometimes it is very perfect. H. 2-53-5. B.B. decrepitates, if powdered arid made into a paste fuses easily and colours flame green. With similar precautions to those used for Gypsum, gives good colouration in the Bunsen flame. On charcoal with Carbonate of Soda fuses, spreads out, and sinks into the coal. The fused mass gives the sulphur reaction on silver. Insoluble in acids. Its high specific gravity, which is quite evident when even a moderate- sized lump is poised in the hand, is one of the first characters that leads us to suspect a mineral to be Heavy Spar. This, and its shape when in crystals, are often enough to make us sure of it. In crystalline masses the cleavage in three directions is a useful character ; this cleavage cannot be mistaken for that of Calcite, though it is sometimes almost as perfect, because two of the planes of cleavage in Heavy Spar are at right angles to the third. The colouration confirms our identification. The presence of Sulphur is ascertained by fusion with Carbonate of Soda, and as no natural Sulphide or Sulphite of Barium is known, a mineral which has the physical characters mentioned, and contains Barium and Sulphur, must be Heavy Spar. The term Cawk is applied to white, massive, or cryptocrystalline Barium Sulphate, which is ground up and used for the adulteration of whitelead. Wit/write. BaC0 3 . Barium Carbonate. G. 4 -29 4 -35. Trimetric. Crystals are twins, like those of Aragonite, with pseudo- hexagonal shapes, such as double hexagonal pyramids, or combinations of hexagonal prisms with hexagonal pyramids. Occurs more frequently in crystalline masses, which often take globular or botryoidal shapes. Cleaves parallel to the faces of the unit prism. H. 3 3*75. Brittle. B.B. fuses easily, colouring flame green. Effervesces feebly with cold acid, dissolves with effervescence in dilute Hydrochloric Acid when warmed. Rock-forming Minerals. 145 The high specific gravity and green colouration indicate Barium ; the crystalline shape distinguishes it from Heavy Spar when it is in crystals. But the easiest test is to dissolve a little in Hydrochloric Acid, and test for Barium in the Bunsen flame. Barytocalcite. (Ba:Ca)C0 3 . Barium and Calcium Carbonate. G. 3-643-66. Monoclinic. Cleaves parallel to faces of the unit prism. H. 4. B.B. fuses on thin edges, and colours the flame with the Barium colour. Fused with Carbonate of Soda on charcoal, the Baryta sinks into the coal, but the Lime remains and glows. The easiest way to recognise this mineral is to dissolve it in dilute Hydrochloric Acid and test in the Bunsen flame. Flashes of Calcium colour show first ; afterwards a steady sheet of Barium colour. Alstonite or Bromlite has the same composition, but is trimetric. F. COMPOUNDS OF STRONTIUM. Strontium produces an intense bright carmine colour in the Bunsen. flame. The colour given by Lithium is somewhat similar but of a more purple hue. Moistening with Hydrochloric Acid intensifies the colour. For the detection of Strontium by flame colouration in the presence of Barium reference must be made to works on the blowpipe. Celestine, ColestinSrSO^ Strontium Sulphate. G. 3'92 3'97. Trimetric. In tabular or columnar crystals. Very perfect cleavage parallel to the base of the unit prism ; cleaves also parallel to its lateral faces/ Very brittle. H. 3 3 -5. Often has a very beautiful faint bluish opalescent colour. Streak white. B.B. often decrepitates; in powder fuses easily, colouring the flame carmine. Gives the sulphur reaction with Carbonate of Soda. In- soluble in acids. The character which first of all leads us to suspect a mineral to be Celestine is its excessive brittleness. Its crystals are often thin plates bounded by basal planes above and below, and laterally by domes ; columnar crystals in which the long faces are pinacoids or domes or both are also common. The basal cleavage often gives the mineral a platy structure. The pale-blue opalescent colour, when it is present, is very suggestive of Celestine. A mineral suspected to be Celestine should be powdered and a little of the powder taken up on a spiral of thin platinum wire moistened with Hydrochloric Acid and tested in the Bunsen flame in the same way as Gypsum. The Strontium colour will then show itself. Test for Sulphur. If both Strontium and Sulphur are present, and the mineral has the proper physical characters, it is Celestine. Strontianite. SrC0 3 . Strontium Carbonate. G. 3 "6 37. Trimetric. Crystals often long and thin and with more or less of a radiated arrangement. Cleaves parallel to faces of unit prism. H. 3-54. Brittle. Of various colours ; streak white. K 146 Geology. B.B. throws out little branching sprouts so that it looks like a bit of moss when taken out of the flame, and fuses only on thin edges : gives very strong Strontium colour to flame, which is intensified by moisten- ing with Hydrochloric Acid. The blowpipe reactions would alone identify Strontianite. It also dissolves with effervescence in cold dilute Hydrochloric Acid and the solution gives a Strontium colour in the Bunsen flame. G. COMPOUNDS OF IRON. Minerals containing a large amount of Iron, if not magnetic, become magnetic in most cases after heating B.B. Treated with Hydrochloric Acid many of them are more or less decomposed. If the solution contains only Ferrous Chloride, it is colourless ; usually it is yellow from the presence of some Ferric Chloride. Potassium Ferrocyanide precipitates from this Prussian Blue. If a few drops of strong Nitric Acid be added to the solution and it be warmed, the colour darkens, owing to the complete change of Ferrous Chloride into Ferric Chloride. Ammonia now produces a rusty-coloured precipitate of Ferric Hydrate. In a Borax bead Ferric Oxide produces the following colours. O.F. With a little, yellow when hot, colourless when cold. With larger quantity, red when hot, yellow when cold. E.F. Bottle green. The bottle green in R.F. is the most characteristic reaction. Manganese and Iron are elements which are constantly associated. Very few ores of Iron are free from Manganese, and some contain it in considerable quantity. Manganese is best detected by fusion with Carbonate of Soda and Nitre. It may also be sometimes recognised thus. A Borax bead, containing only a small quantity of the assay, treated in O.F., is yellow froni Iron when hot ; as it cools the Iron colour fades, and if Manganese is present to any amount the violet of that element makes its appearance. (G 1) OXIDES OF IRON. There are three Oxides of Iron. Percentage of Metallic Iron. Monoxide of Iron or Ferrous Oxide, FeO . . 7 7 '7 Sesquioxide or Peroxide of Iron or Ferric Oxide, FeA - - ... . . 70-0 Magnetic Oxide of Iron or Ferrosoferric Oxide, Fe 3 O 4 . 72-4 The first is an unstable compound, and whenever it is produced is converted into a higher oxide, a carbonate, or some other compound. The other two occur as minerals. Magnetite, Magnetic Iron Ore, Mac/neteisenstein, Magneteisenerz, Fer Oxydule.FQ 3 O 4 . Ferrosoferric Oxide. G. 4 '9 5 "2. Monoinetric. By far the commonest form is the regular octahedron ; rhombic dodecahedrons not rare. Rock-forming Minerals. 147 Brittle. H. 5 -5 6-5. Strongly magnetic and sometimes polar. Iron black. Streak black. B.B. almost infusible. Decomposed very slowly by strong Hydrochloric Acid. Its strong magnetism, black streak, and very common occurrence in regular octahedrons distinguish Magnetite at once from every other common mineral, and any mineral possessing these three characters may be put down as Magnetite with scarcely any risk of error. Two rare minerals, Magnesioferrite and Jacobsite, occur in black, magnetic, regular octahedrons, but the powder of both is brown. Minerals formed of Ferric Oxide. Fe 2 3 . G. 4*5 5 '3. The names Specular Iron Ore, Eisenglanz, Per Oligiste, are applied to the crystallized varieties of these minerals. Rhombohedral. Most commonly in clusters of very flat, knife-edged crystals. H. 6 '5. Generally black or steel colour. Streak red. Some care is needed to determine the streak : the crystals are too hard to be scratched with the knife, even when powdered the powder is often at first black, but when well rubbed down in an agate mortar it turns red. In the same way in the case of some compact and impure varieties which are soft enough to be scratched, the scratch when first made is black, but it reddens if the knife be worked backwards and forwards for a while. The mark made on porcelain is often black at first, but turns red when it is rubbed with the finger. B.B. infusible. The solubility of Ferric Oxide in acids varies ; both soluble and insoluble or imperfectly soluble modifications being known. Natural Ferric Oxide generally gives way to some extent when boiled even in dilute Hydrochloric Acid, but usually decomposition goes on very slowly. The red streak is the most valuable character for identifying Specular Iron Ore : it distinguishes it at once from Magnetite. A mineral that becomes magnetic on heating or gives the reactions for Iron with Borax, and has a red streak and a hardness above 6, is Specular Iron Ore. But the black-bladed crystals, too hard to be scratched by the knife, can scarcely be mistaken, and may generally be safely put down as Specular Iron Ore without testing them. Under the head of Haematite are included varieties of these minerals containing more or less of earthy and other impurities. The amount of admixture varies very much and different forms result of which the following are the most important. Kidney Ore, Rotheisenerz. Kidney-shaped nodules, made of con- centric coats, with or without radiated structure. Compact or crypto- crystalline. Often very pure and then almost as hard as Specular Iron Ore ; soft in proportion to the amount of clayey impurity. Argillaceous Haematite, containing a large amount of clayey matter. It has often arisen from the oxidation of Clay Ironstone. Reddle or Red Ochre.- Clay stained red by Ferric Oxide. All these have a red streak, usually obtained easily. There are also varieties intermediate between Specular Iron Ore and 148 Geology. Haematite. Compact or cryptocrystalline. Black to red. With red streak, obtained with more or less difficulty according to the degree of hardness. Martite is Ferric Oxide crystallized in regular octahedrons, probably a pseudomorph after Magnetite. Titaniferous Iron Ore, Titaneisen, Ilmenite, Menaccanite. A number of minerals containing Oxides of Iron and Titanium are included under this head. The composition of the majority may be represented by the formula m(Fe:Mn:Mg:Ca)OTi0 2 + n(Fe 2 : Cr 2 )0 3 , or neglecting the less important constituents, by the shorter formula mFeOTiO 2 .nFe 2 3 . Other Titaniferous Iron Ores contain no Ferric Oxide and their composition is simply FeOTi0 2 ; and there are others which cannot be brought under either of these formulae, for these see Rammelsberg, Mineralchemie, ii. 148. The amount of Titanic Oxide varies from 10 to 57 per cent. G. 4*5 5. Ehombohedral, but most commonly massive. H. 5 6. Colour iron black ; powder black to brownish red. B.B. infusible in O.F., edges in some cases slightly rounded in RF. Gives an Iron bead with Microcosmic Salt, and if this is removed from the wire and treated with tin on charcoal it assumes the violet colour of Titanium, if that element is present to a sufficient amount. The bead should be saturated or nearly saturated before tin is added, since Titanium does not colour very strongly. When the bead is removed from the flame it should be allowed to stand till it is begin- ning to harden. It is then taken up in a pair of forceps with platinum tops, and the ends are allowed to close slowly. The colour is then easily recognised in the flattened bead. If the tin be applied cautiously so as not to come in contact with the wire, the bead need not be removed from the wire ; but great care is needed, for if any tin gets on to the platinum a fusible alloy is formed, and the bead drops off. Or the mineral may be fused with Bisulphate of Potash. This is conveniently done in a small platinum crucible over a Bunsen flame or a spirit-lamp : the heat should be applied gradually, the crucible being first placed on the tip of the flame and then gradually sunk deeper into it and kept there till the ore is decomposed. The crucible with the fused mass is then warmed in water, but not boiled ; the amount of water used should not be more than is sufficient to dissolve the soluble part. Any insoluble portion is removed by filtration, a few drops of strong Nitric Acid are added to the filtrate, and it is diluted with at least six times its bulk of water and boiled. Titanic Acid separates during the boiling as a white powder. Confirm by testing the powder in a Microcosmic Salt bead B.B. Titaniferous Ironsand or Iserine consists of grains of Titaniferous Iron Ore, which appear to be regular octahedrons. It has been looked upon as a pseudomorph, but it seems likely that the crystals are not really octahedral but are modified rhombohedrons, and that the mineral is only one of the forms of Ilmenite. This Ironsand contains grains of Magnetite, Garnet, Quartz, and other minerals besides that after which it is named. Rammelsberg writes the composition of Iserine 4FeOTiOj.FeA.TiO,, Rock-forming Minerals. 149 Limonite, Brown Haematite, Brauneisenstein. 2Fe 2 O 3 .3H 2 O. Hy- drated Ferric Oxide or Ferric Hydrate; 59 '9 per cent, of metallic Iron. G. 3-64. In stalactitic or botryoidal forms, sometimes with radiated structure ; also massive ; some varieties contain a large amount of earthy impurities. H. of purer forms 5 5 '5 ; earthy forms often softer. Various shades of brown to yellow. Streak yellowish brown. B.B. infusible, but becomes magnetic ; gives water in the closed tube. Soluble very slowly in Hydrochloric Acid. Gothite, Nadeleisenerz, Fe 2 O 3 H 2 0, and Turgite, HydroJwmatite, 2Fe 2 O 3 .H 2 O, are other Ferric Hydrates. Gothite is known crystallized in the trimetric system with very perfect brachydiagonal cleavage. Turgite is mostly more or less red in colour : it flies to pieces in the closed tube. Ferric Hydrates have arisen from the hydration of Haematite, and more frequently from the alteration of Carbonate of Iron. That a mineral is one of the Ferric Hydrates may be shown thus. The presence of Iron is generally sufficiently shown by their rusty appearance ; in case of doubt it may be established by treating them B.B. in a Borax bead or by trying if they become magnetic by heating B.B. They are distinguished from Haematite by their giving water in the closed tube, and by their yellow or brown streak. We may sometimes discriminate the species thus. Limonite and Turgite are not known crystallized. Hence a crystallized Ferric Hydrate is Gothite, the commonest forms are longitudinally-striated prisms, and plates bounded by brachypina- coids. The behaviour of Turgite in the closed tube serves to distinguish it from Limonite. But Limonite can be distinguished from massive Gothite only by analysis, or at least by determining the amount of water lost by heating in a closed tube. (G 2) CARBONATES OF IRON. These have been already described on p. 138. (G 3) SULPHIDES OF IRON. Iron Pyrites, Pyrites, Mundic, Eisenkies* Schwef elides. FeSo. Bisulphide of Iron. Iron 48 per cent, Sulphur 52 per cent., if pure. G. 4-835-2. Monometric. The cube is by far the commonest form, pentagonal dodecahedrons not rare. Very frequently in nodules with radiated structure. * Eisenkies, Iron Flint, on account of its hardness and striking fire with steel. The German names of many other metallic sulphides, which have nothing flinty about them, end in "kies." In fact this affix from having been very happily applied to the name of the commonest of these minerals has been extended to the whole class, and in German mineralogical terminology has become equivalent to "sulphide." Thus the soft Copper Pyrites, which can be scratched with the knife, is in German Kupferkies. 150 Geology. H. 6 6 '5, strikes fire with steel. Brittle. Brass yellow : colour remarkably uniform. Streak black. B.B. decrepitates very often : powdered, made into a paste with water, and gently heated in O.F., gives off strong sulphurous vapours : heated more strongly fuses to black magnetic globule. A sublimate of sulphur in closed tube, vapours of sulphur dioxide in open tube. Very slowly attacked by Hydrochloric Acid, soluble in dilute Nitric Acid with separation of sulphur. Iron Pyrites is the commonest of all minerals and it is scarcely possible to mistake it. It is always brass yellow and varies very little even in tint, and it cannot be touched with the knife ; these two characters are alone enough for its recognition. If crystallized it is nearly always in cubes, frequently with striated faces ; when it assumes other shapes, they can generally be easily seen to be mono- metric. Should the student be in any doubt, simple heating B.B. will generally settle the matter.' There will be a strong smell of sulphur and none of arsenic, and a magnetic residue. The absence of arsenic shows that it is not Mispickel. If not Pyrites it is either Marcasite or Pyrrhotine. The distinctive characters of these minerals are given under their descriptions. Treatment in the closed tube leads to similar results. Marcasite, White Iron Pyrites, StrahlMes (Radiated Pyrites}. FeS 2 . A second form of crystallized Bisulphide of Iron. G. 4*68 4'S5. Trimetric. In modified prisms, often tabular. Usually compact or cryptocrystalline ; radiated nodular forms very common, whence its German name. H. 6 6 '5. Generally of a paler yellow than Pyrites, often inclining to grey, black, or green. B.B. and chemical reactions the same as Pyrites, but very much more liable to decomposition. Except when in distinct crystals Marcasite is not always easy to distinguish from Pyrites. Its colour is a good guide, and the readi- ness with which it decomposes. The student will probably not be long in discovering that some of the nodular masses of Iron bisulphide which he brings home with him crumble down in a few months or even less to a heap of whitish powder. These are Marcasite. Pyrrhotine, Magnetic Pyrites, Magnetkies. Sulphide of Iron whose composition may be represented by the formula Fe n S n+1 , where n may have any integral value from 5 to 16.* Hexagonal, in tabular six-sided crystals, and cross-shaped twins. Cleaves parallel to the base. Cryptocrystalline and compact. H. 3*5 4-5. Brittle. Between bronze yellow and copper red : tarnishes very speedily. Streak dark greyish black. Magnetic, but to a variable degree ; the powder generally attracted by the magnet, even if the mineral fails to affect the needle. On charcoal it fuses easily to a magnetic mass. Unchanged in closed tube ; gives off sulphur dioxide in open tube. Readily decom- posed by dilute Hydrochloric Acid. * Neues Jahrbuch, 1880, vol. ii. p. 303. Rock-forming Minerals. 1 5 i Pyrrhotine is much softer than Iron Pyrites, but this will not always distinguish it from weathered specimens of that mineral. Its magnetism cannot always be relied on. Perhaps it is best known from Iron Pyrites by the ease with which it is attacked by dilute boiling Hydrochloric Acid ; Iron Pyrites gives way very slowly indeed, Pyrrho- tine is at once decomposed. M 1*1 ticket, Arsenopyrite, Arsenical Iron Pyrites, Arsenikkies. FeS 2 FeAs 2 , Sulpharsenide of Iron, or (Fe:As)S 2 , Double Sulphide of Arsenic and Iron. G. 6 6*4. Trimetric. Crystals often very complicated. Very commonly crypto- crystalline. Brittle. H. 5 -5 6. Colour varying from silver white to steel grey with blackish tarnish. Streak dark greyish black. B.B. fuses on charcoal to a magnetic globule, giving off fumes of arsenic which are easily known by their strong garlic smell, and coats the coal with white arsenous acid. Gently heated in closed tube, a sublimate of sulphide of arsenic which is red when hot and brown when cold, and a deposit of sulphur beyond sometimes ; further gentle heating produces a mirror of metallic arsenic beyond which is a black lustrous velvety sublimate, and beyond that there may be a deposit of sulphide of arsenic. Decomposed by dilute Nitric Acid, with separa- tion of sulphur and arsenious acid. The general look of Mispickel is usually marked enough to enable us to recognise it without much risk of error. Its behaviour in the closed tube is characteristic. There is some risk of mistaking it for Smaltine (CoAs 2 ), but by proper treatment Smaltine can always be made to show the colour of Cobalt with Borax, and normal Smaltine does not give a sublimate of sulphide of arsenic in the closed tube. A variety of Mispickel, Danaite, which contains Cobalt, can some- times be distinguished from Smaltine only by analysis ; but the Cobalt may sometimes be recognised by fusing the mineral in a Borax bead and treating the bead with Tin on charcoal. The Iron colour disappears and a faint Cobalt blue becomes visible. Leucopyrite, Fe 2 As 3 , and Lollingite, FeAs 2 , are a good deal like Mispickel. They give however little or no Sulphur reaction. Lolling- ite fuses only on the surface. They are not common minerals. (G 4) SILICATES OF IKON. Iron is an important constituent of many of the Silicates which have been already described. The two following Silicates of Iron are not common, but may be noticed. Ilvaite, Lievrite, Yenite. A hydrated Silicate of Iron and Calcium ; the water is perhaps due to alteration. G. 3 '7 4 '2. Trimetric. Often in clusters of slender complicated prisms ; also massive. Brittle. H. 5 -5 6. Fuses very easily to a black magnetic globule. Decomposed by Hydrochloric Acid and gelatinizes very readily. Ilvaite may be mistaken for some of the ores of Manganese, and it 152 Geology. sometimes contains enough of that element to give a Manganese reaction with Carbonate of Soda. But its easy fusibility at once dis- tinguishes it. Crocidolite is a beautiful blue or green fibrous mineral, allied to Hornblende in composition, but containing alkalies. It is at once recognised by its appearance. G (5) OTHER COMPOUNDS OF IRON. Wolfram. (Fe:Mn.)W0 4 . Tungstate of Iron and Manganese, the relative quantities of the two bases being very variable. G. 7*19 7 '54, for varieties rich in Manganese generally below 7 '25. Monoclinic. Perfect cleavage parallel to the clinopinacoid, by which it is readily split into thin plates. H. 5 5 '5. Blackish. Streak black when Iron, reddish brown when Manganese predominates. B.B. fuses not very easily; the globule crystallizes on the surface in cooling. Fusibility varies somewhat in different specimens, it probably increases with the percentage of Iron. Only very slowly attacked by strong Hydrochloric Acid. Decom- posed much more rapidly by Aqua Regia : after a very short boiling a sufficient amount of Tungstic Acid separates in the form of a yellow powder to be quite recognisable ; some hours' boiling is required for complete decomposition. A more convenient way of detecting Tungstic Acid is to fuse a mixture of the powdered mineral and Car- bonate of Soda with a little Nitre in a platinum spoon. The fused mass is green, indicating the presence of Manganese. If it be dissolved in dilute Hydrochloric Acid, Tungstic Acid separates as a yellow powder. This gives with Microcosmic Salt a bead which is yellow while hot and colourless when cold in O.F., and blue in R.F. The solution after the Tungstic Acid has been separated by filtration gives a rusty pre- cipitate of Ferric Hydrate with Ammonia, and the presence of Man- ganese is further indicated by this turning black when exposed to the air. The behaviour with Microcosmic Salt is somewhat capricious : usually the bead is coloured by Iron, and if it contain only a little of the assay, may show the violet of Manganese on cooling after treatment in O.F. Rarely the bead shows the blue of Tungsten, but I have not been able to ascertain what special treatment it is which produces this result. The perfect clinopinacoid cleavage, the brilliant metallic lustre of its cleavage planes, and its high specific gravity are alone usually sufficient for the recognition of Wolfram. Treatment with Carbonate of Soda is the most useful confirmatory test. The green colour of the fused mass shows the presence of Manganese : the yellow Tungstic Acid separates and may be tested in a Microcosmic Salt bead : Ammonia added to the solution precipitates Ferric Hydrate. Chromite, Chromic Iron, Chromeisenstein, Fer Chrome. (Fe:Mg:Cr) 0.(Cr 2 :Al 2 :Fe 2 )O3. Ferrosoferric Oxide, in which part of the Ferrous Oxide is replaced by Magnesia and Chromous Oxide, and part of the Ferric Oxide by Chromic Oxide and Alumina. G. 4*32 4*57. Rock-forming Minerals. 1 5 3 Monometric, but usually compact. H. 5'5. Brittle. Colour black or brownish black. Streak brown. Sometimes magnetic. B.B. infusible : dissolves slowly in Borax; in O.F. the bead when hot is reddish or yellowish, but on cooling usually assumes the green of Chromic Oxide ; in R.F. the emerald green of that oxide is obtained. It may happen that the colour due to Iron somewhat masks the Chromium colour, but in that case the latter appears when the bead is treated with Tin on charcoal. Treated on charcoal in R.F. with Carbonate of Soda or a mixture of Carbonate of Soda and Cyanide of Potassium, metallic Iron is obtained ; it goes into the coal and may be separated in the usual way by washing. In a mineral suspected to be Chromite the presence of Iron may be ascertained by reducing on charcoal. The only ore of Iron it is likely to be mistaken for is Magnetite, and from that it may be distinguished by its giving the Chromium reaction with Borax. Chromite has very often a blotchy and brecciated appearance, and the fragments of which it seems to be made up are coated with a yellowish or greenish soapy film, probably of some serpentinous material. Vivianite. When unaltered Fe 3 P 2 8 .8H 2 0, or Hydrated Ferrous Phosphate. Usually converted by oxidation into a Ferrosoferric Phosphate. G. 2'58 2 -68. Monoclinic. Perfect cleavage parallel to the clinopinacoid, by which it may be split into thin plates, that are flexible when thin enough. H. 1-52. Sectile. White when unaltered, but usually oxidized to Ferrosoferric Phos- phate and then blue, sometimes green. Crystals often green perpen- dicular to the cleavage and blue parallel to it. Streak colourless to bluish white, dry powder often brown. B.B. fuses easily and gives to the flame the bluish-green colour of Phosphoric Acid, specially if it be moistened with Sulphuric Acid ; soluble in Hydrochloric and Nitric Acids. Water in closed tube. Its blue colour, its softness, and its perfect cleavage generally enable us to recognise Vivianite with ease when crystallized. If any doubt exist, the presence of Iron may be detected in a Borax bead, the flame colouration indicates Phosphoric Acid, and water is obtained in the closed tube. Or the mineral may be dissolved in acid, Iron tested for in the usual way, and Phosphoric Acid detected by adding to a small quantity of the solution a moderately large quantity of a solution of Ammonium Molybdate in Nitric Acid, when a yellow precipitate will come down on warming. There are some comparatively rare Ferric Phosphates which would give similar reactions, but they differ in hard- ness and other physical properties from Vivianite and cannot be con- founded with it. Franklinite. (Fe:Zn:Mn)O.(Fe 2 :Mn 2 )0 3 . Ferro-manganate of Fer- rous, Manganous, and Zinc Oxides. G. 5 '069. Monometric, commonest form regular octahedron ; also massive. H. 5-5 6-5. Brittle. Slightly magnetic. Black. Streak dark reddish brown, that is to say the powder assumes this colour when rubbed down in an agate mortar : the mark made on porcelain is often black. 154 Geology. B.B. infusible. With a small amount of the assay a Borax bead in O.F. may show the Iron colour when hot and the Manganese colour when cold, but Manganese may be more easily detected by fusion with Carbonate of Soda and Nitre. If the powdered mineral is mixed with Borax and Carbonate of Soda and heated on charcoal in R.F. an incrustation of Zinc Oxide is obtained, which is yellow when hot and white when cold. Very slowly decomposed by strong Hydrochloric Acid with evolution of Chlorine. Franklinite resembles Magnetite in its hardness, infusibility, crystal- line habit, and to some extent in its Magnetism. It differs in streak. No amount of rubbing will make the powder of Magnetite anything but black, the powder of Franklinite turns distinctly reddish- brown. It is also highly improbable, though perhaps not impossible,* that any Magnetite would give the strong Manganese reaction of Franklinite. The presence of Zinc too can be ascertained in the way described. In very doubtful cases the mineral may be fused with Bisulphate of Potash, and tested by wet methods. Occurrence of Iron in Rod's. Magnetite, Ilmenite, Specular Iron Ore, and Limonite enter, though only to a comparatively small amount, into the composition of many of the crystalline rocks, and occasionally occur in sufficient quantity to form beds or masses of rock. But for the geologist the most important part played by Iron is as a colouring agent. Scarcely any rock is free from Iron. In many it is present as Ferrous Carbonate, which is white when pure and therefore imparts no colour to the rock. Rocks which contain Iron under this form are usually bluish or greyish, the colour being due sometimes to organic matter sometimes to various inorganic substances. Rocks however seldom show this bluish or greyish hue except at some depth below the surface, or where they have been otherwise shielded from the action of the air. Where they have been exposed they are commonly red, brown, or yellow. Ferrous Carbonate is an unstable compound, and under the oxidizing influence of the atmosphere and of water becomes converted either into Ferric Oxide, t or one of the Ferric Hydrates, and the colours given by these compounds are strong enough to overpower the original grey hue of the rock. Ferric Oxide colours red ; Ferric Hydrate generally produces some tint of brown or yellow, the exact shade depending perhaps on the degree of hydration. The student may observe instances of this change of colour in the sinking of shafts or wells : the sandstones brought up from any depth are almost invariably blue or grey ; the same beds when quarried at the surface are brown or yellow. The same difference in colour may be noticed between the top and bottom beds of a deep quarry. It is not uncommon too to come across blocks of stone which are blue inside, "blue-hearted," and have a brown or yellow outside crust. * An earthy Magnetite containing 17 per cent, of Manganous Oxide is known ; Kammelsberg, Mineralch^mie, ii. 132. t 2 FeC0 3 + O = Fe 2 O 3 + 2C0 2 . Rock-forming Minerals. 1 5 5 This change has naturally gone on to a larger extent in porous rocks like sandstone than in impervious clayey rocks.* The blue colour, of rocks is caused by finely-disseminated Iron Pyrites in some cases,t in others perhaps by Ferrosoferric Phosphate ; the latter salt may. also be the cause of the green colour of certain rocks, while in other cases this colour may be due to a Silicate of Iron, and sometimes perhaps to a Ferric Hydrate, or a Ferrosoferric Hydrate. Magnetite occurs in crystalline rocks very frequently in octahedral crystals, and its sections in thin slices are square or triangular : it also is not uncommon in irregular crystalline grains. It is opaque, and has a bluish metallic lustre by reflected light. Specular Iron Ore occurs in irregular hexagonal plates or in scales or flecks without any regular outline. Usually some part of a piece of this mineral is translucent in thin slices, and is of a red, yellowish- red, or yellow colour, according to its thickness, by transmitted light. By reflected light it is bluish when crystallized, red when in an earthy state. Ilmenite is opaque like Magnetite, and when in irregular grains is not easily distinguished from it under the microscope ; it gives how- ever a brownish sub-metallic lustre by reflected light. When crystal- lized its sections are either irregular hexagons or more commonly are curious skeleton-like frameworks of rhombohedral crystals. Ilmenite too has very frequently associated with it a greyish-white decomposition product to which the name Leucoxene has been given ; it replaces to a greater or less extent the original crystals, running through them in directions related to the crystalline form, and some- times leaving only a mere skeleton of the mineral remaining. Von Lasaulx has found that a decomposition product of Rutile, which he believes to be identical with Leucoxene, is a Calcium Titanate, with the composition CaO.2TiO 2 . He calls it Titauomorphite.% Chromite cannot often be distinguished from Magnetite under the microscope, but it has been observed in certain cases to be traversed by irregular cracks which are lined with a colourless translucent substance that gives colours with polarized light. This is probably the soapy substance which may be so often noticed spreading over the fragments of which a mass of Chromite is composed. Pyrites gives like Magnetite opaque square sections : but by reflected light it has a more or less pronounced glistening yellow lustre. H. COMPOUNDS OF MANGANESE. On account of the strong affinity of Manganese for Oxygen, the commonest natural compounds of Manganese are oxides, but a sulphide, carbonate, and silicate also occur native. If a little Carbonate of Soda be fused in a platinum spoon, and a * See Maw " On the Distribution of Iron in Variegated Strata," Quart. Journ. Geol. Soc. xxiv. (1868) 351. Dawson, ibid. v. 25. The Geology of Rutland (Memoirs of the Geological Survey of England and Wales), chap. vi. t Ebelman, Bull. Soc. Geol. de France, 2nd ser. ix. 221. Neues Jahrbuch, 1879, 568. 156 Geology. small quantity of a mineral containing Manganese be added with a fragment of a crystal of Potassium Nitrate, and the whole fused together, a green mass of Sodium Manganate is obtained. This is a very delicate test. A Borax bead containing Oxide of Manganese fused in O.F. is amethyst coloured when hot and violet when cold : in R.F. it becomes colourless. Alabandite, Manqanglanz, Manganblende.* MnS. Manganese Sul- phide. G. 3-95 4-01. Monometric with cubical cleavage : usually massive. H. 4. Black. Streak green. B.B. fusible with difficulty. Decomposed by Hydrochloric Acid with evolution of Sulphuretted Hydrogen. The green streak of this mineral is very characteristic and distin- guishes it from the natural Oxides of Manganese. It also gives a Sulphur reaction with Carbonate of Soda. Pyrolusite. Mn0 2 . Manganese Peroxide. G. 4 '82. Trimetric. Angle of unit prism 93 40'. Cleaves parallel to faces of unit prism and brachypinacoid. Also massive, and frequently in concentric kidney-shaped nodules. H. 2 2 '5. Black, sometimes bluish. Streak black or bluish black. B.B. infusible. Polianite agrees with Pyrolusite in its composition and physical properties with the exception of its hardness, which according to Dana is greater than 5. Psilomelane. Composition very variable, but may be represented by the formula 4MnO 2 + RO, where R is Mn, Ba, K 2 , or H,. G. 37 47. Not known crystallized; in botryoidal, kidney-shaped, and stalac- titic masses, and compact. H. 5 6. Black or grey. Streak brownish black. B.B. infusible, or fusible only with difficulty. Usually gives water in closed tube. Soluble in Hydrochloric Acid with evolution of Chlorine. Sulphuric Acid added to the solution will generally give a precipitate of Barium Sulphate : if this be separated by filtration, a little taken up on the loop of a fine platinum wire, moistened with dilute Hydrochloric Acid, and held in the Bunsen flame, it will show the yellowish green of Barium. If the precipitate be very small in quantity, a portion of the filter paper may be rolled up and placed in the loop and slowly burned in the flame ; the wire when cautiously moistened with Hydrochloric Acid will generally show the Barium colour. Braunite. M 2 3 . Manganic Oxide. Sometimes it contains Silica, when its composition may be 3Mn 2 3 .MnOSi0 2 . G. 475 4-82. Dimetric. H. 6 6 -5. * Blende is "blind," and the term was originally applied to the natural Sulphide of Zinc, which often accompanies Galena and bears a superficial resemblance to it. When however this ore was found to contain no Lead, it was styled Blind (i.e. false) Lead ore. Subsequently the term Blende became extended to other metallic sulphides, and in German mineralogical terminology it has become equivalent to sulphide. Rock-forming Minerals. 1 5 7 Colour and streak dark brownish black. B.B. infusible. Siliceous varieties gelatinize with Hydrochloric Acid. Manganite. Mn 2 O 3 H 2 0. Hydrated Manganic Oxide. G. 4'2 4 -4. Trimetric. Cleavage parallel to faces of unit prism and brachy- diagonal, the latter the more perfect. H. 4. Black to grey. Streak reddish brown, sometimes nearly black. B.B. infusible. Water in closed tube. Hausmannite. Mn 3 4 . Manganoso-manganic Oxide. G. 4 '7 2 2. Dimetric. H. 5 5 '5. Brownish black. Streak chestnut brown. B.B. infusible. All the Oxides of Manganese just described are decomposed by Hydrochloric Acid with evolution of Chlorine. It is generally easy enough to decide whether a mineral is one of the Oxides of Manganese. They are all black or dark in colour, have a black or dark streak, give a strong Manganese reaction, are infusible or fusible only with great difficulty, and are decomposed by Hydrochloric Acid with evolution of Chlorine. The only other mineral which combines all these properties is Franklinite. If Franklinite is crystallized, the shape at once distinguishes it; the Oxides of Manganese rarely, if ever, contain enough Iron to give the strong Iron reaction of Frank- Unite in a Borax bead ; and Zinc can be recognised in Franklinite in charcoal B.B. But when we have satisfied ourselves that a mineral is one of the natural Oxides of Manganese, it is often by no means easy to decide which. If it gives water in the closed tube, it is either Manganite or Psilomelane. Psilomelane is harder than Manganite, and Barium may generally be detected in it. Also Psilomelane does not occur crystallized, and a crystallized hydrated Oxide of Manganese is there- fore most likely Manganite. Of the anhydrous oxides Pyrolusite is distinguished by its softness : Braunite and Hausmannite are not readily discriminated, but Braunite is the harder and its streak darker. A determination of the amount of Chlorine given off by a given weight of the mineral will show which oxide it is, supposing it to be fairly pure. In rocks Manganese generally occurs as an Oxide in thin films coating the faces of cracks or the surfaces of planes of bedding. Frequently these films assume branching shapes which resemble moss or seaweed ; they are then known as Dendritic Films or Dendrites. Dialogite, Rhodocroisite, Manganspath. MnC0 3 . Manganous Car- bonate. In nature part of the Manganese is often replaced by Calcium or Iron. G. 3 "4 3 '7. Khombohedral, with perfect rhombohedral cleavage. H. 3 '5 4 '5. Brittle. P t ose red, but also occasionally of other colours. Streak white. Its crystallization, cleavage, and colour are often enough to determine 158 Geology. it. In cases of doubt it dissolves with effervescence in warm acid and gives a Manganese reaction with Carbonate of Soda. Rhodonite. MnOSi0 2 . Silicate of Manganese. Part of Manganese often replaced by Calcium, Iron, or Zinc. G. 3 '4 3 '68. Triclinic, with perfect cleavage parallel to the faces of the unit prism. Usually massive. H. 5 '5 6*5. Very tough when massive. Eed, and other colours ; darkens on exposure to the air. Streak white. B.B. fuses easily. Slightly attacked by acids. Generally its red colour and the strong Manganese reaction it gives with Carbonate of Soda will distinguish it. It differs from Dialogite in the absence of rhombohedral cleavage and in its hardness ; generally in not effervescing with acid ; but impure Rhodonite which contains Calcium Carbonate does effervesce. Even when black from exposure, its easy fusibility distinguishes it from the Oxides of Manganese. Besides the black colour is only superficial. /. COMPOUNDS OF ALUMINIUM. Alumina (A1. 2 3 ) is the most important natural compound of Alumi- nium, and the following minerals into the composition of which Alumina enters must be noticed in addition to those already described. White or pale-coloured minerals containing Alumina generally turn a beautiful blue when they are moistened with Nitrate of Cobalt and heated in the O.F. Corundum. A1 2 3 . Alumina. G. 3 '91 4 '16. Hexagonal (Rhombohedral).* Cleavage sometimes wanting altogether ; basal, sometimes perfect but interrupted : rhombohedral, often perfect. H. 9. The pure, clear, brilliantly-coloured varieties are gems. Scqpphire when blue, Ruby when red, Oriental Emerald when green, Amethyst when violet, Oriental Topaz when yellow. The dark-coloured opaque varieties are styled Corundum. Emery, Schmirgel, is granular Corundum mixed with Magnetite or Specular Iron Ore. B.B. infusible. Light-coloured varieties when powdered and heated in O.F. with Nitrate of Cobalt give a blue colour. The moistening will probably have to be repeated several times and the ignition con- tinued for some time before the colour comes, and the blue will not be so intense as in the case of softer compounds of Alumina such as Alunite. Unaffected by acids. Both Corundum and Emery occur as accessory rock constituents. Their very great hardness and infusibility will distinguish them from every mineral. The Diamond alone is harder, and the perfect octa- hedral cleavage and superior hardness of the Diamond distinguish it readily from Corundum. In mixtures of Emery with Magnetite and other substances, pound finely and remove the Magnetite with a magnet. Wash the residue or boil it in dilute Hydrochloric Acid to * M. Mallard is of opinion that Corundum is Trimetric with a prismatic angle of nearly 120'. R ock-form ing Minerals. 1 5 9 remove the clayey and irony stains from the surface of the grains. There will then remain translucent crystalline grains, sometimes showing cleavage, and if a few of these are pressed with the blade of a knife on the basal cleavage plane of a topaz crystal and firmly dragged over it, they will scratch the face. Wavellite. 2(A1 2 2P0 4 ).A1 2 3 3H 2 0.9H 2 = A1 6 P 4 19 1 2H 2 0. Hydrated Aluminium Phosphate. One to two per cent, of Fluorine often present. G. 2-3162-337. Trimetric. Cleaves parallel to faces of unit prism and brachy- diagonal. Usually in globular concretions with radiated structure. H. 3-254. Streak white. B.B. swells up and colours flame green : colouration brought out by moistening with Sulphuric Acid. Blue with Nitrate of Cobalt. In closed tube gives off much water and sometimes enough Hydrofluoric Acid to turn Brazil paper yellow. Soluble in dilute Hydrochloric Acid, and if a moderately large quan- tity of a solution of Molybdate of Ammonia in Nitric Acid be added to a small quantity of the solution, and the mixture warmed, a yellow precipitate will form slowly, showing the presence of Phosphoric Acid. There are several other Hydrated Phosphates of Alumina all of which would yield the above reactions. Wavellite is the only one of the group that can be said to be in any sense a common mineral, and its usual occurrence in radiated concretions is characteristic. Aluminiie, AL.3S0 4 .3AL,0 3 .9HoO. Hydrated Aluminium Sulphate. G. 1-66. Not known crystallized. In kidney-shaped concretions, and massive. H. 1 2. Earthy to touch ; adheres to the tongue. White when pure. B. B. infusible. In closed tube gives water, which at a high temperature is rendered acid by the evolution of Sulphurous and Sulphuric Acids. Blue with Nitrate of Cobalt. Easily soluble in dilute Hydrochloric Acid. At a red heat Sulphur Trioxide is given off and Alumina remains. Alunite, Alumstone, Alaunstein. A1K2S0 4 .A] 2 O 3 3H 2 O. Hydrated basic Aluminium and Potassium Sulphate. G. 2'58 2*752. Rhombohedral with fair basal cleavage, but frequently uncrystallized when it is fibrous, granular, or of a fine impalpable texture. H. 3 '5 4. Brittle. Usually white, but sometimes coloured. B.B. infusible. Water and at a high temperature Sulphur Dioxide and Sulphur Trioxide in closed tube. Blue with Nitrate of Cobalt. Alunite is very slowly attacked both by strong Hydrochloric and Sulphuric Acids. Dilute Hydrochloric Acid seems to produce no appreciable effect : dilute Sulphuric Acid does decompose it slowly and partially. If Alunite be calcined at a moderate heat, the water is driven off from the Aluminium Hydrate (A1 2 O 3 3H>0) and anhydrous Alumina remains. The double sulphate of Aluminium and Potassium (A1K2S0 4 ) remains unaltered. If the calcined mass be treated with water the double sulphate dissolves, and the Alumina remains undis- solved. If the Alumina be separated by filtration, and the solution be evaporated, Potash Alum is obtained in cubical crystals. Great care 1 60 Geology. is required to obtain this reaction, for if the temperature rises too high during calcination the double sulphate becomes insoluble. Alunite and Aluminite are much alike both in appearance and in many of their reactions, but Alunite can usually be distinguished by its greater hardness. The readiness with which Aluminite dissolves in dilute Hydrochloric Acid and the very small effect produced by boiling Alunite in strong Hydrochloric Acid, furnish another easy means of discriminating between them. K. ELEMENTS. Two elements, Carbon and Sulphur, occur as minerals. Carbon presents itself under two forms, as the Diamond and as Graphite. Diamond. Pure carbon. G. 3 '5295. Monometric, with perfect octahedral cleavage. Faces of crystals usually curved. H. 10. Its hardness distinguishes the Diamond from every other mineral. Graphite, Plumbago, " BlodcUad." Pure carbon, but often with some ferric oxide or other impurities mechanically mixed. G. 2 '09 2-229. Sometimes in flat six-sided tables, but generally uncrystallized. H. 12. Colour dark, streak black and shining, the mark made on paper by a blacklead pencil. B.B. infusible. Not affected by acids. Its black metallic streak is the surest sign of Graphite ; this will distinguish it from impure carbonaceous substances which make a mark on paper that is black but not metallic. There is some little risk that an inexperienced observer may confuse Molybdenite and Graphite. The streak of Molybdenite is very like that of Graphite ; if it be rubbed with the finger a greenish hue is generally perceptible in it, the tint is scarcely marked enough however to serve as a means of discrimination : Molybdenite gives the usual reactions for Sulphur, but impure Graphite might do this also : Molybdenite however is soluble in Nitric Acid. But the simplest way of distinguishing the minerals is B.B. Both are infusible; in the forceps Molybdenite gives a yellowish- green colour to the flame which Graphite does not ; and if a little powdered Graphite be strongly heated on platinum foil, it will burn away slowly without flame or smoke, leaving usually a reddish-brown ash ; Molybdenite will not disappear in this way. Again if powdered Molybdenite be heated on charcoal in O.F., an incrustation is pro- duced which is yellow while hot and changes rapidly to white as it cools; if this incrustation be treated with the oxydizing flame it volatilizes and leaves a copper-brown stain. Sulphur. Trimetric. Easily recognised by its yellow colour, and by its burning in a candle with a pale-blue flame, and giving a strong smell of Sulphur Dioxide. Rock-forming Minerals. 1 6 1 L. CHLORIDES AND FLUORIDES. Rock-salt, Steinsalz, Selgemme, Halite. NaCl. Sodium Chloride. G. 2-12-257. Monometric. Usually in cubes, faces of which are sometimes hollowed. Perfect cubical cleavage. H. 2 '5. Sylvite, Sylvine. KC1. Potassium Chloride. G. T9 2. Monometric. Cubical cleavage. H. 2. Hock-salt and Sylvite agree in many of their physical properties and reactions. Both are soluble in water, and both have very much the same taste. Both give the usual reactions for Chlorine. Both are fusible. Sylvine however gives a decided Potassium colour and Rock-salt a decided Sodium colour to the flame, and they may be dis- tinguished by this test. Rock-salt is a very common, Sylvite a rare mineral. Cryolite, Kryolith. 6NaF.Al 2 F 6 . Sodium and Aluminium Fluor- ide. G. 2-93-077. The crystals very closely resemble cubes with their angles truncated by octahedral faces (fig. 19), but they are triclinic. One good cleavage. H. 2*5. Often snow white, sometimes coloured. Fuses in the flame of a candle. Blue with Nitrate of Cobalt. A bit 5x4x2 mm. fused along the edges in a common gas flame : a thin splinter will fuse in the flame of a wax lucifer match. Soluble in strong Sulphuric Acid with evolution of Hydrofluoric Acid. There are several other Sodium and Aluminium Fluorides known. Some which result from the alteration of Cryolite, are hydrated and give water in the closed tube which Cryolite does not. Some are harder than Cryolite. One only, Arksutite, would be difficult to dis- tinguish from Cryolite by physical characters alone, but it is perhaps somewhat less easily fusible. A mineral which satisfies the physical and chemical tests given above for Cryolite must be either Cryolite or Arksutite, and in the majority of cases will be the former. Carnallite. KCl.MgCl 2 .6H 2 0. Hydrated Potassium and Magne- sium Chloride. Sometimes white, but frequently a flesh red, the colour being due partly to ferric oxide and partly to organic matter. B.B. fuses easily. Easily soluble in water, leaving a red residue, which consists partly of scales of ferric oxide. A small piece of Carnallite fused on a loop of platinum wire in the Bunsen flame shows enough Potassium colour to be recognisable by the naked eye ; through Cobalt glass the Potassium colour is very strong. A white residue remains on the wire, which when moistened with Nitrate of Cobalt and heated in the oxidizing flame shows the flesh-red tint due to Magnesia. A small piece fused in a bead of Microcosmic Salt which has been saturated with Cupric Oxide colours the flame strongly with the blue of Chloride of Copper. Or the mineral may be dissolved in water and the insoluble red staining matter filtered off. The filtrate shows strong Potassium colour in the Bunsen flame through Cobalt glass : Nitrate of Silver added to a por- 1 62 Geology. tion produces a white curdy precipitate that darkens on exposure to light : Sodium Phosphate added to another portion produces a white crystalline precipitate. Boracite, Stassfurtit. 2(3MgO.4B 2 O 3 )MgCl2. Magnesium Borate and Magnesium Chloride. G. 2 '9 7 4. Crystals resemble closely Monometric forms, but their optical properties make it doubtful whether they belong to that system.''" H. 7. Fuses easily and colours the flame with the green of Boric Acid. With Nitrate of Cobalt gives the flesh colour of Magnesia. Soluble in Hydrochloric Acid. Boracite takes up water by exposure and passes into massive forms with a hardness of 4*5. For further information on Mineralogy and Crystallography the fol- lowing works may be consulted. A System of Mineralogy. T. D. Dana. Fifth edition. New York. A Text-Book of Mineralogy. E. S. Dana and T. D. Dana. New York, 1877. Manual of Determinative Mineralogy. G. T. Brush. New York. The first contains descriptions and analyses of all the minerals known up to the date of publication ; in the second the elements of Geome- trical and Physical Crystallography are laid down, and an abbreviated description of the principal known minerals is given; the third describes the method of Blowpipe Analysis. Geometrical Crystallography is treated of in Miller's " Crystallo- graphy," and in "Phillips' Mineralogy by Brooke and Miller;" in Naumann's " Elemente der theoretischen Krystallographie," and in his " Lehrbuch der reinen und ungewandten Krystallographie," and in Klein's " Einleitung in die Krystallberechnung. " "Crystallography," by H. P. Gurney (Christian Knowledge Society's Manuals of Ele- mentary Science) is an excellent introduction to the subject, and a good collection of figures of crystals will be found in " Mineralogy," by T. H. Collins, vol. i. (Collins' Advanced Science Series.) For Physical Crystallography the following work may be consulted : " Physikalische Krystallographie," P. Groth, Leipzig, 1876; and if the student wishes to master the application of the Undulatory Theory of Light to crystallographic questions, " Mineralogie Micrographique Roches Eruptives Franaises, par F. Fouque et A. M. Levy," and " Leons d'Optique Physique, par E. Verdet." The optical characters and microscopical appearance of minerals are described in Rutley's " Study of Rocks " (Longman's Text-Books of Science), in Rosenbush's " Mikroskopische Physiographic der Petrographisch wichtigen Miner- alien," and in Zirkel's " Die Mikroskopische Beschaffenheit der Miner- alien und Gesteine." Smaller works on Mineralogy, which will suffice for the needs of the ordinary student, are " Manual of Mineralogy," by T. D. Dana ; " Ele- ments of Mineralogy," by T. Nicol ; " Mineralogy," by F. Rutley * Baumhauer, Zeit. fiir Crystal!, iii. (1879) 337 ; Klein, Neues Jalirb. 1880, ii. Abhand. 209, ibid. 1881, i. Abhand. 239. Lithological Classification of Rocks. 163 (Murby's Science and Art Department Series of Text-Books) ; and " A Glossary of Mineralogy," by H. W. Bristow. In Bauerman's " Systematic Mineralogy," vol. i. (Longman's Text- Books of Science), there is an admirable summary of Geometrical and Physical Crystallography. Two excellent little books on the use of the Blowpipe are " Blowpipe Analysis by T. Landauer," authorized English edition by T. Taylor arid W. E. Kay ; " An Introduction to the use of the Mouth Blowpipe by Dr. T. Sheerer, translated by H. F. Blanford." " Tafeln zur Bestim- mung der Mineralien, F. V. Kobell, Miinchen, 1878," will also be found very useful The most complete treatise on the subject is C. F. Plattner's " Probirkunst mit dem Lothrohre, Fiinfte Auflage," Dr. T. Richter. There is an English translation of the fourth edition by T. H. Cookesley. For the chemical composition of minerals Rammelsberg's " Hand- buch der Mineralchemie " may be consulted. SECTION" IV. LITHOLOGICAL CLASSIFICATION OF ROCKS. The student when he has made himself acquainted with the principal rock-forming minerals may be compared to a child that has learned its alphabet ; and as the next step with the child is to show him how letters are put together to form words, so we must now go on to show the reader how these minerals are combined into rocks. Lithological Classification of Rocks. We will first see what results would be arrived at by Lithology, or an indoor exami- nation of hand-specimens alone. By this method of research one would be led to divide rocks into two great classes Crystalline and Non- crystalline, and a class intermediate between these two. Crystalline Rocks. A large number of rocks can be grouped together on the strength of the following common characters. Their essential constituents are Anhydrous Silicates such as Felspars, Hornblende, Augite, and Micas, with or without Quartz. Many other minerals also occur in these rocks, but it is the presence of these named that gives a character to the group. The minerals are markedly crystalline, and actual crystals are of common occurrence, frequently with irregular boundaries, but occasionally with well- defined outlines, pointed angles, and sharp edges. Where crystalline texture is not recognisable by the unaided eye, the microscope shows these rocks to be composed of an interlacing mass of small crystals. The name "crystalline" is given to these rocks on account of the almost universal presence of crystalline structure or of actual crystals in them. Such rocks would again fall into two sub-classes. In the rocks of the one, which may be called the Confusedly Crystalline, the minerals are not arranged in any order. In the rocks of the other there is a tendency for the minerals to be arranged each by itself in separate layers. This arrangement tends to make these rocks split into plates or leaves, and hence the rocks are called Schistose ( 1. Reddish sand . . . . 2. White marly limestone, upper part fissile or splitting into thin layers, lower part lumpy or rubbly 3. Brown clay splitting into thin layers 4. Soft sand 5. Hard, white, marly limestone 6. Brown clay splitting into thin layers 7. White marly limestone 8. Soft brown sand .... 9. Hard cream-coloured limestone 10. Soft brown sand 11. Solid grey blocky limestone . 3 6 1 9 1 2 1 6 2 1 9 9 8 6 5 4 12. Sandy clay 13. Stiff blue clay Relation between Stratification and Crystalline or Non-crystalline Texture. In a very large majority of cases we shall find that, if a rock is stratified, it belongs to the Non-crystalline class. And we shall also find that a very large number of the Crystalline rocks have no bedded structure or are imstratified. There will be exceptions to these generalizations. We shall meet with rocks which are bedded and crystalline as well ; but when we come to inquire how these rocks were formed, we shall find that, in most cases, either they were originally non-crystalline, and have been subsequently altered so as to acquire a crystalline texture, or that their bedding was obtained in a different way from that of the non-crystal- line stratified rocks. These and a few other exceptions will be better understood when the reader has gone through the chapters on the for- mation of rocks. Fossiliferous and Unfossiliferous Rocks. Again, in many rocks we shall find what are undoubtedly the remains of animals and vegetables, shells of molluscs, corals, bones and teeth of fish, reptiles, and other creatures, leaves, stems, and fruits of trees and plants. Sometimes these are scarcely altered at all from their original condition ; sometimes the substances of which they originally consisted have been replaced by various minerals, the change having occasionally been pro- duced so gradually that not only the external form but all the minute details of internal structure are preserved ; sometimes only an impres- sion or cast remains. All such remains are called Fossils, rocks containing them are spoken of as Fossiliferous; rocks from which they are absent as Unfossiliferous. In nearly every case we shall find that a Fossiliferous rock belongs to the Non-crystalline class. In some rare instances we may meet with fossils in Crystalline rocks, but these will be so very few that we shall come to look upon rocks of this class as Unfossiliferous. 1 68 Geology. Petrological Classification of Rocks. Subject then to cer- tain exceptions, not relatively very numerous, and some of them more apparent than real, Petrological investigations lead us to arrange the rocks of the earth's crust into two classes having the following distin- guishing characters. IST CLASS. 2ND CLASS. Crystalline. Non-crystalline. Unstratified. Stratified. Unfossiliferous. Fossiliferous. Terms connected with Stratification. We may con- veniently define here a few terms used in connection with Stratification. The thicker layers of bedded rocks are usually spoken of as Beds or Strata, and the thinner as Laminae or Stratula. Sometimes each of those portions of a group of bedded rocks, which has the same mineral composition throughout, is called a Stratum ; and, if this stratum can be split up into a number of subordinate layers, each is called a Lamina. Thus in fig. 77 we should say we had a stratum or bed (No. 3) of brown Clay overlying a stratum or bed (No. 4) of soft Sand, the first consisting wholly of Clay, the second wholly of Sand. The brown Clay however can be split up into a large number of thin par- allel layers, each of these is called a lamina, and the rock is said to be laminated or fissile. The distinction between strata and laminae is somewhat vague, but the circumstances do not admit of exact defini- tions or hard lines. Single beds of rock sometimes are as much as two hundred feet in thickness, but such are rare ; about five feet would be a general average. Lamination may go to almost any extent; in some very finely laminated rocks as many as thirty or forty layers may be counted in the thickness of an inch : such beds are usually clayey in composition, and are sometimes called Paper Shales ; very finely lami- nated siliceous and calcareous rocks are however also met with. When, as in fig. 77, the upper and under bounding surfaces of the beds are parallel, so that each bed keeps the same thickness, the bed- ding is said to be Regular. A rock which is regularly and not very thickly bedded, so that it can be split up into slabs for paving, is called Flaggy, or a Flagstone ; if the layers are thin enough for roofing pur- poses, a Tilestone* The majority of Flagstones and Tilestones are Sandstones, but some Limestones, and even some hard Argillaceous rocks, yield Flags and Tiles. When beds thin away, the bedding is Irregular or Wedge-shaped, as in fig. 85. A bed which thins away in all directions is called Lenticular or Lens-shaped. This is all we will give here under the head of Petrology. There are many points yet to be noticed respecting the structure of rocks on a large scale ; but we shall find it the best plan to take these one by one, as opportunities occur, while we pursue our inquiries into the way in which rocks were formed. Whenever, from time to time, we find that we have gathered knowledge enough to enable us to understand * This term, and not Slate, ought to be used for those rocks which split into roofing slabs along planes of bedding. We shall see by-and-by that the planes which bound roofing slates are not planes of bedding. Petrology. how any great structural peculiarity was produced, we will describe that structure and the way in which it arose. Descriptive Geology. Summary. Let us now take stock of the knowledge we have gained from Descriptive Geology. To the substances which make up the earth's crust we gave the general name of Rocks. Rocks we found to be mechanical mixtures of certain definite chemical compounds called Minerals. The number of mineral species which enter to any appreciable ex- tent into the composition of rocks we found to be small, and the chemi- cal elements of which these rock-forming minerals are composed to be not more than twenty in number. By an indoor examination of hand specimens, or Lithology, we were led to a threefold classification of rocks. 1. Crystalline Rocks, in which crystals appear with sharp angles and unrounded edges, but not arranged in any regular order. 2. Schistose Rocks, which differ from the last in having their mineral components arranged more or less in separate layers, a struc- ture which is expressed by the word Foliation. 3. Non-crystalline Rocks, in the typical forms of which the mineral components appear in the shape of grains more or less rounded, or chipped and broken. An examination of large rock masses in the field, or Petrology, leads us to a twofold classification of rocks. 1. Stratified Rocks, which are arranged in parallel layers, beds, or strata. 2. Unstratified Rocks, which possess no such bedded structure, or possess it in a minor degree. We further found that the great mass of the Crystalline rocks are Unstratified, and the great mass of the Non-crystalline rocks are Stratified. In the latter too we frequently meet with Fossils ; from the former Fossils are almost invariably absent. So that a mere examination of the composition, structure, and con- tents of rocks led us to arrange them in the two following classes. 1 ST CLASS. 2ND CLASS. Crystalline. Non-crystalline. Unstratified. Stratified. Unfossiliferous. Fossiliferous. Lastly, we pointed out that this classification was liable to exception, arid was otherwise imperfect, but that it was the best we could arrive at in the present state of our knowledge. This leads us on to inquire whether a further study of rocks will not tell us something more than we yet know about them, which will enable us to arrange them in a more satisfactory manner. In the next three chapters we shall find that what we want for the purpose is a knowledge of the way in which rocks were formed, and that, when we have mastered this branch of Geology, a natural and consistent classification follows from it as a matter of course. CHAPTER III. LITHOLOGICAL DESCRIPTION OF THE NON-CRYSTALLINE ROCKSDENUDA TION. " Process of time worketh such wonders, That water, which is of kind so soft, Doth pierce the marble stone asunder By little drops falling from aloft. " WYATT. SECTION I. PRINCIPLES ON WHICH THE INQUIRY INTO THE ORIGIN OF ROCKS IS BASED. WE have now made the acquaintance of the chief materials out of which rocks are made up, and have learned what are the great classes into which we subdivide the rocks themselves, according as we look at them from a Lithological or Petrological standpoint. It would be possible still to limit our attention for some while to purely Descriptive Geology : we might take one by one each individual species of the great classes of rocks and describe its lithological com- position and the structural characteristics which it shows when studied on a large scale ; and this we might do without saying a word about the causes that produced the rock and impressed on it its peculiar structure. Then we might in separate chapters treat of the origin of rocks and of rock structures. ^< > But, beyond a claim to systematic arrangement, such a scheme would possess no advantage whatever, and it would be attended by a serious evil. It would lay upon the mind of the student a burden too heavy to be borne, in that it would compel him either to carry in his memory a huge mass of bare facts up to the time when he reached that part of the book where the explanation of these facts is given ; or if, as is most likely, he found this beyond his powers, it would oblige him to be continually turning back, when the meaning of any fact was explained under the head of Historical Geology. J;o the description of that fact under Descriptive Geology. *z We will therefore no longer linger in the purely descriptive part of our subject : we will go on to describe the different kinds of rocks in detail, but we will put side by side with the description of each an explanation of the way in which it has been formedHh Thus the subject will be rendered less dry, and a great strain on the^memory will be avoided. *^ Origin of Rocks. 17 1 But the question may be very reasonably asked at the outset, How do we know that there has ever been such a thing as formation of rock ? Are we sure that the rocks have not been all along such as we see them now, and that the earth's crust did not come into being in the identical state in which it is at the present day 1 There are a host of facts that enable us to give a decided No in answer to such suggestions. One of these, the occurrence of fossils in the heart of masses of rock, which has been noticed in the first chapter, is alone sufficient to settle the question. And very slight observation of what is going on every day before our eyes is enough to convince us that, for as far back as the earth has been anything like what it is at present, the rocks of its surface must have been constantly undergoing wear and tear, and that fresh rocks must have been forming without cessation out of their ruins. The whole of this and the next chapter will be taken up with a statement of the facts on which this assertion rests, and when the reader has reviewed the evidence, he will see that but one conclusion, the one just stated, can be drawn from it. Principles on which the Origin of Rocks are deter- mined. The grand principles that must guide us in our speculations as to the origin of rocks are few and simple ; but a very extensive range of knowledge is necessary to enable us thoroughly to apply them. We have first to inquire whether there are any substances now in course of formation which are identical with rocks of the earth's crust, or any which, if not actually identical, could be made so by modifications which it is reasonable to suppose they would be likely to undergo. If we find, as we do, any such substances, we then study the causes which are now producing them, and conclude that the rocks which they resemble were produced in bygone times by similar causes. In this way we are able to give a satisfactory account of the forma- tion of many rocks. There are others which cannot have been pro- duced by any causes that come within the reach of our actual observa- tion ; but even in the case of these we can form reasonable conjectures how they arose and what changes they have gone through. We will begin with an inquiry into the origin of some of the Non- crystalline bedded rocks. These we shall find have been formed by the breaking up of some pre-existing rock, and in nearly all cases, when we trace "back their history far enough, we learn that they sprang first of all from some one of the Crystalline rocks. It would seem at first sight more natural to take the forefathers first and the descendants afterwards ; the present arrangement has been chosen, as better suited to an elementary treatise, for the following reasons. The bedded rocks are more familiar to the generality of people than the great mass of the crystalline. Further, among the causes that took part in the formation of bedded rocks are some of the commonest ope ations of Nature, so common indeed that their import- ance was long overlooked through sheer familiarity. These processes are going on every day under our eyes and can be studied by every one. There is this advantage then in directing the attention of the student first of all to the formation of this class of rocks; he can, 172 Geology. whoever and wherever he be, observe for himself some of the steps by which they have been produced, and test by his own observation the correctness of the teaching which is put before him. The processes, on the other hand, to which a great part of the Crystalline rocks are due, operate unseen to a large extent at considerable depths below the surface ; and when they do break forth and come within the range of observation, their sphere of action is confined to certain limited tracts of the earth's surface, which many persons have no chance of visiting. But before entering on an inquiry into the manner in which rocks of the non-crystalline class were formed, we must give an account of the lithological character of the principal members of that class. SECTION II. LITHOLOGICAL DESCRIPTION OF THE NON- CRYSTALLINE ROCKS. Texture. Under the head of texture we may notice that the binding cement in these rocks may be soft or small in quantity, in which case the rock is crumbly and friable ; or it may be hard and plentiful, when the rock will ~bejirm and solid. According to the size of their particles these rocks may be sub- divided into Coarsely-grained, Finely-grained, and Closely-grained or Compact. In some coarsely-grained rocks lumps, sensibly larger than the majority of the particles, are scattered through the body of the rock. Such are called Conglomerates or Pudding-stones, when the larger portions are rounded ; and Breccias, when they are angular. Subdivisions of the Non-crystalline Rocks. The great mass of the Non-crystalline rocks are made up of one or more of the four substances, Quartz, Clay, Carbonate of Lime, or Carbon. They fall naturally into four groups, according as their prevailing ingredient is the first, second, third, or fourth of these substances, and their classification will be as follows. 1st Class. Arenaceous or /Sandy Rocks. Composed mainly of rounded or broken grains of Quartz. The cement may be either siliceous, argillaceous, calcareous, or a mixture of any of these three substances. 2nd Class. Argillaceous or Clayey Rocks. One or both of the two substances included under the name of Clay is the main constituent of these 'rocks. Besides Clay the majority of the Argillaceous rocks con- tain mixtures of Sand, Carbonate of Lime, and other minerals. 3rd Class. Calcareous Rocks or Limestones. Composed of Carbonate of Lime, with admixtures of Sand, Clay, and other matters. Ith Class. Carbonaceous Rocks. Composed of Carbon, Hydrogen, Oxygen, and Nitrogen, with earthy admixtures. No hard lines can be drawn between these classes, since all sorts of intermediate forms occur. Thus for instance there are many rocks containing both Sand and Carbonate of Lime which might be placed indifferently in the 1st class as Calcareous Sandstones, or in the 3rd as Sandy Limestones ; but in a very large number of the Non- Non-crystalline Rocks. 173 crystalline rocks either Sand, Clay, or Carbonate of Lime is present in so much larger quantity than any other ingredient that we are justified in establishing the subdivisions just given, and able without any difficulty to decide in which of them a given rock ought to be placed. There remain a few rocks of comparatively limited occurrence, such as Rock-salt and Gypsum, which cannot be placed in any of the above classes, and which may be put together in a group by themselves. We may now notice some of the principal varieties of each class of the Non-crystalline rocks. 1. ARENACEOUS OR SANDY ROCKS. A mass of more or less rounded grains of Quartz, not bound together by any cement, constitutes Sand. Rock-sand is a term applied to masses of Sand which hold together sufficiently to stand up in natural rocks, but are not firm enough to yield stone for building purposes. When the Quartz grains are firmly bound together in any way, we get a strong rock and call it Sandstone. The term is generally restricted to those rocks in which there is not much difference in size among the grains. In most cases the solidity of the rock is due to a cement, which fills up the interstices between the grains and binds them together. If this cement be Carbonate of Lime, the rock is called a Calcareous Sandstone ; if Quartzose, a Siliceous Sandstone. Very siliceous Sand- stones with an even close grain are called Cank, Can/cstone, or Galliard. Many Sandstones also contain Clay : such are called Argillaceous Sand- stones. Sandstones containing recognisable bits of Felspar are called Felspathic Sandstones. Sandstones containing a large quantity of Peroxide of Iron are distinguished as Ferruginous or Rusty ; they are red, brown, or yellow in colour at the surface from the oxidation of the Iron (see p. 154). If some of the particles of a sandy rock are larger than others, so that a freshly-broken surface has a rough, gritty feel, the rock is called a Grit or Gritstone. The term however is not generally applied to friable sandy rocks, however coarse they may be, but is restricted to those which are hard and firm. When a sandy rock contains pebbles of Quartz or Quartzose rock embedded in a finer ground-mass of Sand, it forms a Siliceous or Quartzose Conglomerate. The adjective is very generally dropped and the rock styled simply a Conglomerate, because the pebbles of a great majority of Conglomerates are Quartz ; the reasons for this being, first, that Quartz is a substance very plentiful in the rocks of the earth's crust, and secondly, that on account of its great hardness it is able to survive in pebbles of considerable size wear and tear that grinds softer substances to powder. The rounded lumps in Conglomerates are of all sizes from small pebbles up to blocks some feet in diameter. In some cases the pebbles of a Quartzose Conglomerate are cemented together by substances other than Quartz, such as Carbonate of Lime or Oxide of Iron. Some other Quartzose rocks will be noticed under the head of Metamorphic Rocks. 1 74 Geology. The grains of a rock that is decidedly sandy will scratch glass, and this test, which however it is seldom necessary to apply, may be used when there is any doubt about the composition of the rock. When the Quartzose element is disguised by the presence of a large mixture of a softer substance such as Clay, the rock may be pounded, and the powder drawn with pressure between the finger and a plate of glass ; any Quartz grains that may be present will then make scratches. Or the pounded rock may be repeatedly washed with water. The clayey portion will thus be carried off and the sand grains separated. t 2. ARGILLACEOUS OR CLAYEY EOCKS. In the common acceptation of the word Clay is used to denote any earthy substance which can be worked up with water into a plastic mass, that is a mass which may be pressed into any form and will retain the shape given to it. It is also generally understood that it will retain its shape when dried by heat, though this is very imperfectly the case with many of the substances that would be called Clays in common parlance. A clayey substance will often hold together as long- as it is damp, but falls to powder when all the water is driven off. But we have already pointed out that the Clay of ordinary language includes two substances of totally different character. The one is Kaolin or China Clay, whose composition is 2SiO 2 .AL,O 3 .2H.,O, or Silica .. /. /.. /. /. 46-33 Alumina . >f . ' / . '/ .' / .' >/ . ' / . 3977 Water / - ., /'^- / ' ./ ' . ./' ./ ' . 13'90 100-00 China Clay is hydrated ; it is a Silicate of Alumina alone, and contains no alkali ; and it has been produced by the chemical decomposition of a Felspar or a Felspathic mineral. The other clayey substance is com- posed of Felspar or a Felspathic mineral reduced to a very fine powder, but not decomposed. It is eminently clayey in many of its properties ; so finely divided that when mixed with water it takes days to settle to the bottom ; fairly plastic though seldom to the same degree as Kaolin ; and it will sometimes hold together moderately well when baked. But it is anhydrous ; if it be first dried at 100 C. to drive off the mechanically mixed water, it gives off no water in a closed tube at higher temperatures : it is not a simple Silicate of Alumina, but has approximately the same composition as the mineral from which it was derived, a complex Silicate of Alumina, Alkalies, Alkaline earths, and perhaps of other substances, as the case may be ; and it has\ been pro- duced not by the chemical decomposition but by the mechanical tritii- ration of a Felspathic mineral. This substance we QtjJledS Felspathic Mud. Now Clays may be divided into two classes according as ^Kaolin or Felspathic Mud predominates in their composition. These we name and define as follows. Clays. Composed essentially of Kaolin with admixtures of other substances. Non-crystalline Rocks. i / 5 Mudstones. Composed essentially of Felspathic Mud with mixtures of other substances. In practice we shall find a great many clayey rocks which cannot strictly be placed under either of these heads, because they contain both Kaolin and Felspathic Mud in such variable proportions that it is hard to say which ought to be considered the preponderating con- stituent. But on the other hand there are rocks, in which the per- centage of one of the substances, Kaolin or Felspathic Mud, far exceeds that of the other, and these fall at once into the corresponding class. Kaolin and Felspathic Mud are most certainly distinguished by analysis, but the following method will often suffice. The Clay is elutriated with water till all grains of sand or foreign matter are removed. The residue is boiled in dilute Hydrochloric Acid to dissolve off the coating of Oxide of Iron which colours the grains : it is then filtered and well washed : a small portion is pressed with the blade of a knife into a thin plate with a sharp edge and dried. If the clay be Kaolin, this plate will be infusible before the Blowpipe ; if Felspathic Mud, its edges at least may be rounded. It may be here noted that it is the combined water which gives Kaolin its plasticity. If this be driven off by strong heat the residue is no longer plastic. This water for instance is expelled in the burn- ing of bricks, and though powdered brick will absorb a great deal of water, it is impossible to make it in the least degree plastic by any amount of water. The degree of plasticity seems to depend largely on the fineness of the particles. The following are some of the most important varieties of the Clayey rocks. The natural deposits of Kaolin contain grains of Sand, plates of Mica, and other impurities ; and when these are washed out there remains a pure white plastic Clay, used for making porcelain and the finer kinds of pottery. Pipe-day is a similar white pure Clay, which shrinks too much from heat to be available for pottery purposes. It is important that both China Clay and Pipe-clay should be free from Iron, which acts as a flux, and causes the Clay to melt instead of baking in the furnaces. Pot-days are less pure than China Clay, and the ware made from them is coloured and coarser : all that is required of them is that they should form with water a plastic mass, and be capable of baking. Still coarser Clays serve for Brick-days ; the finer varieties, con- sisting of a very finely-divided and intimate mixture of Clay and Sand, are called Brick-earth. Brick-clays should not contain too much Iron, but a moderate quantity of the Protoxide is said to give strength and hardness to bricks. Fire-days are varieties which will stand intense heat without melt- ing, so that bricks made of them do not fuse or soften when exposed to very high temperatures. As far as infusibility goes there is no widely-diffused substance that will resist heat better than Silica, and if finely-divided Silica were plastic we could not have a better material for making bricks capable of standing heat without being fused. But before we can make Silica 176 Geology. into bricks we must have some vehicle to bind the grains together, and this vehicle must be itself infusible. Such a vehicle we find in Kaolin, and hence a theoretically perfect Fire-clay is either Kaolin or a mixture of Kaolin and Silica. Kaolin shrinks and cracks in drying and firing too much to allow of its being used for brick-making, even if it were plentiful enough to be employed for this purpose. But Nature has furnished us with rocks which approach a mixture of Kaolin and Silica in composition very nearly, the Clay for instance whose analysis is given below, and to these this objection does not apply. COMPOSITION OF A FIRE-CLAY. Silica insoluble in hot alkaline ley . . . 59 '95 Silica soluble . . . . 1 '39 Residue after the above Silica is removed Silica . . 45-30 Alumina , , % 34 '08 Iron, Sesquioxide . .. "3 '27 Lime . . 0'87 Magnesia . . 1*14 Potash . . 3-05 Water . . 12 '29 which is very nearly the composition of Kaolin. It is Clays of this class which make the best fire-bricks. It is likely enough that in many cases the Silica which is intimately mixed with the Kaolin in Fire-clay has been derived from the Silicate whose decomposition gave rise to the Kaolin itself. The Alkalies were probably carried away in solution, the insoluble Silica remains behind partly combined with Alumina, partly in a free state. The free Silica occurs most frequently in a finely-divided state mixed with the Kaolin, but occasionally it seems to be represented by nodules of Chalcedony and Opal that are embedded in the Clay.* The ingredients that are most fatal to fire-resisting quality in a Clay are Alkalies (Potash, Soda), Alkaline earths, specially Lime, and Iron, for these act as fluxes and cause the bricks to run. It must not be assumed however that a Clay, because it has a theo- retically suitable composition, will necessarily make a good fire-brick. There are other conditions to be satisfied ; the brick must not crack and fly when exposed to a sudden rise or great extremes of temperature, it must support great pressure at high temperatures without crumbling, and it must resist the corrosive action of some of the slags produced in metallurgical operations. Chemical and mineralogical examination will often enable us to say that certain Clays will assuredly not make fire- bricks ; and it will enable us to say that other Clays are promising enough to make it worth while trying them ; but nothing short of making a test-brick and subjecting it to the heat that it will be required to stand will settle the question. * Zirkel, Petrographie, ii. 608 ; Naumann, Geognosie, i. 726 ; Bischoff, Chemical Geology, ii. 176 ; Wagner, Chemical Technology, p. 293 ; Roth. Allgemeine und Chemische Geologic, i. 141, 142. Non-crystalline Rocks. Fire-bricks are in some cases made out of substances composed almost entirely of Silica. The Dinas brick is made of pounded or weathered Gritstone, which contains between 98 and 99 per cent, of Silica. About 1 per cent, of Lime is mixed with the Sand, and the mixture pressed into moulds, dried, and strongly fired. The Lime causes the outside of the Quartz grains to fuse and adhere together. A stone called Ganister is used for making fire-bricks and linings to furnaces and Bessemer "converters" in Yorkshire and Lancashire. It contains up to 96 per cent, of Silica, and it is probable that the Ferric Oxide, Lime, and Alkalies present in it play the same part as the Lime in the Dinas process.* The following analyses will show the average composition of different kinds of Clay.t (1) (2) (3) (4) (5) Silica - . '. 49-44 46-38 66-68 46-32 65-10 Alumina .... 34-26 38-04 26-08 39-74 22-22 ' Iron Oxides 774 1-04 1-26 0-27 1-9-2 ; Lime .... \ 1-48 1-20 0-84 0-36 0-14 Magnesia .... 1-94 trace trace 0-44 0-18 Water . . 5-14 13 57 5-14 12-67 9-86 (1) Common Pottery Clay, will not stand heat. (2) Best Pottery Clay, burning white. (3) Coarse sandy China Clay. (4) Best Kaolin. (5) Fire-clay, Stourbridge. Loam is a mixture of Clay and Sand, the latter being present in sufficient quantity to allow of water percolating through the mass and to prevent its binding together. Clayey rocks which split into layers along planes of bedding are called Shale; Bind, Blue-bind, Plate, Shiver are other names applied by miners to the same rock. Shales containing a sufficient quantity of Iron Pyrites are used for the manufacture of Alum, and are called Alum Shales. When there is a good deal of Sand present, the rock is . 559 ; Percy, Metallurgy, volume on Fuel, pp. 92, 146, 214 ; Encyclopedia * For further details on Fire-clay see Crookes and Rohrig, Metallurgy, iii. lo Britannica, Arts. "Brick" and "Fire-clay." t See^also Catalogue of Specimens of the Clays and Plastic Strata of Great Britain in the Museum of Practical Geology, London. G. W. Maw. 178 Geology. called Arenaceous or Sandy Shale, or Stone Bind, or Rock Bind. These forms pass gradually into Argillaceous Sandstones and common Sand- stone. Shales stained dark by vegetable matter are called Carbonaceous Shale, Bass, or Batt. When such Shales contain a sufficient quantity of bituminous matter to be used for the manufacture of Paraffine they are called Oil Shales. Such Shales pass gradually into Cannel Coal occasionally. The streak of Oil Shales is usually brown. Marl is Clay containing Carbonate of Lime ; if the rock splits into plates, it is called Calcareous Shale or Marl Slate. Balls and irregularly-shaped lumps of Clay Ironstone and Iron Pyrites are very common in Clays and Shales. Crystals of Selenite are not uncommon, they are generally found in Clays containing Iron Pyrites and some calcareous matter. The oxidation of the Iron Pyrites pro- duces Sulphuric Acid, and this acts on the Carbonate of Lime and produces Sulphate of Lime. Other Clayey rocks will be noticed under the head of Metamorphic Rocks. When Clay is present to any extent in rocks, they give out an earthy smell when breathed upon. Even the hardest Clayey rocks can be worked down by pounding or grinding them with water into a more or less doughy and plastic mass. 3. CALCAREOUS ROCKS OR LIMESTONES. The predominating ingredient is Carbonate of Lime either crystallized as Calcite,* or in an uncrystalline state. Most varieties depend on the extent to which the Carbonate of Lime is mixed with clayey, sandy, and other impurities. Chalk is a white Limestone, usually soft, containing sometimes as much as 94 to 98 per cent, of Carbonate of Lime. The more clayey varieties go by the name of Chalk Marl. * Some other Limestones are as pure as Chalk ; thus some specimens of the Mountain Limestone contain only 4 per cent, of impurities. But in the majority of Limestones foreign matters are present to a large extent. When there is a considerable percentage of Clay, the rock is called an Argillaceous Limestone. The Lime obtained by burning some Argillaceous Limestone forms a mortar that sets under water : such are called Hydraulic Limestones."^ Nothing short of absolute trial will enable us to say for certain that a given limestone has this property, but the following test often indicates hydraulic quality. A small quantity of the powdered rock is calcined, conveniently in a platinum crucible over a Bunsen's lamp. The calcined mass is heated in dilute Hydrochloric Acid at a tempera- ture below the boiling-point for about an hour. Any insoluble matter * Here we have one of the inconsistencies already mentioned of placing a Crys- talline rock in the Non-crystalline class. We shall see further on that the classi- fication we are now using is merely provisional, and that its weak points are not of any moment. t See Wagner's Chemical Technology, and Watts' Dictionary of Chemistry, Art. " Silicates of Calcium." Non-crystalline Rocks. 179 is removed by filtration, and the filtrate is gently evaporated nearly to dryness. If gelatinous Silica separates out, the Limestone is very likely to be hydraulic. Limestones containing a large siliceous element are called Siliceous Limestones. When the calcareous part of such rocks has been dissolved out by the action of carbonated water, a sort of siliceous skeleton is left called Rottenstone. Passages sometimes occur from Calcareous Sandstones into Limestone, and the intermediate forms are called locally Cornstones. Some Cornstones contain so much more Carbonate of Lime than Sand that they are burned for Lime in districts where purer Limestones are not easily obtained. Limestones stained a dark colour by decomposed vegetable or ani- mal matter are called Carbonaceous or Bituminous Limestone; such rocks often give off a fetid smell when struck by the hammer, and are then spoken of as Fetid Limestone or Stinkstone. Limestones occasionally put on a conglomeratic or brecciated form, and contain pebbles or angular fragments of Quartz or other rocks. A Limestone hard and close grained enough to take a polish is called Marble. Some of the so-called Marbles of commerce however are not Calcareous rocks at all. It is extremely common to find in Limestones, and specially in the purer forms, balls and irregularly-shaped lumps of Flint or Chert. These substances also occur in thin beds and in vertical veins. Magnesian Limestones. Nearly all Limestones contain some Carbon- ate of Magnesia. When the percentage of this salt is considerable, they are called Magnesian Limestones or Dolomites. Usually no dis- tinction is drawn between these two terms ; they are used, sometimes one and sometimes the other, as if they were only two different names for the same rock. Possibly however at least three distin- guishable varieties of magnesio-calcareous rocks exist, and it may be convenient to restrict one of these terms to one form and the other to another. Whether the carbonates exist in the rocks we are considering in a state of mechanical mixture, or of chemical combination, is perhaps an open question. Both Forchhammer * and Karsten found that certain Magnesian Limestones could be separated into two parts. One, which was soluble in cold Acetic Acid, had the following composition. Carbonate of Lime ..... 97 '13 Carbonate of Magnesia . . . . . 2 '87 When this portion had been dissolved out, there remained an insoluble granular residue having the composition Carbonate of Lime ..... 53 '38 Carbonate of Magnesia . . . . . 41'42f The facts that the latter portion has nearly the theoretical composi- tion of a double Carbonate of Lime and Magnesia, and that it is * Bischoff, Chemical Geology, ii. 49. t Sterry Hunt has repeated these experiments, and finds that the conclusion arrived at is very approximative^, but not exactly, correct. Silliman's Journ. 2nd ser. xxviii. 180. 1 80 Geology. insoluble in cold Acetic Acid, led to the belief that it was a chemical compound. But insensibility to the action of Acetic Acid is a fact whose value has been somewhat diminished since the time when the experiments were made, for it has been shown that the behaviour of this acid towards carbonates varies very considerably with the circum- stances under which it comes in contact with them.* We cannot say positively therefore whether the insoluble residue of the Limestones operated on is a chemical compound or a mechanical mixture of the two carbonates. The insoluble portion has however the same composition as one of the forms of Bitter Spar, and, like it, is not acted on by cold Acetic Acid : it may therefore be looked upon as probably Bitter Spar. There is then probably one form of mag- nesio-calcareous rock consisting of Bitter Spar and Carbonate of Lime. Other rocks of the same family contain perhaps no soluble portion, and consist essentially of Bitter Spar ; and there may be others wholly soluble, consisting of Carbonate of Lime and Carbonate of Magnesia. If this be so, the following nomenclature may be usefully employed. Dolomite, a rock consisting essentially of Bitter Spar. Dolomitic Limestone, a rock which is essentially a mixture of Bitter Spar and Carbonate of Lime, or of Bitter Spar and Carbonate of Magnesia. Magnesian Limestone, a rock which is essentially a mixture of Car- bonate of Lime and Carbonate of Magnesia. In Nature all these rocks contain frequently large quantities of sandy and clayey impurities, which give rise to sandy or marly varieties. There are other calcareous rocks which have been produced by the alteration of some of the above forms. These will be described in the chapter on Metamorphism. Limestone rocks may be readily recognised by touching them with a little dilute acid, when they will effervesce, the escape of gas being- more plentiful as the rock is more pure. The effervescence indicates that the rock contains a Carbonate, and in most cases this will be Car- bonate of Lime. If any doubt exist as to the base, a small quantity of the rock may be dissolved and the solution tested in the Bunsen flame. If the rock be a Limestone, the Calcium colour will show it- self. Any sandy or clayey impurities will remain undissolved, and by filtering and drying the residue its composition may be roughly deter- mined. If clayey, it will work up into a plastic .mass with water, and give out an earthy odour when breathed upon ; if sandy, its particles will scratch glass. In practice however it is seldom necessary to take all this trouble ; Limestones are soft enough to be scratched even with the blunt edge of a hammer, and have a look which is soon recognised after a little experience. They are also dissolved away by rain-water, and their exposed surface has a cavernous and worn shape not easily described in words, though to get to know the look of it is easy. By such signs the practical geologist soon learns to recognise a Limestone ; he may distinguish the earthy varieties by their smell when breathed * I am indebted to ray friend Dr. Thorpe for calling my attention to this fact. Non-crystalline Rocks. 181 upon, and in any sandy forms grains of Quartz can generally be de- tected with a pocket lens on a freshly-broken surface, or will be seen sticking up on a weathered face, or if they are numerous will scratch glass when a piece of the rock is drawn over it. Magnesian Limestones, which approach Dolomite in composition, effervesce feebly, or not at all, with cold acids ; readily, when powdered, in warm acid. In Dolomitic or Magnesian Limestones it is often necessary to use the method of analysis given on p. 139 to be sure of the presence of Magnesia. Very often however even in these rocks there are cavities, the walls of which are coated with crystals of Bitter Spar, and these may be distinguished from Calcite by their feeble effervescence with acids, and by their faces being frequently curved and somewhat pearly in lustre. By these tests we can sometimes form a fair guess that a Limestone is Magnesian ; frequently however nothing short of analysis will settle the point. 4. CARBONACEOUS ROCKS. There are only two varieties sufficiently common to deserve notice here, Coal and Graphite. Coal is too well known to need any description. We are led to the conclusion that Coal is mineralized vegetable matter by two independent lines of argument. The first is based on its chemical composition and will be understood by the aid of the following table. Chemical Composition of Woody Fibre and the principal kinds of Coal, neglecting the Ask. Carbon. Hydrogen. Oxygen and Nitrogen. Wood . 49-6 6-5 43-9 Humus 54-8 4-8 40-4 Peat . 60-8 5-8 33-4 Lignite 67-5 5-5 27-0 Brown Coal . 72-9 5'4 217 Household Coal . 80-5 5'5 14-0 Steam Coal . 84-3 51 10-6 Anthracite . 95-3 2-5 2-2 From this table we see that the elements which make up Coal are the same as those which compose woody fibre. This fact might seem at first sight to carry no great weight with it, because the proportions of these elements to one another are widely different in different kinds of Coal, and are in no Coal the same as in woody fibre. Woody fibre is about one-half Carbon, in Coal the percentage of Carbon ranges from 67 to 95 per cent., so that Anthracite may be said to be all Carbon. But if we run through the table again we shall note that the departure from the composition of woody fibre is gradual; we pass from one term in the series to the next by a decrease in the percentage 182 Geology. of Oxygen and Nitrogen, and a corresponding increase in the percentage of Carbon. Then the thought strikes us that all these forms of Coal may have been produced by a process which step by step removed the Oxygen and Nitrogen out of wood till scarcely anything but Carbon was left. Chemical investigation proves that the decomposition of wood begins in this very way. The rotting of wood gives rise to a substance of not very definite character known as Humus, whose average composition is shown in the table. Humus contains a somewhat larger proportion of Carbon than wood, and less Oxygen and Nitrogen. Peat furnishes the next step, and here the change in composition is in the same direction. And so we go on step by step till we reach the final term in Anthracite. But it must not be supposed that the transformation of wood into Coal is exactly analogous to the formation of Humus out of wood. When wood decays in the open air, there is a free supply of Oxygen, and Carbon Dioxide and Water are produced. Coal was formed under circumstances which allowed only of a very limited access of Oxygen, and only part of the Carbon lost was converted into Carbon Dioxide, the rest combined with Hydrogen and formed Marsh gas, which now exists in the body of the Coal under the too well-known name of Fire- damp. In the following table the relative composition of Wood and Coal is shown in a different fashion. The Carbon in each variety is taken at the same amount, viz. 100. This table shows like the former the gradual decrease in the Oxygen and Nitrogen, and it also shows that though the Hydrogen appears to remain constant through nearly all the stages in the first table, this is not really the case, and that there is a gradual elimination of this element as well. The gradual increase in weight is mainly due to the fact that Coal has suffered compression during and after the transformation. Weight of one Cubic Foot in Ibs. Carbon. Hydrogen. Oxygen and Nitrogen. Wood 30 100 12-3 86-8 Peat 50 100 97 547 Lignite .... 70 100 8-3 40-0 Brown Coal 75 100 7'4 29-7 "Bituminous" Coal . 80 100 6-4 13-4 Anthracite .... 90 100 2-6 2-3 Non-crystalline Rocks. 183 The inference that Coal is mineralized vegetable matter which a study of its chemical composition leads us to is amply confirmed by the second line of argument. This is based on the fact that vegetable tissue and other remains of plants can be detected in Coal with or without the aid of the microscope. On such general grounds the vegetable origin of Coal has been for a long time universally admitted. Researches of late years have enabled us to go further, and to say what the plants were that furnished the material for certain Coals and what portions of those plants enter most largely into its composition. Some Coals are made up almost entirely of the spores and spore- cases of plants closely allied to the modern Club-mosses. In the cryptogarnic or flowerless plants multiplication is effected by bodies called Spores, which correspond, as far as their ultimate products are concerned, to the seeds of flowering plants. In some cases it is known that there are two kinds of spores, microspores or little spores, and macrospores or large spores, the first producing the fertilizing matter, and the second developing ovules or germs. In the common Club-moss one kind of spores only has been observed, and its mode of repro- duction is not understood. The spores are contained in bags called Sporangia or Spore-cases. In some Club-mosses and Horse-tails the sporangia are placed within cones or spikes, consisting of scales or leaves overlapping each other, and the sporangia are lodged in the spaces between the scales. Now among the commonest of the fossils found in the strata among which Coal occurs is one that goes by the name of Lepidostrobus. In external appearance it resembles strongly the spikes of the modern Club-moss. Dr. Hooker* obtained specimens of these cones with the internal structure preserved, and showed that they consisted of scales supporting sporangia, which contained spores marked with a three-rayed ridge on their under side. In the arrange- ment of the scales, the attachment of the sporangia, and the shape and markings of the spores, these cones correspond with those of the Club-moss. Dr. Eobert Brown t afterwards described a fossil cone called by him Triplosporites, which agrees with a modern lycopodiaceous plant, Selaginella, in containing both large and small spores, the microspores being found in both genera on the middle and upper scales of the cone, and the macrospores on those of the lower portion. Mr. CarruthersJ has since examined another fossil cone, which he has named Flemingites, the sporangia of which show a three-rayed marking on their under side, and agree with those of Lepidostrobus in containing only microspores. Now, as far back as 1840, Professor Morris figured some small bodies found in the Coal of Coalbrook Dale, of the nature of which he was at that time unaware. These bodies agree so exactly in shape, * Memoirs of the Geological Survey of England and Wales, vol. ii. part 2, p. 440 (1848). t Transactions Linnean Society, xx. 469 (1851). J Geol. Mag. ii. 431; vi. 151, 289. In Prof. Prestwick's Paper on the Geology of Coalbrook Dale, Transactions Geol. Soc., 2nd series, vol. v. plate xxxviii. figs. 8-11. 1 84 Geology. size, and the three-rayed marking with the spores or spore-cases detected by Hooker, Brown, and Carruthers in the cones just de- scribed that there can be no doubt that they have been shed from one or other of them, and the Coal in question is certainly made up in part at least of the spores of a lycopodiaceous plant. Similar bodies have been observed by Professors Balfour and Huxley, Mr. Binney, and others in other beds of Coal, and in some cases, the Better Bed of Bradford for instance, the seam is almost entirely made up of them. The plant to which Lepidostrobus belonged is extremely common in the beds associated with Coal ; it is called Lepidodendron, and speci- mens of it with the cones attached to the branches are by no means uncommon. We can now then go further than the general statement that Coal is of vegetable origin ; we know that among the plants which contributed to its formation one of the commonest was a close ally of our present Club-moss, and that, in some cases at least, it was the spores of that plant that furnished nearly all the material of the fossil fuel. It is worthy of note that the spores of the Club-moss are so highly inflammable that they are eminently suited to give rise to a combustible substance like Coal. They are also very resinous, and this w T ould prevent their being wetted, and they would thus be less liable to decay than the other parts of the plant. The amount of spores necessary to form a seam of Coal is so enor- mous that some little hesitation may be felt at first in accepting the view that some Coals are made up of little else but these minute bodies. Large accumulations however of vegetable matter of a similar character have been observed in recent times. Dr. John Davy describes a shower of a " sulphurous substance " in Inverness-shire in 1858.* The " sulphurous substance " was found to be the pollen of the Fir (Pinus sylvestris) ; it lay in some places to a depth of half an inch, and was noticed at points thirty-three miles apart. Sir John Richard- son informed Dr. Davy that the surface of the great lakes in Canada is not unfrequeritly covered with a scum of the same pollen. Similar occurrences have been observed in the forests of Norway and Lithuania. In other instances different portions of other plants have furnished the materials out of which Coal was manufactured. In many cases the bedding planes of Coal are sprinkled over with patches of a dull black, soft, fibrous substance that looks like Charcoal, which is called Mineral Charcoal or Mother of Coal. In the case of certain Nova Scotian Coals Dr. Dawson thinks that this is all that remains of the woody portion of the trees that furnished the material for the Coal ; the purer parts of the Coal, he thinks, are mainly composed of the bark. He relies very much on the fact that bark is far more lasting than woody fibre, and that when trees are found in a fossil state in the rocks associated with Coal, the bark has been turned into Coal, while the interior has rotted away and been filled in with sand and mud. What is believed to be mineralized bark enters also largely into the composition of many Scotch and English Coals. * Proceedings Royal Society, Edinburgh, vol. iv. p. 157 (1859). I am indebted to ray friend Mr. L. C. Miall for calling my attention to this paper, and for other valuable assistance in connection with the subject in hand. Non-crystalline Rocks. 185 In connection with this subject the reader may further consult the papers mentioned in the footnote below.* We will in a subsequent chapter explain how the materials of Coal were collected together and brought into their present shape. The characters of the principal varieties of Coal are as follows. Lignites consist of a matted mass of stems and branches of plants, still retaining their woody fibre and only partially mineralized. They have a low heating power, usually make a good deal of ash, and some- times give off an offensive odour when burning. Brown Coal closely resembles Lignite in many respects, but mine- ralization has proceeded so far that no woody texture can be recognised by the unaided eye. The ordinary Coals used for household purposes vary much in char- acter. Some varieties, known as Caking Coal, fuse into a pasty mass as they burn, and require frequent poking to keep them alight. Others burn without caking. The amount of ash too is very variable ; some Coals choke up the fireplace, others, like the Kilburn Coal of Derby- shire, may be burned the day through and not leave a teaspoonful of ash. A very beautiful variety, known as Cherry Coal in Scotland and Branch Coal in Yorkshire, has a shiny resinous lustre, lights readily, burns cheerfully, and leaves little ash. The above-named and other similar varieties of Coal are usually classed together as " Bituminous :" the term is not chemically correct, for though the Coals contain the constituents of Bitumen, they do not contain Bitumen itself. On the other hand, Anthracite, which is nearly pure Carbon, is described as Non-bituminous. The Coals called Splint , Hard, or Steam Coal are intermediate in composition and pro- perties between " Bituminous " Coals and Anthracite. They are more difficult to light, but have a greater heating power than " Bituminous " Coal. They are of great value for locomotives and marine engines. Some, like the Barnsley Steam Coal, consist of thin semi-anthracitic layers alternating with others of a more " Bituminous " character. A nthracite is heavier, harder, and has a more thoroughly mineralized look than " Bituminous " Coal, qualities well expressed by its popular name, Stone Coal. It rarely soils the fingers, has very frequently a sharp conchoidal fracture, and a brilliant lustre. Other varieties are dull, or break into small cubical lumps. It is difficult to light, but when ignited gives out intense heat, and burns without flame and with little smoke. Cannel Coal differs in many respects from the preceding varieties. The best examples are compact, have a shining lustre and a marked conchoidal fracture, and do not soil the fingers. Cannel Coal is of special value for gas-making : the yield and the illuminating power of * Dawson, Quart. Journ. Geol. Soc. ii. 132, x. 1, xv. 477, 626, xxii. 95 ; Annals of Nat. History, 1871, p. 321 ; Silliman's Journal, Ap. 1871 ; Acadian Geology, 2nd ed. pp. 188, 461, 493. Quekett, Quart. Journ. Microscopical Soc. No. 6, p. 43 ; Transactions Microscopical Soc. ii. 34. Bennett, Transactions Royal Soc. of Edinburgh, xxi. pt. i. p. 173. Balfour, ditto, p. 187. Huxley, Critiques and Addresses, p. 92. Williamson, Macmillan's Magazine, xxix. 404. Binney, Manchester Lit. and Phil. Soc. March 1874. Bischoff, Chem. Geology, i. 258.' Coal, Its History and Uses. 1 86 Geology. the gas manufactured from it are greater than in the case of any other coal. The high illuminating power of " Cannel-gas " is largely due to the fact that Ethane (C 2 H 6 ), which is one of the most important illuminat- ing factors in coal-gas, occurs occluded in the pores of Cannel Coal. On account of the large amount of gas which it gives off Cannel Coal will burn like a candle, whence its name, which is a corruption of Candle Coal. There are however all sorts of inferior varieties, and from the most impure of these a gradual passage often takes place into very carbonaceous black Shale. These imperfect Cannels are called in some parts of England Stone Coal, a term applied in other parts to Anthracite. * The average composition of Cannel Coal is Carbon ....... 847 Hydrogen . . . . . . 6 '4 Oxygen and Nitrogen . . . . . 8 '9 Graphite, Plumbago, or Blacklead occurs as an accessory consti- tuent of Granite, Gneiss, and other rocks, in veins or pockets, and occasionally in a state of approximate purity in beds. It consists of Carbon with about five per cent. of impurities, such as Silica, Alumina, and Oxide of Iron. There is good reason in many cases to believe that it is only an extreme form of Anthracite, that is, it is a Coal from which the gaseous elements have been completely withdrawn. But there are other cases in which the origin of Graphite is still a matter of uncertainty. Other Rocks conveniently placed in the Non-crys- talline Class. Rock-salt. Common Salt (Na 2 Cl) occurs in beds and lenticular- shaped masses often of great size. It contains small admixtures of Magnesium, Calcium, and less frequently Potassium Chlorides, and of Calcium, Magnesium, and Sodium Sulphates. It is also frequently mixed with Clay, Ferric Oxide, and Bituminous matters, which stain it various colours. Ammonium Chloride and compounds of Bromine and Iodine have also been observed in it. Some Rock-salt contains cavities full of gas, which escapes with a crackling noise when the rock is dissolved in water or heated ; these are called by the Germans Knistersalz (Crackling Salt). Some other minerals associated in certain cases with Rock-salt will be mentioned when we come to speak of the way in which that rock was formed. Gypsum and Anhydrite. These minerals occur both in beds and in masses of irregular shape. Gypsum often puts on a fibrous form, the threads being perpendicular to the planes of bedding. It is also met with in coarsely-granular and very finely-grained forms. Gypsum is often rendered very impure by mixtures of Clay, Ferric Oxide, and other substances. Anhydrite is coarsely or finely granular. There are a few other minerals which rarely occur in sufficient quantity to form rock-masses. For an account of them the reader may refer to Zirkel's " Lehrbuch der Petrographie." * For fuller details of the different varieties of Coal, see Crookes and Eohrig, Metallurgy, iii. 413-470; Percy, Metallurgy, i., chap, on Fuel. Denuding Agents and how they work. 1 87 SECTION III. DENUDING AGENTS AND HOW THEY WORK. Example of the Determination of the Origin of a Rock. We may now inquire into the way in which the rocks we have just described were formed, and to make a beginning here is an instance of the manner in which we ferret out the steps by which a particular rock has been formed. Here is a bit of coarse Gritstone, and side by side with it let us lay a piece of coarsely-grained Granite. The two are singularly alike, and in both we can readily distinguish the three minerals Felspar, Quartz, and Mica. The specimens might well be supposed by a casual observer to have been broken off the same crag, and if the two tors from which they were taken are viewed from a little distance, they are so similar in outline and look that any one might be pardoned for supposing they were made of the same rock. But if the blocks be scanned a little more closely, specially if they be examined with a pocket lens, an important difference will be detected between them. In the Granite there are crystals with their angles pointed and their edges sharp ; in the Grit- stone, though the crystals may not be much altered, yet a certain amount of rounding off has taken place in both angles and edges. Also in the Granite the Quartz forms a kind of paste in which the other mine- rals are embedded. In the Gritstone the Quartz is broken up into grains. There can be no question that the Gritstone has been formed by the breaking up of a rock identical with the Granite specimen before us, and that .in the process the crystals have lost something of their sharpness of outline. Our next question will be, What has done this? Let us visit the rocky tor from which our specimen of Granite was taken, and we shall not have to wait long for an answer.* The outer surface of the rock is evidently crumbling away, parts readily fall off in a coarse powder, the grains of which are crystals rounded in just the same way as in the Gritstone before us : large quantities of a similar powder are spread round the base of the tor, and fill cracks or hollows on its surface. The Gritstone holds together a little more firmly than this powder, but otherwise there is no difference between the two, and the conclusion is irresistible that the former is nothing else than a quan- tity of the latter that has been in some way or other bound together into a moderately firm rock. If now we turn over in our minds what is constantly happening to this Granite tor, we shall readily understand how it is that it is crumbling away. Rain beats upon it, and has power to decompose and dissolve part of the cement by which it is held together ; the water also, as it streams off the rock, washes over it the coarse grains lying on its surface, and these grind it away like emery or a file ; the wind drives against it the same wearing implements with the same effect ; frost expands water contained in chinks or crevices, and forces off fragments : these and similar agents are incessantly at work, and by these the crag is being broken up and the heap of debris which * It will be noted here that though mere lithological examination will in this case help us a little towards learning the way in which the Gritstone has been formed, it is only by outdoor study that we can realize all the steps in the process. 1 88 Geology. surrounds and partly buries it has been formed, and is constantly being added to. By methods like those just described the materials out of which our Gritstone is built up were obtained : but the history of its formation is not yet complete. If we visit the quarry from which our specimen was taken, we shall find the rock arranged in layers, beds, or strata, each layer being marked off by a clear plane of division from that above and below it, and likely .enough we may find between some of the beds layers of Clay, Limestone, or other rocks. No such bedded arrangement exists in the debris that surrounds the Granite tor ; and therefore, though the materials for making a Gritstone are certainly to be found there, they must undergo some further process of arrangement before they could give rise to a rock like that of which our specimen formed a part. Not far from the foot of the Granite tor there runs a rivulet, and in fine weather it is clear enough to allow us to count the stones on the bottom. But after heavy rains the water pours down thick and turbid : fill a glass from the brook and let it stand : the water soon becomes clear, and a quantity of sand and mud, which was the cause of the turbidity, settles to the bottom : take out some of the settlings and spread it in the sun to dry, it is nothing else but some of the finer part of the rubbish around the tor produced by the weathering of the Granite. It has been washed down the sloping bank by rain, and is being swept forward by the current. Now return to the rivulet ; by a little atten- tive listening we can detect amid the rushing of the torrent a harsh, grating sound rising from beneath the water ; this is caused by the coarser parts of the rubbish which have been swept down into the stream, but which, being too heavy to be carried suspended in the water, are being pushed and rolled along the pebbly bottom. Thus the ruins of the Granite, coarse and fine alike, are being always carried forward ; and when the rivulet falls into a larger brook and this again into a river, the waste matters travel on along with the contributions of the other feeders. The journey goes on till the river enters the still waters of a lake or of the sea : the stream then loses its velocity, and therefore its power to carry any further ; and the matters which it has brought so far all sink, sooner or later, to the bottom, and are spread out in layers approximately horizontal. In the intervals between two floods each layer will have time to harden a little before the next layer is placed upon it, and so a plane of separation between the two will be produced, and the deposit will have a bedded or stratified struc- ture given to it. Further the velocity, which is sufficient to sweep along fine mud, is not able to move coarse sand, and hence at one time the former alone, and at others the two together, will be laid down beneath the still water. In this way the deposit will come to have in it layers of different degrees of coarseness and different composition. Again the heavy materials can at no time be carried so far out into the lake as the lighter and finer ; and thus the stuff brought down will be, so to speak, sorted, coarse deposits will prevail near the mouth of the river, and will thin away in a wedge-shaped form, and be replaced by finer beds as we advance towards deeper water. All this is found to Denuding Agents and how they work. 1 89 be the case when we drain lakes, and cut into the accumulations which have been formed beneath their waters. The stratified structure of the Gritstone mass, of which our specimen formed a part, is so exactly similar to that of deposits forming now- adays beneath still water in the manner just described, that we have no hesitation in ascribing it to the same cause which is giving them their bedded arrangement, and every step in the history of its formation is now clear to us. Its materials were furnished by the atmospheric wear and tear of Granite : they were washed by rain down hill-slopes into running streams, carried forward during floods into bodies of still water, and there they came to rest in layers laid one over the other by successive freshets : the bands of clay lying between the different beds of Gritstone were formed in the same way when the streams were lower and had not power to carry anything heavier than tine mud. Rocks like this Gritstone, because they are made up of broken pieces of pre-existing rocks, are sometimes spoken of as Clastic (KAao-r'o?, broken) ; and because they have been formed under water they go by the name of Subaqueous. Because they have been derived from pre-exist- ing rocks they are also called Derivative. There are other Derivative rocks which have been formed not by the mechanical wear and tear of pre-existing rocks, but by the chemical decomposition of their con- stituents ; these are sometimes called Dialytic. Repeated observations similar to that just described bring home to us the conviction that the solid matters which form the surface of the earth are constantly undergoing wear and tear ; that the loose rubbish thus produced is being incessantly conveyed by the agency of rain and running water from higher to lower levels, till at last it comes to rest beneath large bodies of still water, where it is spread out in layers approximately horizontal; and that it is in this way that a large part of the bedded rocks of the earth's crust have arisen. Denudation. The process by which the surface of the ground is broken up and its ruins carried away is known as Denudation, and the agencies by which this is effected as Denuding Agents. A thorough grasp of the way in which Denudation works lies at the root of all sound geological knowledge, and we will devote the rest of the present chapter to this subject. Enumeration of Denuding Agents. Denuding Agencies may be classed under the following heads. 1. Rain. 2. Running Water, whether above or below ground. 3. Frost and Frozen Water. 4. Wind. 5. Animal and Vegetable agencies. 6. The Sea. The first five are generally classed together as Subaerial, Atmospheric, or Meteoric Denuding agents. The denudation wrought by the sea is distinguished as Marine. The Work of Denuding Agents. We will now look at the way in which each of the denuding agents just enumerated works, and the results it produces. We shall have to consider each first as a producer ', secondly as a car- rier of waste. 190 Geology. 1. RAIN. Rain acts on the surface of the ground in two ways, Mechanically and Chemically. Of the water which falls upon the earth from the clouds part rises again into the air by evaporation, or is taken up by plants ; part streams over the surface, and is at last collected into brooks ; part sinks into the ground, and, after pursuing for a longer or shorter distance an underground course, rises again in springs. We have here to deal with that part which flows over the surface before it becomes gathered into a definite channel. Mechanical Action of Rain. Water during this portion of its course exerts an important mechanical effect as a carrying agent : any loose surface matter produced by the decomposition of the rock beneath is swept on by it, and advanced a step forwards on its road to the rivulet, which will at last receive both it and the agent which moves it. At the same time, as these loose materials roll over the ground, they still further abrade and wear away the surface. The power of a gentle shower to move fine mud may be seen any rainy day either in a ploughed field or by a roadside : heavy storms, even in temperate climates, carry far coarser materials than most people, are aware of, specially if the slope of the ground be steep. I recollect well enough having to leave a rock, which was affording me some small shelter during a thunderstorm in the centre of England, on account of the shower of stones which the rain washed over the edge, preferring the certainty of being wet through to the chance of having my head broken. In the tropics, where not only is the rainfall very large in amount, but also where an enormous quantity comes down in a very short time, the carrying and wearing effects become enormously increased. We must also recollect that rain, besides acting as a carrier of loose matters which it finds ready to hand, softens many rocks, such as Clay, and so renders them an easier prey to showers that come after. The earth pillars of the Tyrol furnish an excellent instance of a case of denudation on a large scale, which can have been produced only by the action of rain. There is a very full account of them in Lyell's " Principles," 10th ed. vol. i. p. 335. Chemical Action of Rain. Rain also has the power of acting chemically on certain rocks, and carrying away some of their consti- tuents in solution. The rock most largely attacked in this way is Limestone. There is a gas popularly known as Carbonic Acid, but which is styled by chemists Carbon Dioxide (C0 2 ). A solution of Carbon Dioxide in water is supposed to form a weak acid, Carbonic Acid (H 2 C0 3 ). Carbonic Acid, or a mixture of Carbonic Acid and water, has the power of dissolving Calcium Carbonate (CaC0 3 ), or Carbonate of Lime, as it is often called. Now Carbonate of Lime, it will be remembered, is the chief constituent of Limestone, and whenever water impregnated with Carbonic Acid comes in contact with Lime- stone, the Calcium Carbonate is dissolved out and carried away in solution. Almost all water on the earth's surface contains more or Denuding Agents and how they work. 191 less Carbonic Acid ; Carbon Dioxide exists in small quantity in the air, and the rain, as it falls, takes up some, and so becomes mixed with Carbonic Acid; the same result is produced when rain-water comes in contact with decaying vegetable matter as it flows over the surface of the ground. Almost all surface water then has the power of attacking Limestone, the dissolving away of that rock goes on incessantly, and the amount which is thus slowly and insensibly carried away becomes in time very considerable indeed. It is in this way that the caverns and underground watercourses, with which all Limestone countries abound, are formed. It is curious on taking up a good topographical map of certain districts to note that there is a line on reaching which all the streams suddenly cease. This line marks the boundary between a tract of Limestone and some other rock insoluble in water : over the latter the water runs in brooks, but on reaching the former it has by degrees dissolved away the rock and eaten out underground channels, into which it sinks and flows away out of sight. It is for this reason too that Limestone districts abound with funnel-shaped cavities, descending from the surface vertically into the rock, into which water sinks and disappears. They are often called Swallow-holes or Swallows. Wherever there was any little depression in which water could lodge, the bottom was eaten away lower and lower, and a pipe formed at last leading from the surface into an underground channel. Districts composed of very pure Lime- stone show bare rock up to the surface, because the rock is entirely soluble ; when however the Limestone contains insoluble impurities, these remain behind, and give us the means of forming a rough estimate of how much has been removed. We have a good instance of this in the south of England. The surface of the Chalk Downs there is often covered, as shown in fig. 78, with a red Clay full of Fig. 78. CLAY-WITH-FLINTS IIESTING ON CHALK. a. Clay-with-Flints. fc. Chalk. Flints, known as Clay-with-Flints, the origin of which is as follows.* Chalk contains from 94 to 98 per cent, of Carbonate of Lime, mixed with from 2 to 6 per cent, of clayey and earthy matters ; it has also embedded in it many nodules of Flint. The two last are practically insoluble, and therefore remain behind when the Carbonate of Lime is dissolved and carried away by percolating rain-water. The Clay- with-Flints often reaches a thickness of many feet, and testifies to the large extent to which the rock has been insensibly dissolved away, f In the same way many parts of Palestine [{: are. thickly strewn with * Geology of Parts of Middlesex, Herts, Bucks, Berks, and Surrey (Memoirs of the Geological Survey of England), p. 63. t See also Bischoff, Chemical Geology, iii. 194. % Dictionary of the Bible, Art. "Palestine." 192 Geology. loose lumps of Flint, which have remained behind, while the Lime- stone, in which they were originally contained, has been carried off in solution. There is scarcely any of the manifold denuding processes which is of such importance and so constantly brought before the notice of the geologist as the dissolving away of Limestone by Car- bonated Water. In some cases, where Limestones contain a large admixture of siliceous matters, a sort of skeleton of the latter remains behind when the Carbonate of Lime is dissolved out, forming what is known as Rottenstone. Another important decomposition effected by atmospheric water is that of Felspars and Felspathic minerals which we have already seen results in the production of Kaolin, and furnishes an important ingredient of many clayey rocks. The formation of Kaolin goes on naturally to a large extent in many Granite districts, Cornwall for instance,* and deposits of the Clay are formed in hollows or flats, to which it has been carried by running water : such deposits contain grains of Quartz, imperfectly decomposed fragments of Felspar, scales of Mica, and other impurities, which are separated by washing. The Oxygen of the air contained in rain-water also enables it to oxidize, or raise the degree of oxidation of some constituents of rocks. This happens most largely to the ferruginous constituents, and the process has been explained on p. 154. Of the substances acted on chemically by rain-water the one most largely dissolved is Carbonate of Lime, partly because it is so readily soluble, and partly because Limestone is a rock so universally diffused. This being the case, it certainly seems strange that this salt can scarcely be detected at all in solution in sea-water ; Sulphate of Lime and Magnesian Salts, which have doubtless come there in the same way as Carbonate of Lime, we do find in sensible quantities, but the last only in the smallest amount, or not at all. These apparently contradictory facts are capable of easy explanation, as we shall see when we come to look at the formation of Calcareous rocks. Besides the substances mentioned, rain-water dissolves and carries away in solution others less common, as Rock-salt, Sulphate of Lime, Sulphate of Magnesia, and under certain circumstances Silica. 2. RUNNING WATER. Rivers as Carriers of Sediment. As the portion of the rain that streams over the ground becomes gathered into definite chan- nels, it brings into the brooks so formed the matters which it has swept mechanically along with it, or which it holds in solution, and the first function which running streams perform is to carry these on in their downward course. In this way alone streams and rivers are most important auxiliaries in the work of denudation, they prevent accumulations of debris from * De la Beche, Report on the Geology of Cornwall, Devon, and West Somerset, p. 509. The Hensbarrow Granite District and the China Clay Trade ; J. H. Collins. Denuding Agents and how they ^vork, 193 acting as a shield against the action of denuding agents, and allow ;i bare surface to be always maintained for the latter to work upon. We are apt at first sight to underestimate the carrying power of running waters, and to take notice only of the light matters which float on the surface, overlooking the far more important burden of fine mud they hold in suspension, the matters carried down in solution, and all they move forward by pushing them along the bottom. It is only when the amount of matter carried by rivers is subjected to actual measurement that we come to realize how large it really is. To take two instances, it has been determined that the Mississippi carries 7,459,267,200 cubic feet of sediment every year into the sea; and the Rhone 600,381,800 : the first of these quantities would cover a square mile of ground to a depth of 268 feet.* We must also recollect that since the specific gravity of rocks lies between 2 and 3, they lose from a half to a third of their weight in water. During floods the carrying power of rivers becomes very much increased, but I believe very few people are aware how enormous the increase is, unless actual instances of the work done by violent rushes of water happen to have come under their notice, t Here then are the details of a couple of actual cases. In 1866 twenty inches of rain more than falls in many places in England in the course of a year fell in Scinde in twenty-four hours, and the Mulleer river rose in consequence to an unusual height. The valley was crossed sixteen miles above Kurrachee by a bridge con- structed of wrought-iron girders. The flood banked up wood and grass against the bridge, and at last threw it over, and one of the girders, weighing eighty tons, was carried two miles down the river and buried in sand. It is probable that in this case the accumulation of drift- wood served in some measure to buoy up the girder, but even allowing for this, the transporting power of the current must have been astonishing. J In 1864 a frightful flood was caused by the bursting of a reservoir above Sheffield. The rush of water was most violent, for it was esti- mated that 40,000 cubic feet passed along the narrower part of the valley per second. The official report states that 92,000 cubic yards of the embankment were swept away in less than half an hour, and mentions one stone weighing thirty tons which was moved; I saw myself a stone of about two tons, which I could identify by its shape as having formed part of a weir more than a hundred yards up the valley. Whole acres of meadow-land were deeply buried beneath heaps of debris, consisting mainly of large angular blocks of rock, which the torrent had torn off from the banks as it rushed along. Some small farmhouses which stood across its path were sliced in two * The student will do well to consult the admirable and exhaustive treatment of this subject by Professor Geikie, in Jukes' Students' Manual of Geology, 3rd ed. pp. 420-429, and in Trans. Geol. Soc. of Glasgow, iii. (1868) 153; or Croll, Climate and Time, chap. xx. + Mr. Hopkins states that the weight which a current of water can move in- creases as the sixth power of the velocity. Quart. Journ. Geol. Soc. of London, vol. viii., Presidential Address, p. 27. Quart. Journ. Geol. Soc. of London, vol. xxiv. p. 124. H 1 94 Geology. as neatly as if they had been cut through with a knife, one half carried away and the other left standing. At sharp bends in the valley where the water had impinged on projecting spurs of the bank, or where it had been driven into a recess, it had excavated in the solid sandstone rock large hollows, which any one who was not aware of the circumstances of the case would have supposed to be quarries. Besides mechanically-formed sediment, rivers also carry away large quantities of matter in solution which has been brought into them by rain or spring water, or dissolved out in their passage over soluble rocks. Thus the waters of the Nile contain 14 parts in 100,000, those of the Rhone 17, those of the Main 24, and those of the Thames 40 of matter in solution. Denudation wrought by Rivers directly. But besides acting as the bearers of matter brought into them, streams take also a direct part in the work of denudation. Running water has of itself little or no power to abrade rocks, ex- cept in so far as it may in some cases soften them and destroy their coherency by soaking into them; but the sediment with which all streams are charged enables them to effect a very large amount of destruction. This wears away the banks as it passes, and portions from time to time become undermined and topple over into the current, there to be ground fine and in the end swept away. The process may be seen going on even in rivers which flow peacefully through com- paratively flat districts ; and in more rugged tracts, where the stream runs at the foot of a lofty clilf, the amount brought down by each fall is proportionately increased. The undercutting will evidently go on faster if the base of the cliff consists of a rock softer than that on the summit, or if there be springs bursting out on its face. The bed of the stream is at the same time ground away by the rough materials rolled over it. Thus rivers are always performing a twofold w r ork ; they sweep along debris brought into them by rain, and this enables them to wear away their banks and beds, and to grind small the masses detached by its action, while it is itself at the same time still further comminuted, and rendered capable of being carried more easily and to longer dis- tances. Daubre"e has made a series of very instructive experiments which illustrate the way in which rocks are ground down by the action of running water.* He placed his materials with water in a cylinder which could be rotated at a known velocity about a horizontal axis. In this way he reduced the fragments to a mixture of pebbles, sand, and mud. When operating on Granite the sand grains were as a rule very slightly rounded or even angular. The explanation given depends upon the fact that rounding goes on only so long as the fragments are rolled along the bottom; when they become small enough to be carried in suspension they no longer grind one another. Now the quartz of Granite being a good deal fissured and brittle, is quickly broken up by the mutual impact of the blocks into fragments small enough to * Geologie Experimentale, i. p. 248, et seq. Denuding Agents and how they work. 195 swim along with the stream, and hence comparatively little rounding is produced.* When Orthoclase was treated in the same way in pure water, a certain amount of decomposition was effected, for the water contained in solution Potash and to a smaller extent Soda and Alumina. In carbonated water and in lime-water the decomposition went on to a larger extent than in pure water ; the presence of common salt checked it. The product most largely obtained was that we have described as Felspathic Mud. The fact that some decomposition took place how- ever shows that the process by which Kaolin is produced had at least begun, and we can understand by the light of these experiments how the two main elements of the clayey rocks have been formed by the mechanical action of running water on Felspathic minerals. Underground Streams. We will next turn our attention to the water that circulates underground. In the case of rocks not acted upon chemically by rain, this finds its way down through cracks, or between the beds, or, in the case of a very open porous rock like un- consolidated sand, through the body of the rock itself. If its down- ward course is stopped by reaching a bed through which it cannot force its way, it flows along the top of this bed, and escapes, when the bed comes to the surface, either in springs or by a general oozing out above the outbreak of the impervious stratum. Or it may be that the cracks by which it is descending become so narrow that friction against their sides seriously impedes its further progress ; if in this case it meet with a wider fissure opening out upwards, it may be easier for it to be forced up this by hydraulic pressure than to continue to descend by gravity, and then it will mount up and issue as a spring.t The natural pipes which feed springs of this class will not generally have a very large bore ; but in the case of rocks which are chemically acted on by rain, there is scarcely any limit to the size of the under- ground channels which water makes for itself. Among the widely- diffused rocks Limestone is the one most readily soluble, and in it accordingly are these underground watercourses most frequently met with : the water bursts out of them, not as a spring, but as a full- grown brook ; and they sometimes swallow up, and after a time dis- charge again, the contents of good-sized rivers. It is scarcely necessary to give instances, but we may mention the Holy Well, at the town of that name in Flintshire, which was estimated by Pennant to discharge twenty-one tons of water in a minute. The fable of " Divine Alpheus, who by secret sluice Stole under seas to meet his Arethuse," was evidently based on a knowledge of the facts we have been describ- ing, examples of which are extremely common in the calcareous dis- tricts of Greece. Underground streams, provided their course is throughout down- * Dr. Sorby has noticed the angularity of the grains of sand in many Grit- stones. Address before the Royal Microscopical Society, 1877. t Geikie, Primer of Physical Geography, p. 46, lig. 6. Comus. 196 Geology, wards, may and do produce and convey mechanically-formed sediment, just as rivers above ground, but the amount of it will obviously be small. Their principal share in the work of denudation is dissolving and carrying away in solution anything they can act upon chemically, and the amount removed in this way, so to speak invisibly, is very large indeed. In volcanic districts, or where springs descend to a great depth, their waters become heated and impregnated with alkaline solutions, and are then able to dissolve Silica and other matters, which otherwise they would not be able to attack so easily. The increase of pressure at great depths also allows water to become more largely carbonated, and otherwise increases its dissolving power. Water accumulating below ground assists in another way in bringing about denudation of the surface. When large quantities of soluble rocks, such as Limestone or Rock-salt, have been dissolved away, the ground above falls in, and thus new channels are formed for rivers to run in, and carry on in their own way the wearing away of the surface. Thus many of the depressions, in which the lakes called " Meres " in Cheshire lie, have probably been formed by the sinking of ground beneath which thick masses of Rock-salt have been dissolved away ; * and many of the " Dales " of Derbyshire and other Limestone districts have all the look of having been once caverns, the roofs of which have fallen in.t 3. FROST AND ICE. We now come to the denuding effects of water in its solid shape as ice. If water be gradually cooled, it contracts as the temperature de- creases till 39 Fahrenheit or 4 Centigrade is reached ; it then begins to expand and continues expanding till it is converted into ice at 32' Fahrenheit or Centigrade. Frozen "Water. Just at the point when water is becoming solid the expansion is very rapid, and the efforts of the molecules to get further apart are so exceedingly powerful that if the water be shut up in a close vessel they rend the latter open, even though it be formed of iron half an inch thick. Just the same result follows when water which has soaked into the cracks and crevices of a rock freezes. The expansive force tears the rock open, forces off pieces from it, and throws them down to be worked small by rain and other denuding agents. The amount of ruin wrought in this way will evidently be very con- siderable. All rocks admit water, and wherever frost occurs, it becomes one of the most powerful agents in their destruction.^ Glaciers. We have already seen how important a denuding agent * Ormerod, Quart. Journ. Geol. Soc. iv. 262. Recent experience shows that where salt has been removed in large quantities by pumping it up dissolved in water as in Cheshire, there is great risk that the sites of some of the most nourish- ing towns will before long be occupied by Meres. T The Geology of North Derbyshire (Memoirs of the Geological Survey of England), p. 2. See Geol. Mag. vol. vi. p. 491, for a good instance. Denuding Agents and Jiow tJiey work. 197 water is as it flows over the surface in its liquid state ; in those cold regions where water ean exist only in a solid condition, the place of streams and rivers is taken by Ice Rivers or Glaciers, and these also do their share of denuding and transporting work. There is a certain line, called the limit of perpetual snow, whose height in the tropics is some 15,000 or 16,000 feet above the sea, and which gets lower and lower as we go northwards or southwards, till at last it comes down to the sea-level. Above this line the temperature never rises for long together above the freezing-point, and all the moisture which falls from the sky comes down, not in the shape of rain, but as snow. On a tableland which rises above this limit snow alone will fall ; very little of it is ever melted, and the amount that disappears by evaporation is far less than is supplied from the atmosphere ; and so layer after layer will be added, and the pile will be always growing in thickness. The snow thus heaped up is compacted into ice in various ways : the weight of the mass forces the air out from, between the crevices of the snow- Hakes and binds them together ; and the water, which the thawing of the surface by the midday sun produces, trickles down into cracks and crevices, and becomes frozen there when night comes on. In this way in such situations enormous heaps of snow and ice arise. It would seem at first sight that under these circumstances the ice-heap must increase in thickness every year, and that in consequence the tableland on which it rests will get higher and higher as time goes on ; but this is not the case, snow-capped mountains and ice-clad tablelands retain the same elevation in spite of the constant additions to their covering, and there must therefore be some means by which ice is carried away from them as fast as it is being added above. Now if ice were a body rigid like glass, this increase in the height would take place wherever the ground rises above the snow-line, and there would be scarcely any limit to the depth of the accumulations which would be formed in such .situations. But though to look at it any one might well suppose that ice has as little power to change its shape, bend, or mould itself as glass, such is really far from being the case. Under a sufficient amount of pressure ice can be forced into new forms almost as readily as moist clay or dough, though the amount of pressure required to mould ice is far greater than suffices for the modelling of these evidently plastic materials, and the change of shape does not take place exactly in the same manner-in the two cases.* Now suppose we laid clay on a table with a slight bulge upwards towards the middle, from which boards sloped down to the floor, and kept adding to it above, what would happen 1 For a time we might go on adding to the heap without pro- ducing any effect, but as we kept putting more and more on, the weight of the upper part would squeeze out some of the clay below, and at last force it over the edge of the table and down the boards, and as long as we kept heaping on above, clay would continue to be squeezed * The student who wishes fully to understand how ice is able to manage this, must consult Tyndall, The Glaciers of the Alps ; Lyell, Principles, vol. i. chap, xvi. ; Canon Moseley, Proceedings of the Royal Society, xvii. 202 ; Croll, Phil. Mag. March 1869, and September 1870 ; Climate and Time, chaps, xxx. and xxxi. ; also Nature, i. 116, iv. 447, v. 185, vi. 396. 198 Geology. out below, and would slide in an unbroken sheet down the boards on to the floor. If there were grooves or hollows in the surface of the boards, the flow of clay would evidently take place chiefly along them. This is exactly what happens when a great heap of snow and ice has been piled up on a lofty tableland ; the weight of the huge mass drives portions over the edges of the tableland and down its slopes ; and as the pressure from behind is kept up by the additions which are always being made to the pile at top, a continuous and steady flow is maintained. In high latitudes, where the snow-line comes down to the sea-level and the whole land is cased in ice, there is a discharge of the latter all along the coast-line into the sea ; in more temperate climates, where snow accumulates permanently only on very high ground, the ice drains off down the valleys in the form of long tongues, known as glaciers, which are really ice-rivers, always sliding downwards, and whose motion, except that it is slower, differs in no respect from that of streams of liquid water. Glaciers descend far below the snow-line, but sooner or later reach a level at which they can no longer remain in a frozen state, when they melt and become rivers. The under surface of a glacier is just at the melting-point, and the water derived from the thawing of the bottom layer of the ice, together with that which sinks down crevasses when the upper surface is melted by the midday sun, runs in a stream between the base of the glacier and the rock on which it rests, and issues, often in considerable body, from beneath the snout. The tableland on which snow accumulates is called the "gather- ing-ground," and the parent mass of snow the " snowfield " or "neve." Glaciers, like rivers, act as carriers of debris brought on to them. On the bare mountain slopes, which rise above the ice, atmospheric weathering goes on largely, and the loose matters thus produced roll clown the hillsides and fall on to the surface of the ice. Thus long lines of dirt, stones, and large angular blocks of rock are always found fringing the edges of a glacier. These are called Lateral Moraines. When two valleys meet and their respective glaciers unite, the two inner lateral moraines run together into a heap of rubbish in the middle of the glacier and form what is called a Medial Moraine. In the case of a large glacier formed by the junction of many tributaries, there will be many of these medial moraines, so that in some cases the surface is so thickly strewn with dirt and rubbish that the ice can scarcely be seen through it. All this burden is carried slowly forward by the downward movement of the glacier, and at last shot over the end, where it is piled up in a heap called a Terminal Moraine. The Terminal Moraine is constantly being worn and wasted by the stream which issues from beneath the snout of the glacier, and its materials are ground fine and swept down and go the way of other products of Atmospheric Denudation. Fig. 79 is a somewhat diagrammatized view of a glacier formed by the junction of the ice-streams of two valleys : the outer lateral Denuding Agents and hoiv tliey work. 199 moraines fringe the edges, a medial moraine is seen formed by the union of the two inner lateral moraines : at the extremity the lateral and medial moraines are being shot over to form a terminal moraine. A stream rushes out from an ice-cave beneath the glacier, which has cut its way through the terminal moraine, so that only small portions of the latter remain on each side. In the extreme dis- 2OO Geology. tance we catch a glimpse of snow-clad hills forming part of the snow- field* As rivers abrade their banks and beds by the aid of the sediment they carry along, so glaciers wear away the bottoms and sides of the valleys along which they flow. Stones fall through the fissures or crevasses which traverse the ice, or are picked up from the bed of the glacier, and get firmly frozen into its base. The under surface is thus converted into a great rasp, which grinds the rocks over which the glacier passes, and wears them down into the finest and most impalp- able mud. The stream below the glacier carries this on, and rushes out from beneath the end largely charged with mud ground so fine as to be easily carried to long distances, which is sometimes called " Flour of Rock." The great network of the tributaries of the Rhine, for instance, is formed of streams draining the northern flank of the Swiss Alps, the fine glacial mud brought down by them is carried on by the river, and out of its settlings the flat lands of Holland have been in great measure formed. Continental Ice-sheets. In Arctic or Antarctic regions, where the conditions for the accumulation of large masses of snow and ice are present, the ice is not confined to the valleys, but the whole land becomes cased in a widespread sheet of it, which wraps everything in one unbroken covering from the highest ground down to the sea-level. The best known case of this sort is that of Greenland : the interior of this country, wherever an attempt has been made to penetrate into it, has been found to be buried in ice, and it is probable that the ice-sheet stretches without break over the whole land. In some parts the frozen mass reaches quite down to the coast, and terminates in an abrupt wall, not uncommonly from one to three thousand feet high, and sixty miles or more in length. Elsewhere strings of hills stand up like islands between the interior ice and the sea, and in the valleys and fiords which separate these detached masses of bare land the ice passes out to sea in great glaciers, t Just in the same way as glaciers, masses of continental ice are always slowly moving from the interior to the coast, whence they continue their motion over the sea-bottom till the water is deep enough to buoy the end up : huge masses then break off and float away as icebergs. Icebergs are also formed by portions which tumble into the water from the wall in which the ice-sheet sometimes terminates. Just as in the case of glaciers too continental ice-sheets grind the surface of the ground over which they pass into fine mud, and discharge large quantities of it by subglacial streams directly into the sea. But it also seems likely that huge ice masses, such as we are now considering, will grind and tear up the ground underneath them * The reader who wishes to have in a few words a graphic description of glacier regions should turn to Professor Geikie's Primer of Physical Geography, p. 75. For details see the works of Agassiz, Charpentier, J. Forbes, Tyndall, and the publications of the Alpine Club ; also the "Report on Ice as an Agent of Geolo- gical Change," Reports of British Association, 1869, p. 11. t See the works of Dr. Kane and Dr. Hayes; Dr. Brown, Quart. Journ. Geological Soc. of London, xxvi. 671 ; Rink, Journal of Royal Geographical Society, xxiii. 145. For the Antarctic Regions, see Sir J. Ross's Voyages. Denuding Agents and Junv tJiey ? to a much larger extent than even the largest valley and that there will be thus formed beneath them a great mass of stones and dirt. To this the name of Moraine Profonde or Grund- morane is given. This pell-mell assemblage will be pushed to and fro by the moving ice, and if it be driven into a valley or sheltered recess, the ice-sheet may ride over it and disturb it no further, and there it may remain, and if any improvement in climate cause the ice to dis- appear, it will furnish a proof of the former presence of an ice-sheet long after the latter has ceased to exist.* If any peaks of bare rock rise above the surface of a sheet of con- tinental ice, moraines will be formed on the latter just as they are ibrmed on glaciers, and carried down to the coast. Icebergs bear away portions of this moraine matter, and also stones from the moraine pro- fonde frozen into their base, and drop their burden as they melt. In this way rubbish and large unrounded blocks of rock are deposited on the sea-bottom f;ir away from the source from which they were derived. Coast Ice. Ice also does important work as a carrier of denuded matter in the shape of coast ice and ground ice. In high latitudes it often happens that from the melting of snow on the shore the water adjoining the land becomes freshened far enough to allow of its freez- ing at a higher temperature than the body of the sea, and a belt of ice, known as coast ice or "the ice foot," is formed along the shore.t On to this fringe of ice debris rolls down from the land, and shingle gets frozen into its under surface. The coast ice is lifted and at last de- tached by the rise and fall of the tide, and portions of it with their load of detritus float away, and drop what they carry as they melt. Ground Ice. Ice known as "ground ice" or "anchor ice" forms sometimes, in a way not perfectly understood, at the bottoms of lakes and rivers while the rest of the water remains unfrozen. J Pebbles and other loose matters are frozen into the under surface of this sort of ice, and lifted by it and floated away when it rises to the surface. 4. ACTION OF WIND. The share which wind takes in the work of denudation, though not very large, ought not to be entirely overlooked. Its effects are best seen in those isolated blocks or pinnacles which often rise from the surface of a country composed of coarse sandstone. These are very frequently undercut or worn away below, taking the form of anvils or one-legged tables "Shapes The sport of nature, aided by blind chance, Rudely to mock the toiling works of man." In these cases we find the surrounding ground strewn with coarse sand produced by the decomposition of the rock below. This sand the * J. Geikie, The Great Ice Age. chaps, vi. and vii. t Quart. Journ. Geol. Soc. xxxiv. (1878) 565. See Engelhardt, Smithsonian Reports, 1866, p. 425, and Ann. de China, et de Phys. 1866. Also Geol. Mag. [2] iii. (1876) 459. Wordsworth. 202 Geology. wind drives against anything that stands up above the surface and grinds it away, but as the wind can lift the sand only a short distance from the ground, the wearing is confined to the lower portion of the obstacles. Fisr. 80 is an instance of one of these undercut rocks : in Fig. 80. UNDERCUT TABLE or GRITSTONE. this case probably the process has been helped by the coping-stone be- ing harder than the beds below, but plenty of cases occur where pillars of rock of equal hardness throughout are hollowed out underneath in exactly the same way. Similar forms are very common in Granite. Rocks weathered in this way are frequently mistaken for " Druidical remains." The most striking instances of this method of denudation are met with in the western territories of the United States.* In deserts, and other large sandy tracts, the drifting of sand by the winds grinds and wears rocks that stand in its way, and produces very remarkable polished surfaces and scratches upon them, not unlike those due to the action of moving ice.t By processes like these no inconsiderable amount of rock is worn away. Denudation of this sort is sometimes called ^Eolian. Wind also acts as a transporting agent ; sand and dust, and any loose matters produced by the weathering of rocks, are swept by it into running water or the sea : but perhaps the most important work it does in this way is by transporting the light ashes thrown up by volcanoes ; these are carried by it to vast distances ; if they fall on the land, they are ready to be swept further on by rain and rivers ; or they may fall directly into the sea ; in either case they furnish materials for subaqueous strata. Wind also aids the sea and other large bodies of water in the work of denudation by causing waves and unusually high tides. In this * See Gilbert, Geol. of Parts of the Western Territories, p. 82 ; Bulletin of the United States Geol. and Geog. Survey of the Territories, iv. (1878) 831 ; Reports of ditto, 1873, pp. 32 and 36, 1874, plate iv., 1875, p. 156. t Geol. Explor. of the 40th Parallel, ii. 159. Denuding Agents and hoiv they work. 203 way they are enabled to act more energetically in the destruction of their coasts and banks, and the rush back of the pounded-up water, when the gale abates, sweeps before it with more than usual force the stuff which the storm has brought down. 5. ORGANIC DENUDING AGENTS. The help given by plants and animals towards denudation, though not very important, calls for a passing notice. Burrowing animals, such as rabbits and moles, undermine the ground, give passage to rain, and so weaken the surface and render it an easier prey to other denuding forces. The matter they throw out is ready to be carried away by rain : in this respect even so insignificant a creature as the earthworm has been thought worthy of being noticed by geolo- gists.* Marine boring-shells and land-snails bring about the destruc- tion of Limestone to a small but appreciable extent. Trees destroy rocks mechanically by forcing down their roots into crevices and so splitting off portions : but plants do their most impor- tant denuding work indirectly ; by their decay they furnish Carbonic Acid to water and thus enable it to dissolve Limestone. Professor Ansted mentions cases of holes drilled in this way to great depths, and sometimes right through blocks of Limestone, eacli hole containing the stem of a plant which by supplying Carbonic Acid to water had been able to work through the rock as effectually as a boring-drill, t GENERAL VIEW OF SUBAERIAL DENUDATION. Such then are the main subaerial denuding agencies, and the ways in which they each individually do their work ; let us next see what is the result of the joint action of all. The first step in the process of subaerial denudation is the formation of soil. Formation of Soil. Actual bare rock is a thing not often seen at the surface ; in a vast majority of cases what we first come to on breaking up the ground is a layer of soil. This is almost universally the case in flat countries, very generally so in hilly ones, and it is only in mountainous tracts that we find any widespread exceptions to the rule.J Now it is in most cases easy to see that this coating of soil does not consist of matter brought from a distance and spread over the solid rock beneath, but that it is nothing more than the upper portion of that rock itself broken up and converted into sand, clay, or some other incoherent material. Various subaerial denuding agents have worked together to produce this result. Eain has softened and in some cases decomposed chemically the constituent minerals of the * The Formation of Vegetable Mould by the Action of Worms, C. Darwin. f For the action of plants in helping the accumulation of silt see Nelson, Trans. Geol. Soc. [2] v. Pt. 1, p. 103, and Proceed. Geol. Soc. June 1852. This statement is not strictly true of Limestone countries, which show always a tendency to a bare rocky surface, because their Carbonate of Lime is dissolved by the rain-water and carried away in solution. 2O4 Geology. rock ; frost has shivered it ; the heat of the sun and other atmospheric agencies have dried, cracked, and pulverized parts of it ; the roots of trees and perhaps burrowing animals have had some share in breaking it up : by these and suchlike forces any exposed surface of rock is inces- santly being attacked until a portion of it is converted into loose soil. The natural planes of division, known as joints and cleavage, by which rocks are traversed, aid materially in this work of destruction. They allow of the percolation of water into the body of the rocks, and are planes of weakness along which fracture is readily produced. On exposed mountainous countries the light matters thus formed are washed away by rain, or roll down the hillsides by their own weight as fast as they arise, and therefore in such situations the rock is con- stantly kept bare in spite of the attempts made by subaerial denuding agents to bury it in its own debris : in countries less hilly, and with a smaller rainfall, the process of removal goes on to a smaller extent ; while in flat countries, where the fall of the ground is small, and the carrying power of water running over it is little or nothing, soil forms faster than it can be carried away, and the solid rock is everywhere deeply buried in its own ruins. These accumulations of rain-borne decomposed rock go by the general term of " Eain-wash ;" they may be distinguished (1) by their materials being strictly local in their origin, (2) by their stones, if they contain any, being not water-worn but angular, or at most showing only so much rounding as might be produced by the chemical dissolu- tion of their angles and edges. Deposits of rain- wash of various kinds occur in the south of England.* Deposits of this class are also very largely developed in Spain, in the flatter parts of which we may travel for hundreds of miles without seeing a bit of rock except in the deepest railway cuttings ; so that from a general point of view the country may be said to present only two physical features, broad plains of rain-wash and mountainous Sierras. The causes of this are twofold : the large arid violent rain- fall, and the great extremes of temperature which often prevail, give rise to rock-disintegration on a large scale, and the plateau-like form of the ground prevents the debris so formed from being carried away. The surface disintegration of rocks is nowhere better shown than in the case of Granite and some Traps. In Granite this is due mainly to the atmospheric decomposition of the Felspar, and in this way the rock is often reduced to a mass of loose fragments, which can be shovelled out with a spade to a very considerable depth, t A large tract of country round Madrid, which is coloured on some geological maps as Diluvium, is covered by decomposed Granite, in which it is easy to pick out bits of the rock only partially disintegrated. In the same way some diorites weather down to loose earth, in which are embedded portions rudely spherical in shape that have been hard enough to resist decomposition ; and the whole has so exactly the look * See Goodwin Austen, Quart. Journ. of the Geol. Soc. vi. 94, vii. 121 ; Foster and Topley, ditto, xxi. 446. t See De la Beche, Geological Observer, pp. 3, 4. Denuding Agents and how they work. 205 of the accumulation of sand and boulders formed by a mountain torrent that it might well be mistaken for a mass of water-worn materials. Removal of Soil from higher to lower Levels. We must now advance a step further ; only a portion of the disintegrated rock remains to form soil : in some cases we have seen it is swept away as fast as it forms, and even when the rate of formation is greater than that of removal, some part of the loose matter is con- stantly moving onwards. Sooner or later the products of surface weathering find their way into a brook, and are swept forward by it, either in suspension or by rolling along the bottom, till the brook joins a larger stream, along which they travel till the stream falls into a river, and along this they continue their course till the river is lost in the sea. At the same time these transported materials enable the water of the streams, which by themselves have little or no abrading power, to wear and grind the banks between which they flow and the bed of their channel, and thus to add to the amount already being carried downwards by them. The sum-total of transported matter is also swollen by salts and other substances dissolved by rain in its course over the ground, brought up from below in solution by springs, or taken up by the waters of the river itself in its passage over soluble rocks. The amount of these chemically-dissolved substances is far from incon- siderable ; thus Professor Ramsay tells us that the Thames carries every year into the sea 33,497 tons of matter (chiefly Carbonate of Lime) in solution ; and this, if precipitated and compressed into Lime- stone, would form a bed a yard thick and more than seven acres in extent. The above are the principal steps in the process by which water in its liquid state transports the products of subaerial denudation con- tinually from higher to lower levels. The share which it contributes to the work, in its solid shape as ice, falls next to be considered. We have seen that the streams which issue from beneath the snouts of glaciers are largely charged with sediment already ground so fine that it is at once carried forwards. The coarser matters shot over to form the terminal moraine are attacked by various subaerial denuding agents, in the end ground fine enough to be moved, and then are carried away. Where glaciers or masses of continental ice come down to the sea- level, the streams beneath them discharge directly into the sea large quantities of finely-comminuted mud ; and the icebergs which break off from them carry away coarse materials and large unrounded blocks of rocks, and deposit them far away from the spots from which they were derived. In a word, the surface of the ground is constantly acted on by a number of agencies, which all work together to wear and break it up : the loose matters so produced are carried downwards, and at the same time added to, by moving water either in a liquid or a solid state, till they at last come to rest at the bottom of large bodies of still water. This chain of events, all intimately connected with one another, con- stitutes the process of subaerial denudation. 206 Geology. 6. MARINE DENUDATION. The sea to a very large extent only finishes work begun for it by subaerial denuding agents. The coarser stuff brought into it by rivers is tossed to and fro by the tides, till it is ground fine enough to allow of its being swept away altogether. In the case of a coast bounded by cliffs, the expansion of frozen water, the undermining caused by the outbreak of springs, or the unequal yielding to the weather of beds of different hardness, and other similar causes, break off and throw down large masses, and the sea completes the work by grinding these into mud, shingle, or sand, and then by the aid of tides and currents sweeping them away. But the sea may also claim a certain amount of the denuding work which it effects as entirely its own. In the same way as the sediment carried by running waters enables them to grind away their banks, the sea uses the boulders and shingle of the beach as instruments for the destruction of its cliffy shores. Waves, rolling in from open ocean spaces and driven forward by gales of wind, have force enough to lift and hurl against anything that comes in their way enormous masses. By this means the loose blocks that fringe in heaps rocky shores are dashed against the cliffs, and by this incessant pounding and battering fresh portions are from time to time brought down, to be used in their turn as instruments for further destruction. That the waves beat upon the land with fearful violence when the wind blows inshore, specially if these waves have gathered head by rolling across a broad expanse of open ocean, we are all ready to admit. But unless the figures have been actually laid before them, few people have an adequate idea of the feats of which such waves are capable. Those who have felt a large ocean-going steamer shiver like a live creature from stem to stern when she is struck by a heavy sea, can realize what the effect of such a wave will be when it breaks in shallow- water. We get the most exact measures of the power of storm-waves from the experience of those who have to deal with the construction of breakwaters and lighthouses. Even in the sheltered estuary of the Firth of Forth stones more than a ton in weight have been torn out of a sea-wall and transported for a distance of thirty feet. On Plymouth breakwater blocks of rock weighing from three to five tons have been wrenched from their bed, tossed about like pebbles, and rolled up the inclined plane of the breakwater ; one block of seven tons was rolled a hundred and fifty feet. Professor Geikie mentions that at Barra Head, in the extreme north of the Hebrides, a block, estimated to weigh forty-two tons, was moved five feet, and came to rest only because it became jammed by a fragment broken off it. No rock however hard and tough would stand the battering by blocks far smaller than these just mentioned, and the hardness of the rocks of a coast, so far from protecting them from the encroachments of the sea, makes them in some measure a more certain prey to its wearing action, for the harder the rock is, the more destructive will be the ammunition furnished by its ruins. Where the coast is composed of soft rocks, it is eaten into all the more easily ; its destruction is incessant, and the Denuding Agents and hoiv tJiey work. 207 advance of the sea becomes rapid enough to be obvious even to the most casual observer. It must be noted that the destructive action of the sea is confined almost entirely to the belt between high- and low-water mark. With- in that space the rise and fall of the tides and the force of the breakers grind down any loose matter exposed to their action. These agencies however cease to have any effect on a bottom covered by a moderate depth of water, and hence very nearly all the denuding work of the sea is Coast Denudation. The drifting of rough sediment over the bottom by under-currents may produce some abrasion, but its amount cannot be very large.* Relative Importance of Subaerial and Marine Denuda- tion. We may here note that marine denuding agents, such for instance as the beating of the waves on an exposed rocky coast, are far more striking and appeal far more forcibly to the imagination than the slow and almost insensible action of subaerial denuding forces. The latter, partly because they are so common, partly because their action is so gentle, and partly because they operate so slowly that their results are inappreciable unless they are very carefully measured or observed over a long period of time, are apt to be overlooked, and indeed were for a long time, if not actually overlooked, yet denied their true im- portance by geologists. It is now however very generally recognised that they perform by far the larger part of the denudation that is going on before our eyes, and they have at last had their true place granted them in the roll of denuding forces, a place to which Hutton long ago pointed out that they were entitled. It seems almost past belief that the importance of subaerial denuda- tion should have been so long overlooked, and that truths so simple and apparently so self-evident as those, of which w r e have just given an abstract, should not have forced themselves on the notice of geologists from the very first birth of the science. The explanation however is easy enough, and furnishes so useful a lesson to the would-be cultivator not only of geology, but of any other science, that it is worth calling- attention to it. Men failed to see these obvious truths simply because they did not look for them ; because, instead of going forth and mark- ing Nature, they amused themselves with weaving ingenious conceits in armchairs at home. Hutton was the first clearly to enunciate the laws of denudation, which he had learned by observation ; the sum- mary of one of his chapters is worth quoting. " Whether we examine the mountain or the plain ; whether we consider the degradation of rocks or of the softer strata of the earth; whether we contemplate Nature and the operation of time upon the shores of the sea or in the middle of the continent, in fertile countries or in barren deserts, we shall find the evidence of a general dissolution on the surface of the earth, and of decay among the hard and solid bodies of the globe." Playfair puts it more tersely thus. " The consequence of so many minute, but indefatigable agents, all working together, and having * For Marine Denudation see Lyell's Principles, vol. i. chaps, xx. and xxi. ; De la Beche's Geological Observer, sec. v. ; Geikie, The Scenery of Scotland viewed in connection with its Physical Geology, chap. iii. 2o8 Geology. gravity in their favour, is a system of universal decay and degradation, which may be traced over the whole surface of the land, from the mountain-top to the sea-shore." Another useful lesson may be learned from this bit of the history of Geology. It is now nearly ninety years since the Theory of the Earth was published, and it is only quite lately that geologists have come to recognise the truth of its teaching. So slow are men, even when the right road is pointed out to them, to leave a groove which they have been for a long time following.* * In connection with the subject of Denudation the student will do well to read Theory of the Earth, Part I. chap. i. sec. iv., Part II. chaps, iii. to vii. ; Playfair's Illustrations, sec. iii. ; Scrope's Volcanoes of Central France, 2nd ed. chap. ix. ; Ramsay, The Physical Geography and Geology of England and Wales ; A. Geikie, The Scenery and Geology of Scotland ; A. Streng, Uber den Kreislauf der Stoffe in der Natur, Neues Jahrbuch, 1873, p. 33; Gilbert, Report on the Geology of the Henry Mountains, chap. v. CHAPTER IV. WHAT BECOMES OF THE WASTE PRODUCED AND CARRIED OFF B Y DENUDA TION. THE ME THOD OF FORMA TION OF BEDDED ROCKS, AND SOME STRUCTURES IMPRESSED ON THEM AFTER THEIR FORMATION. " The Rhone by Leman's waters washed, When mingled and yet separate appears The river from the lake, all bluely dashed Through the serene and placid glassy deep, That fain would lull her river-child to sleep. "BYRON. SECTION I. MATTER MECHANICALLY CARRIED. HAVING now passed in review the various agencies which at all times and in all places are at work breaking up the surface of the earth, and having convinced ourselves that the larger part of the waste which results from their action finds its way sooner or later into running waters, and is carried on by them in their downward course, our next task is to inquire what happens to the matters thus swept away when the streams which bear them along are lost in bodies of still water, such as the sea or a large lake ; and we will begin with the mechani- cally-transported matters, those namely which are carried in suspension and those which are swept along the bottom. When the velocity of the stream is by degrees checked, and at last destroyed altogether, it can no longer bear along its burden, the stuff that has been rolled along the bottom comes to rest, and the sediment held in suspension sinks clown. But the suspended matters will not fall all together : there will generally be some of them heavier than others, some coarse and others more finely divided. Long after the current has ceased to be able to hold up the heavy and coarse part, it may retain velocity enough to carry forward the light and more finely divided, and the latter will travel much further than the former from the mouth of the river before they reach the bottom. Dr. Sorby has found that grains of sand TOO f an i ncn * n diameter sink in still water one foot in ten seconds : while the finest Kaolin takes five days to sink through the same space. If there be a current running it is evident that Kaolin will be carried much further than sand before it comes to rest on the bottom. That finely-divided sediment does o 2io Geology. settle down very slowly was well shown by one of the soundings on the Challenger Expedition made about four hundred miles off the coast of Africa. Even at that distance the bottom was formed of brownish mud evidently coloured by debris brought down by some of the small African rivers (Voyage of the Challenger, ii. 84). Arrangement of Mechanical Deposits according to Size and ^ATeight. In any large body of water then, fed by running streams, we should find deposits on the bottom arranged somewhat in the following order. Fringing the coast, and especially facing the mouths of the rivers, there will be a belt of banks of coarse pebbly materials. In the case of lakes or tideless seas these may stretch out for a considerable distance, for when the water has been shallowed for some way out by the formation of a bank of shingle, pebbles may be rolled on in the shallow water on the top of the bank and shot over the end, and the front in this way be continually pushed forward. Along the shore of the open ocean no broad accumulation of shingle can take place, for the pebbles are always being ground fine, and swept out to sea by the rise and fall of the tides and the beating of the breakers. When these shingle-banks can be formed, they will evidently be thickest on the shore side, and thin away in a wedge-shaped form as we advance into deeper water. Beyond the shingle-banks, and resting on their thin edges, there will be other banks of a similar shape, but formed of materials a little less coarse : and because the components of these can be borne rather further than the coarsest shingle before they come to rest, these banks will not thin away quite so rapidly, and will form wedges with angles more acute than those of the banks next the shore. In this way as we leave the shore we shall find the deposits becoming less and less coarse, and arranged in wedges getting more and more acute, till at last their upper and under surfaces become approximately parallel, and they take the shape of beds or strata. The lightest and most finely-divided matters will sink very slowly through the water and travel far before they reach the bottom, and will come to rest in layers or laminae, which keep nearly the same thickness over large areas. Arrangement of Mechanical Deposits according to Mineral Composition. Besides these differences in arrangement there will also be a difference in the mineral composition of the deposits fringing the shore- and those remote from it. The sediment carried down mechanically by running water consists mainly of two kinds, sandy or siliceous, and clayey or argillaceous. Now Quartz, of which sand is composed, is very hard, and will stand a great deal of wear and tear before it gets ground fine ; hence the coarser deposits will be mainly sandy. Clay, on the other hand, is soft, and easily worn down into the finest impalpable mud ; hence the finer deposits will be mainly clayey. Therefore we shall find, as a rule, that near the shore wedge-shaped banks of coarse, sandy materials prevail, while further out regularly bedded layers of fine day cover the bottom. Matter mechanically carried. 211 This order will not be without interruption, because during floods the coarser materials will be carried further out than usual, and so wedges of sand will come to be interleaved among the evenly-bedded clays ; and in the same way, when the rivers are low, clayey deposits will be formed in the sandy region : still upon the whole the general arrangement of the deposits will be such as has been described. Finally, if the ocean be large enough, there will be a limit beyond which no river-borne sediment will be carried, and no mechanical deposit formed on the sea-bed. These regions however, as we shall see further on, will not be without deposits, for in them marine ani- mals build up great masses of pure Limestone. General Arrangement of Mechani- cal Deposits. In fig. 81 an attempt is made to show diagrammatically the arrange- ment of mechanically-formed deposits. The dark part is the solid rock forming the land and the sea-bed ; the straight line the sea-level. Looking at the lower and therefore first-formed accumulations of sediment, we see close to the shore a bank of large pebbles with a steep face seawards : beyond this there follow other banks, the first of pebbles not so large as on the bank nearest the shore, the next of coarse sand, the next of finer sand, and so on : and the seaward faces of all these banks get less steep as we leave the shore. Beyond the finer mud stretches out in thin layers, becoming more and more nearly horizontal as we get out to sea. After these beds had been laid down, the streams ceased for a while to have the power to carry coarse sediment, and could not bear even fine mud as far as before : consequently the latter .sank down nearer the shore, filling up the hollows between the banks, and levelling over their uneven surface. Subsequently coarse matters were again brought into the water, and a range of banks similar to those in the lower part of the section were piled up on the top of the layers of mud, while finer deposits again began to accumulate in the remoter parts of the ocean. In the diagram the transitions are, for the sake of distinctness, made abrupt ; but in Nature the passage from shingle to sand, and from sand to mud, would be much more gradual, and, as the material grew finer, almost insensible. In the crust of the earth we meet with rocks which, except that they are harder and bound more firmly together, bear the closest resem- 212 Geology. blance to the accumulations that have just been described. Con- glomerates are composed of exactly the same materials as the shingle of the beach and littoral zone ; Sandstones of all degrees of coarseness find a parallel in submarine sand-banks ; there are sandy Shales, which are mixtures of fine sand and mud ; and the finely-laminated argillaceous Shale cor- responds exactly to the evenly-bedded deposits of small silt and mud. And among these rocks we find also just the same arrangement as has been described in the last few pages. In the great masses of pebbly Sandstone, as in shingle-banks, the pebbles are found to grow smaller and smaller as we trace the rock in a certain direction, and at last disappear altogether, so that the bed passes insen- sibly into a rough Gritstone, this again still further on merges into a finer Sandstone, and perhaps at last is found to tail out altogether in a wedge-shaped form, and to be replaced by deposits of more regularly-bedded hardened sandy mud, and these in turn shade off into more purely clayey mud. We also find alternations of sandy and clayey rocks, corresponding to the alternations of banks of sand with deposits of mud, which are now being pro- duced by variations in the transporting power of currents. Further we find that masses of rock of coarse sandy composition contrast in shape with those of a fine clayey nature in exactly the way that the above considerations would lead us to expect. It is scarcely ever strictly correct w to speak coarse Sandstone. On the lare scale coarse Sandstones occur in lenticular-shaped masses, and these masses are themselves made up not of beds but of wedge-shaped portions dovetailing into one another. The clayey deposits on the other hand spread over large areas with comparatively little variation in their aggre- gate thickness, and are made up of numerous laminae, each of which is equally constant in thickness. We may then talk of "beds" of Shale, but if we would be very accurate we ought to speak of " banks " of coarse Sand- stone. The best proof of these facts is found by compar- ing the strata passed through in sinking coal-pits to the same bed of Coal at different spots. Fig. 82 shows an actual case in the Yorkshire Coalfield. In all the sinkings to a bed of Coal AB along a distance of twelve miles, the seam is overlaid by some forty yards of Shale. At C a bed of Sandstone puts in. It is at first very thin, but in the space of less than half a mile it swells out to a thick- ness of thirty yards or more, and it forms the roof of the Coal for a distance of ten miles. At C and for some miles to the north it is very coarse, occasionally a Conglomerate, but as we go northwards it grows smaller in grain, and at last becomes a rock of remarkably even and fine texture. Still further to the north it grows clayey and merges into a mass of sandy Shale. Matter mechanically carried. 21 In fact so variable is this Sandstone, that we should hardly believe that we had been all along dealing with the same rock, if it had not been for the constant presence beneath it of a bed of Coal about whose iden- tity throughout there can be no doubt. It is likely that originally the Coal was covered throughout by a band of Shale similar to that found above it to the south of C. This was stripped off by denudation to the north of C. Into the hollow so formed a mixture of clayey and sandy detritus was carried by currents coming from the south : the stuff was sorted in the way which has been described, the coarser part dropped near the margin of the hollow, while the finer floated away towards its central part.* An admirable instance of the gradual replacement of a Sandstone by Shale may be actually seen in a railway cutting south of Normanton Station in Yorkshire. At the south end of the cutting about thirty feet of solid Sand- stone are exposed : a little further to the north little wedges of Shale may be observed thrusting themselves in between the lenticular masses of which the Sandstone is made up. The thin edge of each wedge points south, and each gradually thickens out, while the Sandstone which separates them from one another thins away, till in the space of less than a quarter of a mile there is no Sand- stone left, and the sides of the cutting are Shale from top to bottom. Fig. 83, which is a section exposed in a railway catting near Pontefract. I ; ' shows a still more complicated interlacing of Sandstone and Shale ; the dotted parts are Sand- stone, the parts marked by fine parallel lines are Shale, and the black patches nests of Coal which will be noticed further on. Some substances remain suspended in water not so much on account of their low specific gravity as in virtue of their form. One of the commonest of these is Mica. The thin laminae into which this mineral splits present a broad sur- face to the water and give rise to an amount of resistance to sinking very large in comparison with their weight ; henco they settle clown very slowly and regularly. In this way the surfaces of the beds of regularly-stratified deposits are thickly flecked over with little spangles of this mineral. Sandstones and Shales of this character are very common, and are called Micaceous Sandstones and Shales. That Mica does behave in the way described is found out in the washing of China Clay : owing to the way in which the water holds up its thin plates, it is the most difficult impurity to get rid of. * The Geology of the Yorkshire Coalfield (Memoirs of the Geological Survey), p. 434. 214 Geology. Influence of Salt in promoting Deposition. It has been noticed that suspended matters sink down more rapidly in salt than in fresh water. This is only a particular case of the general truth that there are a large number of substances which when dissolved in water hasten the precipitation of suspended matter, many indeed to a much larger extent than common salt. This property would doubtless tend to promote deposition to some extent ; but there are substances which have exactly the opposite effect. One of these, silicate of soda, must often be present both in river- and sea-water.* Horizontal Growth of Coarse Deposits. Accumulations of shingle such as we have dealt with form a fringe of comparatively small breadth along the margin of the water, and wedge-shaped in form. But coarse Grits and Conglomerates do occur extending over areas hundreds of square miles in extent with a fair persistency in thickness ; and we must next explain how these acquired their large horizontal dimensions. Suppose ACE (fig. 84) to be the cross section of a fringing belt of coarse deposit. The result of this accumulation will be to make the water between A and C so shallow that pebbles and coarse sand can now be rolled as far as C \ arrived there they will roll down the face CJ3, and extend the bank further out. By successive additions of this kind the bank will continue to grow out- wards ; and there will be nothing to stop its increase provided the water does not deepen too rapidly. Given then a broad spread of shallow water, the formation of a broad spread of Conglomerate pre- sents no difficulty. We can also readily imagine that with shallow water and shifting currents banks of sand may be at one time piled up, and at another, when the current changes its direction, may be broken down and spread out, and that repeated washing to and fro may at last arrange the material in a mass of fairly uniform thickness. Vertical Growth of Coarse Deposits. We have yet another difficulty to meet. The coarse deposits we have so far accounted for can never be of large thickness. But we do find Grits and Conglomerates not only spreading over large areas but also reach- ing thicknesses of hundreds of feet. How was this thickness obtained 1 We shall see by-and-by that the surface of the globe is never at rest, that portions are rising and others sinking, and that these changes of level have been going on throughout all past time. Now suppose the bed of a shallow sea to be sinking, say at the rate of a foot in a century, and also that in the same time one foot of sediment is laid down all over the bottom ; then, since the accumulation of sediment fills up the water at exactly the same rate as the sinking would deepen it if no deposition were going on, it is clear that the depth will always remain the same, and that the conditions necessary for the accumula- tion of coarse deposits will always be preserved ; and with such an adjustment any thickness whatever of such deposits may be obtained. * Robertson, Trans. Geol. Soc. Glasgow, iv. 257 ; Ramsay, Quart. Journ. Geol. Soc. xxxii. (1876) 129. For a full statement of the facts aiid an ingenious explanation see Jevons, Quart. Journ. Science, April 1878. Matter mechanically carried. 215 Wherever then we find a great thickness of coarse beds, \ve know at once that during their deposition the sea-bed must have been sinking at about the same rate as they were accumulating. And this explanation will apply not only to Conglomerates and Grits, but to all rocks formed in comparatively shallow water. For instance in the Coalfield of South Wales rocks certainly subaqueous, but none of which can have been deposited in deep water, are piled one on the top of another to a thickness of ten thousand feet. These must have been laid down on a subsiding bottom. Drift- or Current-bedding. We may now look a little more closely at the way in which the materials of the banks of coarse sedi- ment are arranged on a small scale. The process is exactly the same as goes on on the large scale. In fig. 84 AE is the surface of the Fig. 84. FORMATION OF CURRENT BEDDING. water, and CD the front of one of these banks in course of formation. From A to C the water is shallow enough to allow the current to retain velocity sufficient to roll sand or pebbles along the bottom, but in the deep water beyond C this velocity becomes suddenly checked and their further progress arrested. The coarse matters are therefore shot over the edge of the bank, and arrange themselves in a little sloping layer, CDEF: by this means the extent of shallow water will be a little Fig. 85. QUARRY IN CURRENT-BEDDED ROCK. increased, and another sloping layer added above CD. And so the process goes on, till in the end a bank is formed made up of thin 216 Geology. sloping layers all dipping in the direction in ivhich the current is moving. If a current with a different direction pass over the same spot, another bank will be piled up, composed like the first of thin sloping layers, but with its layers dipping towards a different quarter. By a repetition of this process we shall obtain a deposit made up of wedge-shaped beds, each of which is traversed by smaller planes of division crossing the main lines of bedding obliquely. Such a structure is called False-bedding, Cross-bedding, Current- bedding, or Drift-bedding ; an example of it is given in fig. 85. Rock possessing this structure is sometimes called "Eddy Rock" by quarry- men and well-sinkers. Ripple-drift. Let us next consider what will happen on a sea- bottom on which a current is forming ripple-marks. The shape of the bottom is such as is shown in %. 86a, the arrow being the direction of the current, and each ripple having a long gentle slope on the side from which the current comes, and a steeper slope on the opposite side. If no sediment is being brought into the water, the current will roll sand up each gentle slope, and the latter will fall down over the steep slopes, and the only thing that will happen will be a general movement of the ripples in the direction of the arrow. But if sand be sinking through the water, it will, as it falls on the gentle slopes, be rolled up them and over their edge, and fall down in a layer over each of the steep fronts, AJJ, CD, EF, as in fig. 866. In this way the steep fronts will always be added to, till at last the old surface, ABCDEFG, will be effaced by the filling in of the hollows ABC, CDE, EFG, and a new ripple-marked surface, aAlCcEd, formed above it, fig. 86:. (1864) 358. Mantell, Ann. and Mag. Nat. Hist, xvi. (1845). OF THE Dissolved Matters. U IT I V BsR S I !] precipitated in various forms, and that in many portions of the Chalk ; and that the pectous condition i probably be when first thrown down has been in the course of time gradually exchanged for a more or less crystalline state. In the case of the Carboniferous Limestone of Ireland, Messrs. Hardman and Hull have shown that whole beds of Limestone have been more or less completely replaced by Silica and converted into Chert;* and Professor Renard has proved the same for the Chert (Plithariite) of the Carboniferous Limestone of Belgium, t Professor Sollas has detected sponge spicules in slices of the Irish Chert, so that in this case the Silica was probably furnished by these organisms.^ Glauconite Grains, Greensands. In some earthy organic Limestones, and in sandy rocks of a mechanical origin, we meet with grains of Glauconite or Glauconitic matter. Sometimes they occur in sufficient quantity to give a distinct green colour to the rock, which is then called a Greensand if it be of a sandy character. Accumula- tions of similar green grains have been observed on modern sea-beds. The grains are found in many cases to be casts of the shells of Foramini- fera and other hollow organic structures. It seems likely that the soluble Silica set free from siliceous organisms combined with Alumina, Ferric Oxide, Potash, and other substances present in the water, and formed this substance. The Firestone of Ventnor contains grains of Glauconite, which seem to have been formed contemporaneously, but in most cases are not moulded in the chambers of Foraminifera. Red Clay of the Atlantic. There are some deposits, apparently differing in their origin from any yet described, which the soundings of the Challenger expedition have shown are now in process of forma- tion over the very deepest parts of the bed of the ocean. The general results of these explorations of the sea-bottom are, as far as they have been at present made public, as follows. The Atlantic or Globigerina Ooze covers, as has been already men- tioned, very extensive tracts ; down to a depth of two thousand fathoms the shells retain nearly all their Carbonate of Lime, and the deposit is essentially calcareous ; beyond that depth this calcareous slime gradually becomes more and more clayey, arid passes into a deposit to which the explorers have given the name of Grey Ooze. In the Grey Ooze the shells of the Foraminifera can still be detected, but they have lost much of their sharpness of outline, assume a kind of rotten look and a brownish colour, and become mixed with a fine amorphous red-brown powder. As the depth increases the proportion of this powder grows larger and larger, the traces of calcareous matter decrease and at last disappear altogether, and at a depth averaging from two thousand one hundred to two thousand three hundred fathoms the deposit assumes the form of a Red Clay in the finest possible state of subdivision. The Red Clay is found to be a silicate of Alumina and * Sci. Trans. Roy. Dublin Soc. New Series, i. t Bull. Acad. Roy. de Belgique, [2] xlvi. See also A. Knop. Neues Jahrb. 1874, p. 281. J Aim. and Mag. Nat. Hist. [5] vii. (1881) 141. Sollas, Quart. Journ. Geol. Soe. xxviii. (1872) 399; Aim. and Mag. of Nat. Hist. (1880), 448. 234 Geology. Ferric Oxide. In volcanic regions and their neighbourhood it often contains fragments of Pumice and volcanic products. Lumps of Peroxide of Manganese, ranging in size from small particles up to good- sized nodules, also occur in it. We might be inclined at first sight to think that this red mud is, like the clayey deposits we have hitherto been dealing with, of mechanical origin ; and that it is found only at great depths and far from land, because, being very fine, it took a very long time to settle down ; that it is in fact the impalpable residue of river-borne sediment which remained in suspension after the coarser part had sunk to the bottom. But a very little consideration will show us that such an explanation will not fit in with the facts of the case. If this were the origin of the Red Clay, we ought to be able to trace a connection between it and the land by whose wear and tear the materials for it were furnished ; we ought to be able to follow it, growing gradually coarser and coarser, up to the rivers that brought these materials into the ocean. But no such connection exists ; the red mud occurs only over the deepest parts of the sea-bed, and between it and the land there lie broad tracts covered with Globigerina Ooze, and practically free from any trace of mechanical deposit whatever. Its isolation therefore proves that it did not come from land in the fashion of ordinary deposits of clay, and it must have arisen in some way or other on the areas to which it is confined. How it was formed is still far from settled ; the passage from Globigerina Ooze through Grey Ooze into the Red Clay is so insensible that it seems highly likely that the two last have been formed out of the first by the gradual removal of its Carbonate of Lime, and Professor Wyville Thomson originally sug- gested that this may have been brought about in the following manner. He believes that the Globigerina live on the surface,* and that when they die their shells sink slowly to the bottom ; as they pass downwards the Carbonate of Lime is dissolved out by the aid of the Carbonic Acid contained in the sea- water ; the greater the depth through which they sink, the longer will they be exposed to this action, and the more com- plete will be the extraction of the Carbonate of Lime ; if the depth be great enough, the Carbonate of Lime will be entirely taken away, and there will remain only the earthy insoluble portion of the shell, and this he thought was probably the material of which the Red Clay is composed. Sir W. Thomson still holds to a certain extent to his old explanation. He thinks that when organic tissues decay under water containing sundry matters in suspension or solution, the inorganic salts which exist in these tissues may combine with some of the suspended or dissolved matters, and form such compounds as this silicate of Alumina and Ferric Oxide. It is certainly in favour of his view that whereas Foraminifera and Pelagic Mollusca swarm on the surface of those parts * Mr. H. B. Brady has come to the conclusion that Foraminifera live on the bottom as well as at the surface of the Atlantic (Microscopical Journ. New Series, xix. 84), and actual dredgings show that this is so in the case of some seas. But the observations of Sir W. Thomson seem to place beyond all question their great abundance at the surface, which after all is the point on which his explanation rests. Dissolved Matters. 235 of the ocean where the bed is formed of Red Clay, scarcely a trace of their shells is found in that clay. In regions where the depth is less, the shells sink down and form Globigerina Ooze. Why does not this happen in the deepest part of the ocean 1 Why do the shells show more and more signs of decay as the depth increases, and what becomes of them in the most profound basins 1 They have somehow disappeared, and the theory of Sir W. Thomson furnishes a reasonable explanation of their absence. Mr. Murray is inclined to think that the larger part of the Bed Clay is decayed Pumice and similar volcanic products. Such substances are met with everywhere in the deep-sea deposits. They seem to have come from volcanoes on land ; they may perhaps sometimes have been shot directly into the sea, or they may have been washed down by rivers. They float for long distances, but are sooner or later decom- posed, and the products of their disintegration sink to the bottom. Sir W. Thomson is quite prepared to allow that these materials form an important element in the Red Clay, but he still holds that the source he originally suggested has contributed no inconsiderable share towards the formation of that deposit.* It seems indeed scarcely possible to ignore altogether the organic element in the formation of the Red Clay. The geological bearing of this discovery is most important, for it shows us that clayey rocks are not necessarily confined to regions adjoining land, but may be formed in the most remote and deepest parts of the ocean. Sir W. Thomson points out the possibility of some fine homogeneous clayey rocks of the earth's crust having been formed by methods similar to that which is now producing the Red Clay.t In some of the deepest soundings in the Atlantic the shells of Radiolarians preponderated to such an extent in the Red Clay that it again assumed the character of an organic deposit. The reason why the proportion of these shells increases with the depth seems to be this. Radiolarians are not largely confined to the surface like Foraminifera, but live at all depths. Hence the greater the height of the column of water that lies above any part of the sea-bed, the larger will be, cceteris paribus, the number of Radiolarian shells that fall on that part. % * Voyage of the Challenger, ii. 299. t See Nature, xi. 95, 116; xii. 174. The Voyage of the Challenger, i. 223. Also Professor Huxley, "On some of the Results of the Expedition of H.M.S. Challenger," Contemporary Review, March 1875, in which paper the reader will find an admirable summary of what was known up to that date on the subject of organic deposits, and references to original memoirs. C. W. Gumbel, Die am Grunde der Meeres volkommenden Manganknollen, Sitz. K. B. Ak. Wiss. Math. Phys. A. 1878. The Voyage of the Challenger, i. 234. 236 Geology. SECTION III. MATTER CARRIED IN SOLUTION AND THROWN DOWN BY PRECIPITATION. Of the various matters carried down to the sea or lakes dissolved in river-water Carbonate of Lime is almost always the substance present in the largest quantity, and in the last section we have seen how by the agency of animals it is rendered available for the manufacture of Limestone. But there are rocks the constituents of which occur in a state of solution in river- and sea-water, but which cannot have been formed in this way, for the simple reason that no animals are in the habit of extracting these constituents from the water. In their case then we seek for a different explanation of their origin, and the explanation that most naturally suggests itself is that the materials have been directly precipitated from solution without the intervention of organic agents. Dolomite is a good case in point. Magnesian Salts are present in sea-water, but Carbonate of Magnesia does not enter into the composition of the hard parts of any marine creatures except to an extent that is all but inappreciable. Rock-salt and Gypsum are further instances, and in their case their crystalline condition suggest precipitation very strongly ; for when substances are thrown down from solution, they frequently assume a crystalline form or structure. Means by which Precipitation may be brought about. A substance like Common Salt, soluble in pure water, can be thrown down by evaporation alone. Any solvent can dissolve only a certain quantity of the substances soluble in it. If the dissolving agent is carried away by evaporation, the proportion of dissolved matter increases, and the solution is said to become concentrated. When this has gone on till the solution holds as much as the solvent can dissolve, it is said to be saturated. If more of the solvent is removed, some of the dissolved matter must be thrown down or precipitated. In Nature this will take place if a solution of any salt is poured into a basin, and if water is carried off by evaporation faster than it is supplied. The solution will grow more and more concentrated, will become sooner or later saturated, and then precipitation will follow. Again there are matters which cannot be dissolved in pure water, but which water charged with certain substances can take up. We have had an instance of this in the solubility of Carbonate of Lime in carbonated water. In such a case, if the substance which gives water its solvent power be removed, the dissolved matter will be precipi- tated. This process goes on to some extent with every spring in Limestone districts, and very largely in the case of those springs which rise from considerable depths. During their underground course their water is under a greater pressure, and can therefore hold more Carbon Dioxide and dissolve more Carbonate of Lime than at the surface. But directly the air is reached, or the pressure is in any way lessened, Carbon Dioxide escapes and some Carbonate of Lime is thrown down. Minute subdivision of the water also favours the release of the Carbon Dioxide which escapes by diffusion. For this reason when the water Matters in Solution. 237 of calcareous springs breaks into spray at a waterfall or rapid, every- thing around becomes coated with a film of Carbonate of Lime. Further some liquids can dissolve a larger percentage of certain substances when at a high than when at a low temperature. Decrease of temperature is therefore another cause of precipitation. A fourth way in which precipitation may be brought about is very largely made use of in chemical analysis. It often happens that when solutions of two salts are mixed together, the dissolved salts are decomposed, new compounds are formed some of which are insoluble in the liquid, and these are precipitated. Suppose for instance that we mix together a solution of Carbonate of Lime in carbonated water with a solution of Sulphate of Magnesia in water. The Magnesium and Calcium change places in the manner shown in the equation below, CaCO 3 + MgSO 4 = CaSO 4 + MgCO w and Sulphate of Lime and Carbonate of Magnesia are produced. Sulphate of Lime is practically insoluble in water and is precipitated. Conditions necessary for Precipitation. If precipitation takes place in consequence of evaporation alone, the solution must be saturated. Now it is in the highest degree improbable, we might almost say impossible, that the waters of a large open ocean can ever be saturated with the substances brought down into them by solution. The amount carried in this way is very large, and while water and what- ever it holds in solution are both of them constantly pouring in, the former alone is removed by evaporation, and the proportion of dissolved matter, if there be no cause which extracts and removes it, is con- stantly on the increase. But on the other hand the bulk of the water through which this dissolved matter is to be distributed is enormous, and though the degree of concentration tends to increase as time goes on, a very long time indeed must elapse before anything like saturation can be arrived at ; and long before this time has gone by some one of those changes in physical geography, which are always going on, will come in to alter the circumstances of the case. We must also take into account the enormous amount of dissolved matter which is carried back imperceptibly by the action of the wind on the surface. Every particle of spray which is lifted into the air carries with it its salt, and, as the water evaporates, this is separated in a state so finely divided that it floats about and is blown back to land. It is partly to this cause that " sea-air " owes its bracing quality. Any one who has watched the snowlike flakes of foam that are driven far inland during a gale will realize how important a part this process plays in restoring to the land the salt which running- water has carried off into the sea, and in hindering the saturation of ocean-water. As a matter of fact in land-locked seas portions often do become considerably concentrated, but no case is known in which any- thing approaching saturation is arrived at. In the eastern part of the Mediterranean for instance the amount of dissolved matter rises to .38 and 40 parts in 1000 of water, the average of sea,- water being 34 -3 ; but there is no precipitation, the heavy salt water sinks to the bottom and flows out as an under current through the Straits of Gibraltar, 238 Geology. while an upper current brings in sea-water of average content from the Atlantic. But it is altogether different with inland bodies of water of moderate size : in their case there is the same machinery at work tending to produce concentration, and owing to the much smaller mass of water acted upon, saturation will be reached in a shorter time. In lakes which have an outlet, if the discharge is sluggish, the evaporation vigorous, and the incoming streams powerfully charged, a state of saturation may ensue and precipitates be formed ; but where there is no outlet, it is evident that the solution must grow more and more concentrated till this takes place. For similar reasons all closed bodies of waters, even if originally fresh, must become salt in time, because their feeders bring in water plus dissolved matter, and evaporation incessantly removes the first and leaves the second behind to accumulate. The Dead Sea for instance may have been once as fresh as the Lake of Tiberias : we know that in the case of the former water has been for a long time back drawn off by evaporation faster than it is poured in, because there is proof that the lake was once much bigger than it is now, and the result has been a concentration of dissolved matter in it till its present intense saltness was arrived at. From the Lake of Tiberias on the other hand water runs out as fast as it runs in, and hence it remains per- fectly fresh. Whenever then we find among the rocks of the earth's crust deposits like Rock-salt, which can have been produced only by preci- pitation, we have proof that these deposits were formed not in the open ocean, but in inland bodies of water, and the probability is very strong indeed that these bodies had no outlet. In the case of precipitation brought about in other ways it is not absolutely necessary that the solution should be saturated ; but if the deposit is large, the solution was probably at the least very highly concentrated, and concentration is much more likely to be arrived at in a closed body of water than under any other circumstances. Rocks formed by precipitation very generally possess a crystalline structure ; and this is one of the exceptions, which the student was told to expect, to the generalization that bedded rocks are non-crystalline. Instances of Rocks formed by Evaporation of satu- rated Solutions. Rock-salt. Beds of Rock-salt several hundred feet in thickness are not rare, the deposits for instance of Cheshire, Saltzburg, and Wieliezka ; Rock-salt is also met with disseminated to a greater or less extent in clay and other mechanical deposits. It can scarcely have been formed in any other way except by the evapora- tion of saturated solutions. Salt is scarcely, if at all, more soluble in hot than in cold water, so that decrease of temperature cannot have been the cause. Double decomposition is not admissible, for nearly all the compounds of Sodium are soluble in water, the few exceptions being such as are certainly not likely to occur naturally to a large amount. There are two ways in which the necessary saturated solutions may be obtained. We may have a closed lake into which rivers bring salt Matters in Solution. 239 in solution ; or a portion of sea-water may be cut off from connection with the main ocean by the formation of sandbanks or by the up- heaval of a barrier of land, and concentrated by evaporation. A layer of salt will thus be formed of no great thickness ; but if the barrier be breached and the basin refilled, and if the breach be then healed, a thing likely enough to happen if it be a sandbank, a second layer will be added, and by a repetition of this process often enough a deposit of any thickness may be obtained. The Dead Sea and the Great Salt Lake of Utah are familial- instances of closed salt lakes. Both were once considerably larger than now. We learn this by observing that along the hillsides sur- rounding both there are horizontal terraces composed of gravelly and sandy deposits such as are formed at the edges of lakes. These mark their former margins ; when several lines of terraces occur at different levels we have proof of a gradual shrinking in size interrupted by pauses. In the case of the Salt Lake, and probably in that of the Dead Sea also, there has been a decrease in the rainfall and perhaps a corresponding increase in the rate of evaporation. The diminished rainfall, and a gradual elevation of the ground around its margins, caused the level of the Salt Lake to sink below its outlet, and it became vastly diminished in size and finally closed. That the water was originally fresh is proved in the case of the Salt Lake by the pre- sence of fresh-water shells in these terrace deposits. There is very good reason to think that the Jordan originally flowed all along the Wady el Arabah and emptied itself into the Gulf of Akabah, and that the Dead Sea was then a fresh-water lake. Subsequently an elevation of the ground south of the Dead Sea has dammed the river back and closed the lake. In both cases the result has been to produce closed bodies of water in which water was removed by evaporation faster than it was brought in. In consequence the percentage of dissolved matter increased till saturation was reached, and now deposits of Rock-salt are forming on the beds of both these lakes. The following tables show the amount and composition of the average solid contents of sea-water and of the waters of these two inland seas ; in each case the first column shows the number of parts by weight of the different salts in 1000 parts of water, the second the percentage composition of the total solid contents. * AVERAGE SEA- WATER. Sodium Chloride .... 26-86 78-32 Potassium Chloride .... 0-58 ; 1-69 Magnesium Chloride .... 3-24 9-44 Magnesium Sulphate .... 2-20 6-40 Calcium Sulphate .... 1-35 3-94 Bromides, Iodides, and other Salts 0-07 0-21 34-30 100-00 * For. the Great Salt Lake see The Geological Exploration of the 40th Parallel, i. 488 ; for the Dead Sea, Lartet, Armalcs des Sciences Geologiques, i. 274. 240 Geology. Dead Sea at depth of 300 metres. Great Salt Lake. Sodium Chloride 3878 13-95 118-63 79-61 Magnesium Chloride . 170-33 61-27 14-90 10-00 Potassium Chloride 8-92 3"21 Calcium Chloride 50-34 18-10 Magnesium Bromide . Sodium Sulphate 8'70 0-95 3-13 0-34 9-32 6-24 Potassium Sulphate 5'36 3-60 Calcium Sulphate ... 0-86 0-55 278-02 100-00 149-07 100-00 Again in cases when an area is periodically inundated by the sea and then left dry, or where portions of sea-water are cut off from the main body by temporary barriers, such as shingle-banks, sand-dunes, or the upheaval of a land barrier, evaporation would produce deposits of salt ; and if the process be repeated often enough and gentle sub- sidence go on meanwhile, any thickness may be accumulated. One instance of this kind is found in the Runn of Cutch mentioned by Lyell (Elements, 6th ed. p. 446; Principles, 10th ed. vol. ii. p. 98).* The great deposits of Rock-salt in the Bitter Lakes of the Isthmus of Suez seem also to be a case in point. This bank is estimated to have contained 970,000,000,000 kilogrammes of salt, its superficial area is about 66,000,000 square metres, and it is composed of layers vary- ing in thickness from 5 to 25 centims. The basin in which this deposit lies seems to have been every now and then filled by inunda- tions from the Red Sea, and during the intervals between two succes- sive incursions evaporation concentrated the solution and threw down a layer of salt.t An excellent instance of lakes which have been once portions of the open sea and have been permanently cut off by the upheaval of land barriers is furnished by the group of inland seas and sheets of salt water, of which the Caspian and the Sea of Aral are the best known, that extends from the Sea of Azov eastwards through the low-lying plains of the desert region of Turkestan to Lake Balkash. A great portion of this district is below the sea-level, and is covered by sandy and clayey deposits, which contain recent marine shells, and there is every reason to think that at a geologically modern date a vast sheet of salt water stretched over this region and was connected with the Arctic Ocean by a long tongue running up towards the Gulf of Obi. U"p- heaval of the ground towards the north has blocked the connection with the ocean, evaporation has reduced the bulk of the water, and it is only in the deeper hollows that we now find the dwindled remnants of this old expanse of salt water. The northern and shallower part of the Caspian is so much freshened by the large volume of sweet water poured into it by the Volga that it contains only 13 parts in 1000 of dissolved salts, but in the Bay of Karasu, on the eastern shore, the * Also Sir Bartle Frere, British Assoc. Reports, 1869, Trans. Sections, p. 163. t Comptes Ren'dus, June 22, 1874. Matters in Solution. 241 amount rises to 5 6 '28 in 1000. This degree of concentration is not sufficient to produce deposition of salt, but on the south of Karasu lies the broad and shallow Bay of Kara-bogas, 3000 square miles in area. It is cut off by a large sandbank, breached by a narrow channel, and through this channel the salt water, already highly concentrated, is constantly streaming in : evaporation proceeds rapidly in this large natural salt-pan, and Rock-salt and Gypsum are deposited. Within the Aralo-Caspian area there are more than 2000 salt lakes, from one of which, Lake Elton, 200,000,000 Ibs. of Common Salt are obtained every year. Among the most important deposits which have been produced by the evaporation of saturated solutions is the great mass of Rock-salt at Stassfurt in Prussian Saxony, which is more than 200 yards in thickness. Its large depth and certain peculiarities in the arrangement of its constituents to be mentioned shortly, seem to show that this mass was deposited in a basin which was many times over filled by sea- water, cut off from connection with the sea, and subjected to evaporation. This deposit possesses a special interest because different parts of it are characterized by the presence of different salts, and these salts occur exactly in the order that would be expected if it had been formed by precipitation from a solution growing gradually more and more concentrated, the least soluble salts occurring in the lowest and first-formed portions, and the other salts coming on above in the inverse order of their solubilities. The whole series has also been here preserved, and we have a complete record of every step in the process. The deposit may be divided into three parts, which are described below in the order in which they occur, beginning at the top. 3. Rock-salt with beds and nodules of Kieserite (Hydrated Mag- nesium Sulphate), Carnallite (Hydrated Magnesium and Potassium Chloride), Polyhallite (Hydrated Magnesium, Potassium, and Calcium Sulphate), Sylvine (Potassium Chloride), Bromides, and other salts, called " Abraum Salts," i.e. salts that have to be removed before you can get to the Rock-salt : " cover," as an English quarry man would call them. 2. The main mass of the Rock-salt, consisting of beds of Rock-salt parted by layers of Gypsum, or Anhydrite, or saliferous clay. 1. Rock-salt with beds of Gypsum, Anhydrite, and Calcium Car- bonate. Now when sea-water is artificially evaporated the dissolved salts are thrown down in the following order. First, the Carbonates of Lime and Magnesia. Secondly, when the water has been reduced to about one-fifth of its original bulk, most of the Calcium Sulphate and the remainder of the Calcium Carbonate. Thirdly, after further con- centration, the larger part of the Common Salt and small quantities of Magnesium Chloride and Sodium Bromide, and a little Magnesium Sulphate. Fourthly, we obtain a highly concentrated solution or "mother liquid," occupying about ^ of the original bulk, which con- tains about 369-19 parts of dissolved matter in 1000 of water. The composition of this residue is as follows. Q 242 Geology. Sodium Chloride . . . . . 30*55 Magnesium Chloride ..... 37 '55 Potassium Chloride . . . . . 6 '30 Sodium Bromide . . . . . 3 '90 Magnesium Sulphate . . . . . 2170 100-00 If this be compared with the composition of average sea-water, it will be seen that the proportion of Sodium Chloride is vastly reduced, and that the percentages of Magnesium and Potassium Chlorides and Magnesium Sulphate have risen very considerably. It is scarcely necessary to point out how exactly the order in which the different salts follow one another in the natural deposit corresponds with the results obtained by artificial evaporation. In the lowest beds the comparatively insoluble Calcium Sulphate and Calcium Carbonate predominate. During the next stage, when Common Salt was the main precipitate, we have evidence that the basin was many times over refilled with sea-water. Every time a new supply arrived a layer of Gypsum or Anhydrite was thrown down first and a deposition of Common Salt followed ; but for a long period the concentration never went far enough to give rise to the precipitation of the most soluble salts ; whenever it approached this stage, the solution was diluted by an influx of sea-water. Occasionally the sea-water brought with it a supply of mud, and then a layer of clay was laid down in the interval between the formation of two beds of salt. Lastly the basin remained closed for so long a time that a degree of concentration was reached high enough to ensure the precipitation of the " Abraum Salts," and these are the very salts which the composition of the " mother liquor " of sea-water leads us to expect would be thrown down at this stage, namely, Chlorides and Sulphates of Magnesia and of Potash mixed with Common Salt.* Rocks formed by Precipitation consequent on the removal of a Solvent. The commonest instance of a rock of this class are the deposits, called Calcareous Tufa or Travertine, which are formed by springs holding Carbonate of Lime in solution, and they are so very common that little need be said about them. There is no Limestone district which will not furnish instances of such de- posits ; perhaps one of the best known cases is that of the so-called Petrifying Springs of Matlock, which are really incrusting springs. They rise through a thick mass of Limestone and come to the surface highly charged with Carbonate of Lime ; these waters are allowed to drip upon birds' nests, wigs, sucking pigs, and other objects which seem to commend themselves to the tourist, and these become coated over with a crust of Carbonate of Lime. The deposits from these springs have formed a terrace on the hillside on which part of the village stands. The tufa is quarried for ornamental rockwork ; it is friable and porous, and twigs of trees, snail-shells, and other land organisms are embedded in it. * Bischof, Lehrbuch der Chem. und Phys. Geologic, ii. 11. English readers will find it more convenient to refer to "River-water, Sea-water, and Rock- salt," by Justus Roth, in the Contemporary Review for August 1880, p. 233. Matters in Solution. 243 'Deposits of a similar nature are formed on a much larger scale by springs still more powerfully charged in many volcanic districts, such as Central Italy. It is by a similar process that those long rootlike pendants, called Stalactites, which hang from the roofs of Limestone caverns, are formed ; as also the sheets of Carbonate of Lime which coat their walls, and the lumps and bosses of that substance, called Stalagmites, which rise from their floors. Small hollow Stalactites may often be seen hanging from the underside of a bridge, the Carbonate of Lime of which they are formed being extracted by percolating water from the mortar.* Occasionally we meet with rocks of much greater antiquity than those just mentioned which are Calcareous Tufas : very good in- stances occur among the Tertiary strata of the Isle of Wight. In the neighbourhood of North Berwick the "Burdie House Limestone" is a very finely banded rock, closely resembling the incrustations of Car- bonate of Lime formed in boilers. It was doubtless deposited in pools into which powerful calcareous springs discharged their waters, and it was thrown down in consequence partly of evaporation and partly of loss of Carbon Dioxide. The upper part of the bed sometimes contains volcanic lapilli, so that the springs had probably a volcanic origin. Rocks formed by Precipitation consequent on de- crease of Temperature. Hot springs abound in many volcanic districts, and frequently no doubt rise from considerable depths. Water at a high temperature and under great pressure t can alone dissolve Silica, and in consequence of this power the waters of these springs are often highly charged with that substance. In many cases their waters contain alkalies, and this materially increases their power of dissolving Silica. The Silica too is doubtless often derived from the decomposition of Silicates, and Silica newly set free from combina- tion is soluble much more easily than in the crystallized state. As the water cools Silica is deposited, and forms a deposit known as Siliceous Sinter. If is often friable and porous, but when the precipitation has been slow and has occurred at intervals, it becomes compact and assumes a finely banded structure. The incrustations round the Geysers of Iceland and of the Yellow- stone Park in North-Western America J are well-known instances. Deposits formed by precipitation from siliceous springs are occa- sionally met with among the rocks of the earth's crust. On the coast near North Berwick, in the middle of a great mass of rocks of volcanic origin, there are thin beds of very finely banded siliceous stone which seem to have been formed in this manner. Rocks formed by Precipitation consequent on Chemi- cal Reactions between Dissolved Salts. We have already pointed out that organic agency cannot have produced Dolomite directly in the same way as it has furnished the material for Lime- stone. When we turn our attention to the question how Dolomite * Professor Geikie gave an account of a very striking case in Nature, x. 8. t On the influence of pressure in increasing the solvent power of liquids see Sorby's Bakerian Lecture, 1863. U. S. Geol. Survey of the Territories, Reports for 1871 and 1872. 244 Geology. was formed, and examine into the nature and surroundings of Dolo- mites and Magnesian Limestones in different parts of the world, we are struck by the fact that a vast number of the deposits of these rocks, though they may differ from one another in many respects, agree in exhibiting several very marked peculiarities. First comes the significant fact that Dolomite, Gypsum, and Rock- salt constantly occur together. Rock-salt was certainly formed by deposition from solution in closed lakes, and this suggests that com- panions so constant were probably formed in the same way. An observation of Dr. Sorby's * leads to a similar conclusion. He states that some very solid Dolomites contain even now T about one-fifth per cent, of salts soluble in water, Sodium, Magnesium, Potassium, and Calcium Chlorides, and Calcium Sulphate. These are doubtless re- tained in fluid cavities, which were formed at the same time as the Dolomite, and caught up some of the water from which it was de- posited. This water then must have been a concentrated solution, such as would be arrived at only in a closed lake. Secondly, Dolomites of this class are usually unfossiliferous ; when they do contain fossils, these are not found generally throughout the rock, but are restricted to certain localities. The forms of life too show little variety, and usually include only a comparatively small number of species ; further, the animals are often dwarfed, stunted, and deformed, as if the cir- cumstances under which they lived had prevented healthy growth. In this respect Dolomite contrasts forcibly with Limestone. Many Limestone-rocks swarm with the remains of marine animals ; the number of species is usually large, and the individuals are well-grown, healthy animals. Such Limestones rarely contain to any amount salts soluble in water, because a saturated or even highly concentrated solution cannot be arrived at in an open ocean. These facts agree well with the view that such Dolomites were thrown down from solution in closed lakes. Strong saline solutions are not desirable habitats ; only a few hardy species could live in them at all ; these would shun the more concentrated portions and would congregate where the conditions were least trying ; and even those creatures who did manage to hold out would frequently suffer from the unhealthiness of their surroundings and would fail to reach their normal size, or would give evidence by their deformity and distorted shapes of the evil plight under which they had passed their lives. Lastly the mechanical deposits of Shale and Sandstone interbedded with the associated masses of Dolomite, Gypsum, and Rock-salt are unfossiliferous and of a deep red colour. This red colour is caused by a thin film of Ferric Oxide, which coats every grain. If a bit of one of these Red Sandstones is crushed and boiled for some time in Hydrochloric Acid, the colouring matter will be dissolved, and the grains of sand will be found, after having been washed and dried, to be colourless and translucent. The water then in which these beds were deposited must have been largely charged with some compound of Iron, either mechanically suspended in a fine state of division or in solution. The colouring matter may have been brought into the lake mechanically as finely divided Ferric Oxide ; but * Report of British Association, 1856, Transactions of Sections, p. 77. Matters in Solution. 245 it is more likely that it came in solution as Ferrous Carbonate. That salt is very unstable, and in the presence of Oxygen would be con- verted into insoluble Ferric Oxide, which would be precipitated and coat each grain of sand or clay as it sank through the water. Iron is universally present in the rocks of the earth's crust, but a percentage of Ferric Oxide such as these red beds possess is exceptional. The colouring matter could hardly accumulate to this extent in an open sea or a lake with an outlet, but a closed body of water acts like a trap and retains whatever is brought into it. The view then that Dolomite, Gypsum, and Rock-salt, when they occur together, were formed in closed, bitter lakes not only accounts for the peculiarities which these rocks themselves present, but also explains equally well why the mechanical deposits interstratifiecl with them are so constantly of a deep red colour. On these broad general grounds then we infer with a high degree of probability that the Dolomites and Gypsums of which we have been speaking were formed by precipitation in inland bitter lakes in conse- quence of reactions between Magnesian and Calcareous salts held in solution by the waters of the lakes.* Methods by which Dolomite and Gypsum may have been formed. Various attempts have been made to imitate ex- perimentally the process l)y which Dolomite may be supposed to have been produced by precipitation from a solution of Salts of Lime and Magnesia ; but the efforts of chemists in this direction have not as yet been particularly successful. Several of the proposed reactions are clearly inadmissible, because, though the experimenters have succeeded in throwing down either Dolomite or a mixture of the Carbonates of Lime and Magnesia, they were obliged to employ temperatures and pressures far greater than we are at liberty to suppose prevailed during the formation of actual Dolomite, t Among the methods which naturally suggest themselves as likely to have produced Dolomite, the one that first occurs to the mind is pre- cipitation from the waters of mineral springs. Cases where Magnesian Limestones and possibly Dolomitic Limestones have been thus formed are known ; J but it is not certain that Dolomite can be formed directly in this way. Bischof's experiments rather tended to show that this was not likely. When he attempted to precipitate Carbonate of Lime and Carbonate of Magnesia together by evaporation from a solution in * There is a good account of the association of red beds with chemical deposits in the Explication de la Carte Geologique de la France, ii. 90-94. See also Nature, vi. 142, 242 ; and Ramsay, Quart. Journ. Geol. Soc. xxvii. 189, 241. t Under this head we may reckon the explanations proposed by Forchhammer, Ann. de Chem. et Phys. xxiii. ; A. Favre and Marignac. Leonhard's Jahrbuch, 1849, p. 472, Bull. Soc. Geol. de France, 2nd ser. vi. 318 ; Haidinger, Poggend. Annal. Ixxiv. 591 ; Leonhard's Jahrbuch, 1847, p. 862 ; Morlot, Leonhard's Jahrbuch, 1847, p. 862, Liebig and Kopp, Jahresbericht, 1848, ii. 500. Some of the above references are second-hand, and I have not been able to verify them. See also Bischof's Chemical Geology, iii. chap. 53 : Naumann's Geognosie, i. 523, 714 ; Zirkel, Petrographie, i. 243. Liebig and Kopp, Jahresbericht, 1853, p. 929 ; Leonhard's Jahrbuch, 1S38, p. 62, 1840, p. 372 ; Zirkel, Petrographie, i. 243 ; Naumann, Geognosie, i. 523, note, 714, note. 246 Geology. water, lie found that the first salt, because it was the least soluble, was at first thrown down almost exclusively ; then a mixture of the two salts was precipitated, and then Carbonate of Magnesia alone ; and he thought a deposit produced in this way would consist of Limestone at the bottom, possibly a little Dolomite in the middle, and Carbonate of Magnesia at the top.* It is worthy of notice however that he speaks of the strong "tendency to the formation of double salts which char- acterizes Magnesia ; " and it is possible, that even supposing the two carbonates to be precipitated separately, this tendency might lead them afterwards to rearrange themselves and form Bitter Spar. In connec- tion with this part of the subject, an observation of Sterry Hunt's also seems important. In repeating an experiment of Marignac's, he found that Carbonate of Magnesia is quite ready at the moment of its forma- tion to unite with Carbonate of Lime into Dolomite ; but when he substituted Magnesite for newly-formed Carbonate of Magnesia, he found that it showed no such aptitude to combine with the Carbonate of Lirne.t It would therefore seem to make all the difference in the world in this case, whether the Carbonate of Magnesia is in its ordi- nary condition, or whether it is newly set free from combination or newly formed; in the first case there will be no tendency to form Dolomite with any Carbonate of Lime that may be present, in the second the union with Dolomite may take place. Dr. Sterry Hunt has given special attention to the subject now be- fore us, and he has suggested two reactions, by one of which Dolomite alone, and by the other Dolomite and Gypsum, may be produced by precipitation. J Carbonate of Soda is largely produced by the decomposition of Soda Felspars and other Silicates of Soda. Carried down in solution it acts on the Calcium and Magnesium Chlorides in sea-water and produces Calcium or Magnesium Carbonate and Common Salt. + CaCl. = CaC0 3 + 2NaCl. The salts of Calcium are affected first, and Carbonate of Lime is precipitated accompanied by two or three hundredths of Carbonate of Magnesia. When this has been effected, a solution of Carbonate of Magnesia remains, which on evaporation deposits Hydrated Magnesian Carbonate. He remarks that the separation of Magnesian Carbonate does not suppose a high degree of concentration, and may have gone on when animal life was present, so that Magnesian beds formed in this way may be fossiliferous. But for the formation of any bulk within a reasonable time a degree of concentration would probably be necessary greater than that which animals can bear, and a more or less sudden destruction of the forms of life would occur, giving rise to a deposit abundantly fossiliferous. When the water had been cleared by preci- pitation, fresh individuals might migrate into it, to be in their turn destroyed and entombed when a state of saturation was again arrived at. * Chemical Geology, iii. 167, 169. t Silliman's Journ. 2nd series, xxviii. 184. Ibid. 2nd series, xxviii. 170, 365. Report on the Geology of Canada to 1863, p. 575. Matters in Solution. 247 Again Dr. Hunt lias suggested the following as a method by which Dolomite and Gypsum may be formed together. When a solution of Carbonate of Lime in carbonated water is mixed with a solution of Sulphate of Magnesia in water, double decomposi- tion ensues, and Carbonate of Magnesia and Sulphate of Lime are formed. CaC0 8 + MgSO 4 - CaSO 4 + MgCO,. Dr. Hunt found that if the solution be concentrated by evaporation, Gypsum is first thrown down, the Carbonate of Magnesia remaining dissolved. If now an additional supply of water holding Carbonate of Lime in solution be furnished, further evaporation may cause the two carbonates to fall down in a state of intermixture,* and thus a precipi- tate containing the elements of Dolomite will be obtained. It will be noticed while both these reactions give us the elements of Dolomite, neither of them produces directly Dolomite itself. The union of the carbonates into true Dolomite or Dolomitic Limestones, Dr. Hunt thinks, must have been brought about afterwards by the aid of pressure and temperature, but he states that the lowest temperature at which the combination can be effected has not been ascertained. Bearing in mind the tendency towards the formation of double salts which Magnesia is stated to exhibit, it seems not impossible that com- bination may take place slowly by simple chemical affinity without the aid of any very large amount either of pressure or heat. We cannot say then that the problem of forming Dolomite by direct precipitation has yet been solved. It has not been found possible, under the conditions which we can command, to effect this experiment- ally; but it by no means follows that such a method of formation is impossible : the ingenuity of chemists has hardly exhausted every possible combination that might lead to such a result ; and, even were this the case, it is perfectly possible that the necessary conditions may be such as we cannot imitate in our laboratories. In fact, imperfect as has been at present the success of experimenters, they have got quite far enough to justify the belief that the process consisted in some reaction between calcareous and magnesian salts in solution. What those salts were, and what was the exact nature of the reaction, have yet to be learned. If we substitute Carbonate of Barium or Strontium for Carbonate of Lime, Sulphates of these metals, that is Heavy Spar and Celestine, are formed in the place of Sulphate of Lime. Both these minerals are frequently associated with Dolomite and Gypsum. Besides the Dolomitic and Magnesian Limestones, of which we have been treating, there are others which have been formed by the altera- tion of ordinary Limestone. These will be dealt with in the chapter on Metamorphic Rocks. We have now seen our way in a dim sort of fashion to a method by which Gypsum and Dolomite might be formed together ; we have yet to explain the origin of great masses of Gypsum unaccompanied by Dolomite. Several explanations have been offered to account for such * Bischof 's experiments, described a little way back, seem against this result. 248 Geology. deposit by chemical precipitation. It is evident that, if streams hold- ing Sulphate of Lime in solution discharge themselves into a closed body of water, a saturated solution would at length be produced, and the salt would be thrown down. The precipitate might take the form either of Anhydrite or Gypsum; we do not know the conditions which determine w T hich of the two it will be, but it has been suggested that pressure will decide whether it is hydrated or anhydrous. Again it has been suggested that submarine volcanic outbursts may discharge Sulphurous Acid, which would be converted into Sulphuric Acid, and this acting on the Carbonate of Lime in solution, would give rise to Gypsum. Another suggestion is that solutions of alkaline Sulphates have been, poured into sea-water, and that mutual decomposition has gone on between their contents and the Chloride of Calcium of the sea- water, which has resulted in the formation of Sulphate of Lime and soluble alkaline Chlorides.* Other Gypsums which are probably the products of alteration of Limestone or Anhydrite will be considered under the head of Meta- morphic Rocks. Some Peculiarities of the Red Beds associated with Dolomite and Gypsum. The absence of fossils from these red beds is easily accounted for. Very few creatures can abide water stained by Ferric Oxide ; those that can get away, like fish, fly to clearer water ; the animals that live on the bottom are killed. This has been repeatedly observed to be the case when large volumes of red mud have been poured into the sea or into lakes. But Sir H. De la Beche has pointed out that though the animals which live on the sea-bed cannot exist upon a bottom of red mud, if the water above be clear, fishes could swim about in it ; t and this is the reason why the latter are found fossil in some red beds where the remains of molluscs are scarce or altogether absent. If there were intervals during which no sediment was brought down, the water might become bright enough to tempt fish into it, and an irruption of red mud might kill and bury them, and so they would be preserved in a perfect state. The remains of land plants and of terrestrial or amphibious animals are met with in the red beds ; but these have been drifted down from the adjoining land where they would not be affected by the character of the water. The surfaces of red beds are often blotched over with blue and green patches, sometimes blue and green bands are interstratified with them, and the faces of the joints, and the portions of the rock in the immediate neighbourhood of these planes of division, show the same colours. It is probable that the change in hue has been pro- duced by the decay of plants, which robs the red colouring matter of part of its oxygen and converts it into lower states of oxidation. The beds are also sometimes traversed by tubular pipes which may have been formed by the escape of gases generated during the process. J * The student may refer to Bischof, Chemical Geology, i. chap. 19; Naumann, Geognosie, i. 760 ; Zirkel, Petrographie, i. 268-273. t Memoirs of Geological Survey of Great Britain, i. 51. * De la Beche, Memoirs of Geological Survey of Great Britain, i. 53 ; Maw, Quart. Joum. Geol. Soc. ot London, xxiv. 351. Matters in Solution. 249 A change of colour produced in this way may often be noticed where peat rests on red or brown sand. The decaying organic matter reduces the Ferric Oxide or Ferric Hydrate, which stains the sand, to colour- less Ferrous Carbonate, and the sand round the edges of the peat is bleached to a pure white. The reaction may be expressed thus Fe 2 3 + 2CO, = 2FeCO 3 + O. Dr. Dawson has described a case in which red mud has had its colour discharged in a somewhat different way. Large quantities of red mud are carried into the harbour of Pictou ; the deposit, however, forming on the bottom is not red but dark grey, and seems to owe its colour to finely disseminated Sulphide of Iron ; this bottom mud too when first taken up emits Sulphuretted Hydrogen. Large quantities of sea- weeds grow on the mud flats adjoining the harbour, and it seems likely that their decay reduces the suspended Ferric Oxide and certain sulphates held in solution and produces Sulphide of Iron, which, as it is precipitated, mixes with the subsiding mud. Dr. Dawson has further suggested that certain red beds containing masses of Gypsum have been formed by the denudation of strata con- taining large quantities of Iron Pyrites. The oxidation of this mineral furnishes Ferric Oxide and Sulphuric Acid. The former gives the red stain ; the latter, acting on accumulated masses of the hard calcareous parts of animals, converts them into Anhydrite or Gypsum.* The so-called pseudomorphs of crystals of. Common Salt described on p. 82 are very common in the red beds associated with deposits of Rock-salt. The casts of the hollows produced in one bed by the disso- lution of salt crystals adhere to the under surface of the bed next above, and form cubical-shaped prominences. We also frequently find on the surfaces of these beds curious warty protuberances, which are sometimes like flattened spheres, sometimes crescent-shaped, and some- times take less regular forms. These are probably cavities produced by the dissolution of effloresced masses of salt, which were afterwards filled in by mud or sand.t Though red beds are so constantly associated with rocks formed by precipitation, we must not assume the converse to be true and assert that a red colour is in itself a proof of inland-sea origin. It is a strong presumption that way, but requires confirmation by other tests. There can be no difficulty in understanding how red beds may be formed beneath the sea. The waste produced by the denudation of red rocks will be red, and when deposited on the sea-bed will give rise to marine rocks of a red colour. And now that the Challenger sound- ings have made us acquainted with the vast deposits of red clay that are in process of formation beneath the deepest part of the ocean, the idea that all red beds are necessarily of inland-sea origin cannot be entertained for a moment. Red beds may therefore be formed under * Quart. Journ. Geol. Soc. v. 29. The whole paper is well worth careful study. t Mr. Binney has noticed these on the surface of heds of Permian Marl, Mem. of Lit. and Phil. Soc. of Manchester, vol. xii. ; Geology of the Country round Stockport (Mems. of Geolog. Survey of England and Wales), p. 36 ; see also Jahrbuch der k. k. Geol. Reichsanstalt, xxiii. 252. 250 Geology. any conditions ; and as the presence of Ferric Oxide in any quantity generally drives away animals, they will seldom contain fossils enough to enable us to determine whether they are marine or not : it is to the rocks associated with them that we must look if we want to solve this problem. Some other Rocks formed by Precipitation. Deposits of Sulphur are occasionally associated with Gypsum and other mem- bers of the class of chemically-formed rocks. One great source of Sulphur seems to be the oxidation of Sulphu- retted Hydrogen Gas, which produces Sulphur and Water. Some of the Sicilian deposits of Sulphur appear to have arisen in this way. In the middle of a mass of beds deposited in salt water there occurs a band of marly Limestone (Formazione zolfifera) which from its fossils seems to have been formed in a fresh-water lake : the Sulphur occurs in lenticular-shaped beds, from one to three metres in thickness, interpolated in this Limestone ; very finely laminated portions also occur consisting of thin alternating layers of Sulphur and Limestone. This Sulphur is not in a crystallized state. A study of what is happening in some of the lakes in the volcanic district of Central Italy enables us to explain how these Sicilian deposits were formed. There is a deep lake between Tivoli and Rome called the Solfatara di Tivoli. Its water has a temperature of 22- 24 C. (71-75 F.), and contains much Carbon Dioxide, Calcium Carbonate and Sulphate, Sulphuretted Hydrogen, Strontium, Magne- sium, and Iron. The water is in a state of constant effervescence from the discharge of gas, but is usually quite clear and of a blue colour. Every now and again however the water is rendered milky by the precipitation of Sulphur, which slowly subsides and forms deposits on the bed and margin of the lake. From this occasional turbidity the lake also goes by the name of Acqui albuli. Calcareous Tufa is at the same time thrown down and in it nodules of Sulphur occur. The periods of precipitation coincide with times of high concentration, pro- duced either by fluctuation in the proportion of supply to outflow or by rapid evaporation, and their recurrence depends on the seasons and the state of the weather. We have evidently here a deposit now in the course of formation exactly corresponding to the Formazione zolfifera of the Sicilian beds. Sulphur also occurs in other members of the Sicilian series, but not in beds ; it is found in a crystallized state lining fissures or cavities, and has evidently been introduced after the deposition of the rock; probably it is the result of the condensation of sulphurous vapours rising from volcanic sources below.* Deposits of Sulphur may also be produced in this way. If water containing Sulphates of the Alkalies or Earths come in contact with decaying organic matter, the Sulphates are reduced to Sulphides ; these are further decomposed by carbonated water and give off Sul- phuretted Hydrogen ; the oxidation of this yields Sulphur. Daubree * A. v. Lasaulx, Neues Jahrbuch, 1879, p. 490. Terrestrial Deposits. 25 1 has noticed naturally formed Sulphur in the rubbish filling up an old moat in Paris; this rubbish contains large quantities of Plaster of Paris and also much organic debris, and the mutual action of the two has resulted in the production of Sulphur in the manner just described.* There are other rocks, comparatively rare, which have been formed by chemical action ; the Trona (2Na 2 O.3CO 2 .3H 2 O) of Fezzan in North Africa for instance, and some deposits of Phosphorite, for an account of which the student must refer to treatises on Petrology. Sources of the Materials for Deposits formed by Pre- cipitation. To a certain extent these may be furnished by the disintegration of pre-existing rocks. Streams flowing over Limestone, Dolomite, and Gypsum will carry away in solution Carbonates of Lime and Magnesia and Sulphate of Lime. The decomposition of Felspars and other Silicates gives rise to soluble Silicates of Soda and Potash. The Potash is mainly taken up by plants. If the solution of Silicate of Soda enter sea- water it acts on the Calcium and Magne- sium Chlorides, and insoluble Silicates of Lime and Magnesia together with Common Salt are formed. Si0 2 .Na 2 O + CaCl 2 = SiO 2 .CaO + 2NaCl. If a solution of Silicate of Soda is mixed with a solution of Calcium or Magnesium Carbonate, we get Sodium Carbonate and insoluble Silicates of Lime or Magnesia. Si0 2 Na 2 O + CaC0 3 = SiO. 2 CaO + Na 2 CO 3 . Thus solutions of Common Salt and Sodium Carbonate are obtained, which when concentrated furnish deposits of Rock-salt, and Soda salts. But far more efficient agents in supplying the requisite ingredients are mineral springs. Mineral springs are found almost exclusively in districts which either are now or once have been the scene of volcanic action, and it is probable that in many cases the materials necessary for the formation of chemical precipitates came directly from a volcanic source. In the case of many chemical deposits we know that there were active or decaying volcanoes in the neighbourhood at the very time when these deposits were forming, and here there can be little doubt that powerfully charged springs furnished the materials required. When the disintegration of rock has yielded the supply, these rocks probably themselves drew their ingredients originally from a volcanic source. It is likely then that in all cases we must look upon volcanic action as the agent which, either directly or ultimately, furnished the materials out of which the rocks we have been treating of were manufactured. We shall have more to say on this subject in the chapter on Volcanoes. SECTION IV. TERRESTRIAL DEPOSITS. We have now dealt with that portion of the waste of denudation, by far the larger part, which is carried down into bodies of still water, * Comptes Rendus, Jan. 17, 1881. See also an account of the Sulphurous Waters of Eugheim, Armales des Mines [7] xviii. (1880) 102. 252 Geology. and have described the different ways in which it becomes arranged in bedded deposits. A certain portion however of denuded matters is a long time in making its journey, and often tarries on its way, forming accumu- lations on dry land distinguished as Terrestrial. The Terrestrial deposits of the present day are far from being insignificant ; and now that we have found so many rocks that bear a close resemblance to denuded matters which have been arranged under water, the question naturally suggests itself, whether there are any rocks approaching in the same way the accumulations of the products of denudation on dry land. We cannot reasonably expect to find among the rocks of the earth's crust many which correspond to the Terrestrial deposits now going on. These accumulations are so liable to be broken up and carried away by a continuance of the denuding processes which gave rise to them, that it is only as it were by some happy accident that they survive at all the wear and tear which the surface is always undergoing, and when they do manage to escape total destruction, as a rule only frag- ments of them are preserved. But we can imagine that if ground is let down very gently beneath water, the loose matters lying on its surface may become submerged without being destroyed, may be covered up by subaqueous deposits, and may be handed down as the relics of a land surface that has long passed away. Soil and Rain-wash. Under the present head come the de- posits of rain- wash described in the last chapter, and the surface soil formed partly by the breaking up of the underlying rock and partly by the decomposition of vegetable matter. The remains of old soils, still penetrated by the roots of plants that grew in them, and with the stools, and occasionally the trunks, of trees in the position in which they grew, are now and then found among solid rocks. One of the best known cases is the " Dirt-bed " of the Island of Portland and the adjoining coast, a section of which is given in fig. 93. The lowest beds (1) are Limestones proved by their fos- sils to have been formed beneath the sea ; on these there rests a thin band (2) of dark earth, full of angular fragments of the underlying Limestone and containing the stools of large plants allied to the modern Cycas, with here and there prostrate trunks of trees : above this come other Limestones (3), containing fossils which show them to be .of estuarine origin. The surface is formed by a "brashy" or stony soil (4), composed partly of dark vegetable matter and partly of frag- ments of the rock (3). Now even if the plants in the band (2) did not clearly bespeak its origin, we could not fail to be struck by the close resemblance which it bears to the present surface soil (4). There is no rounded or foreign stone in it, all the fragments are of the rock immediately beneath, it is as true a " brash " as the loose matter at the top of the section ; and the dark earth, in which the stones are embedded, is unmistakably vegetable soil. The roots of the plants, though their evidence is scarcely needed, furnish additional proof that we have here an old land surface ; that before the deposition of the beds (3) the lower Limestones had been raised into the air and sup- Terrestrial Deposits. 253 ported vegetable growth ; that partly by the decay of plants and partly by the atmospheric breaking up of the rock the band (2) was formed ; that the whole was then sunk beneath water in which the rocks (3) accumulated, and that the submergence was so gentle that the loose surface covering was not swept away. The lower portion of the beds (3) will be observed to be bent up over one of the stools, which projects above the surface of the " Dirt-bed." Occasionally two dirt-beds are seen in the section, showing that the process happened twice over. 254 Geology. We shall notice still more striking instances of the preservation of old vegetable soils, when we come to consider the formation of Coal. In some cases old soils have been sealed up and preserved by sheets of lava that have flowed over them. Thus in Madeira Sir C. Lyell has described red partings of Laterite or red ochreous clay between sheets of Basalt. " These red bands vary in thickness from a few inches to two or three ,i| || | Jim |||{ feet, and consist sometimes of layers of tuff, some- times of ancient soils derived from decomposed lava, both of them burned to a brick-red colour, and altered by the contact of melted matter which has flowed over them." * Similar intercalations of red earth, which also probably represent old land sur- faces, occur among the Basalts of the north-east of Ireland and of the western islands of Scotland. They consist of bands of clay and earth, usually only a few inches in thickness, of a bright red colour, and appear to be beds of soil formed by the weathering of the surface of one lava stream, which were afterwards burned to their present colour when they were overwhelmed by the next sheet of lava. Accumulations of vegetable matter, some- times converted into Charcoal and sometimes form- ing Lignite or Coal, are also met with in similar positions, and these may occasionally be observed to rest on a soil in which the roots of the plants can still be detected, t Screes. At the foot of cliffs, both inland and on the sea-coast, and on steep rocky hillsides, frag- ments of disintegrated rock accumulate in great piles of angular blocks, which are known as Screes. These are sometimes rolled by the waves and spread out in sheets of coarse shingly Conglomerate ; some- times they are covered up pretty much as they lie and give rise to Breccias. We can readily understand how deposits of this nature arise if we take a stroll over the stony up- lands of the Carboniferous Limestone of Derbyshire or Yorkshire. The hills rise by steps and terraces ; steep rocky scars run in long lines along the slopes, and on climbing one of these you step on to a level plateau which stretches away up to the foot of the next scar. Piled up against each scar lies a huge heap of Screes. The darker part in the section of fig. 94 shows the general outline of the ground. If now this country were to be gently submerged beneath water into which mud and sand were being carried, scars and Screes would gradually be buried beneath the deposits formed therein, * Elements of Geology, 6th ed. p. 639. t Judd, Quart. Journ. Geol. Soc. of London, xxx. 227. s i Irr~ Terrestrial Deposits. 255 and we should have a group of rocks consisting mainly of fine bedded Shales and Sandstones, but containing at various levels wedge-shaped tongues of Breccia and Conglomerate. This is what has actually happened far back in the south-west of England, and the general nature and arrangement of the de- posits so produced is shown in fig. 95. The country crossed by the section is a plain of red Clays and Sandstones (c\ in which there stand up every here and there isolated hills of hard Limestone (a). Every one of these hills is fringed by a bank of coarse Conglomerate and Breccia (6), known as the Dolomitic Conglomerate, made up of rounded boulders, pebbles, and angular blocks of the Limestone (a) ; in each case the Conglomerate is thickest in the imme- diate neighbourhood of the Limestone hill which it surrounds, grows thinner as we recede from that hill, and at length wholly disappears. The Conglomerates and Clays are inter- bedded in such a way that it is evident that the formation of the two must have gone on ^ ..... together, and the steps of the process must have been as follows. The country was at one time covered by a broad sheet of water, in the middle of which bosses of the Lime- stone (a) stood up as islands. Into this water rivers carried down red mud and sand, which furnished the materials for the * beds (c). On the exposed surfaces of the islands subaerial waste gave rise to an accumulation of Screes, at the same time that the red beds were being regularly laid down in the surrounding water. After a time the land sank, and the water en- croached over a portion of what had pre- viously been dry land ; the submerged part of the Screes became thus covered up by layers of red beds, and appeared as a wedge- ^ . . . . shaped mass of Conglomerate interstratified with the latter. By a repetition of this process the successive alternations of red Clay and Conglomerate were produced. Along the margin of the islands it would also happen that the waves would play upon the accumulations of debris, round its fragments into pebbles, and spread them out in layers of shingly Con- glomerate among the more quietly deposited strata of Clay. In many s = o -~ 5 O 256 Geology. instances carbonated water, percolating through the accumulation of loose blocks, has dissolved and redeposited Carbonate of Lime and Dolomitic compounds, and bound the whole into a very solid rock. The subaerial character of these Breccias is borne out by the fact that they contain the bones of two genera of terrestrial reptiles.* There are some curious brecciated deposits among the Permian rocks of Cumberland and Westmoreland, known by the name of " Brock- ranis" and "Crab Kock," some of which seem to be old Screes. t Blown Sand. Another very common form of terrestrial accumu- lations is that of Blown Sand. In many cases we find the sea-shore fringed by a belt, often of considerable breadth, of hillocks or dunes of sand, which has been dried by the wind and blown inland from the beach. Similar piling up of sand goes on in large deserts and other sandy tracts of the earth's surface. These sandy accumulations often show, when cut into, rude bedding, and the action of the wind pro- duces in them structures exactly analogous to the current-bedding and ripple-drift of subaqueous Sandstones.^ In some cases the sand is mixed with broken shells, and water, percolating through the mass, dissolves out their Carbonate of Lime and redeposits it as a cement, so that a hard calcareous Sandstone is produced. Should any of these accumulations of Blown Sand be preserved in the manner just de- scribed, it might be difficult to distinguish them from Sandstones formed beneath water, unless they happened to contain land shells or land plants in the position in which they grew. In one respect however piles of Blown Sand do differ from accumu- lations of river-borne sand formed under water. The grains are usually far more completely rounded in Blown Sand than in subaqueous Sand- stones. Both Sorby|| and Daubreell have dwelt on the angular char- acter of the grains of many Sandstones, and have given reasons why this should be so. As long as Quartz grains are rolled along the bottom by running water they do suffer wear and tear, but on account of the brittle nature of the material it is more breaking and chipping than rounding that goes on. When the grains are so far reduced in size that they can be carried suspended in the current, they suffer little alteration and retain the angularity they received from their mutual impact while travelling on the bed of the water. But when blown about by the wind, there is much more friction between the grains than when they are buoyed up in water, and their angles and edges are worn down to a much larger extent. We must also take into account the fact that under the same circumstances large grains * De la Beche, Memoirs of the Geological Survey of Great Britain, i. 240 ; Geological Observer, p. 550; Etheridge, Quart. Journ. Geol. Soc., xxvi. 174; Huxley, ibid. 42. t Quart. Journ. Geol. Soc. xx. 149, 152. For another case see Judd, Quart. Journ. Geol. Soc. xxx. 281, fig. 9, and 286. The Voyage of the Challenger, i. 309, tig. 73. Darwin, Volcanic Islands, pp. 86-88. || Anniversary Address to Koyal Microscopical Society, 1877, p. 20 ; Quart. Journ. Geol. Soc. xxxvi. (1880), Ann. Address, 58. See also J. A. Phillips, Quart. Journ. Geol. Soc. xxxvii. (1881) 6. U Etudes de Geologie Experimental, i. 248 et scq. Terrestrial Deposits. 257 are worn much faster than small ones. The amount of wear will be about proportional to the friction, which may be taken to vary as the weight and therefore as the cube of the diameter; the surface over which the wearing action goes on varies as the square of the diameter. Hence the amount worn off a grain varies as the diameter, and will be less and less as the grain decreases in size. If then the grain remained angular down to a small size, its angularity could then be reduced only very slowly indeed. The buoying up of small grains by the water has not been taken into account in this estimate ; it will evidently still farther reduce the amount of wear in minute grains. Rocks of Vegetable Origin. Perhaps the most important of terrestrial deposits are those of vegetable origin. There are plants, such as the peat-mosses, which in cold temperate climates form in swampy situations and hollows broad and thick sheets of vegetable matter known as peat-mosses. When the lower parts die, the upper surface lives on and grows upwards, and the sheet of vegetable matter continues to increase in thickness.* The peat-bogs of our own islands and the Great Dismal Swamp of Virginia are well-known instances of these vegetable accumulations. t In the Falkland Islands there are deposits of Peat formed of the roots and matted leaves and stems of several species of plants. In the upper parts the several parts of the plants are distinctly recognisable, lower down all structure is obliterated, and the whole is reduced to an amorphous carbonaceous mass. J We must now explain how such layers of dead plants have given rise to rocks. Coal. We can readily imagine an accumulation of dead land- plants, like a peat-bog, being let down gently beneath water, covered up by deposits of sediment, and preserved in the middle of a mass of bedded sand and mud. This is now admitted to have been the origin of beds of Coal, the conclusion having been come to by the following line of reasoning. As has been already explained, it has long been allowed on all hands that Coal is of vegetable origin ; but at one time great difference of opinion existed as to how the vegetable matter out of which it is formed was brought together. Some geologists would have it that Coal was an accumulation of drift plants, just as Sandstones and Shales are accumulations of drifted sand and mud. There are several very strong objections to this view. Many Coal seams extend with a fairly regular thickness over tracts hundreds of square miles in area, and it is not easy to see how such a light matter as dead wood could be spread out in even and regular layers of such great extent. Again the better kinds of Coal are nearly pure vegetable matter, and contain only a very small percentage of sandy and clayey admixtures. Such purity of composition is hardly explicable on the Drift theory, for the water that carried down the dead plants would bring also sediment, * There is a good account of Peat and its growth in a paper by Dr. James Geikie, Transactions Royal Soc. of Edinburgh, xxiv. 363. t Sir C. Lyell, Travels in North America, i. chap, vii.; Russell, My Diary North and South, i. 127. The Voyage of the Challenger, ii. 213. R 258 Geology. the two would be inevitably mixed up together, and the result would be a sort of Coal certainly, but a Coal far more earthy, and producing when burned a far larger quantity of ash, than the majority of Coals in use. It is found too in some cases that the small quantity of im- purity which Coal does contain agrees in amount and composition with the earthy portion of living plants. So that the whole of Coal, both its carbonaceous part and its ash, may have come from a vegetable source. Some other explanation had therefore to be sought for, and the first step in the right direction was made by Sir W. Logan. He pointed out that every bed of Coal rests on a peculiar clay, to which the names Underclay, Seatearth, or Warrant are given, and which may be suit- ably designated by the general term Seatstone. Seatstones vary very much in mineral composition and other respects, but they all agree in two points : they are unstratified and break up into irregular lumpy fragments, and they always contain a peculiar vegetable fossil known as Stigmaria. These Stigmaria are long, branching, cylindrical bodies, dotted over with regularly arranged pits or scars, from which long ribbon-shaped filaments run out in all directions, till the Seatstone is sometimes one thickly-matted mass of them. The Stigmaria lie parallel to the bedding, and their position and the rootlike processes that spring from them suggest naturally the idea that they are roots. That this is really their character was first proved by Mr. Binney. He had his attention called by Sir John (then Mr.) Hawkshaw to a railway cutting near Manchester, in which the trunk of a tree, very commonly found fossil in the measures associated with Coal and known by the name of Sigillaria, stood embedded in rock, erect as it grew and still connected with its roots. These roots were Stigmaria and the bed into which they struck down was a Seatstone. Many similar cases have since been observed. The mystery was now fully solved : each Seatstone is an old terrestrial soil, and the trees and plants that grew upon it, as they died and fell to the ground, formed a layer of nearly pure vegetable matter : after a time the surface was lowered beneath water, but so gently that the soft pulpy mass was not disturbed, sand and clay were laid down on the top, and the band of dead plants was thus sealed up, preserved from decay, and converted by pressure and chemical changes into a seam of Coal.* Fig. 96 shows a Sigillaria rooted in a Seatstone, and running up across the overlying beds of Shale and Sandstone. As long as the bed of the water continued to sink, deposits of mud and sand were formed one on the top of another. But after a while the sinking stopped. The water, which never attained any great depth, was now gradually filled up ; banks of sand were piled up and grew into islands, the channels between the islands were by degrees choked with sediment, till at last a swampy plain or morass was produced. * Steinhauer, American Phil. Transactions, new series, vol. i. ; Logan, Trans- actions Geol. Soc. of London, 1842 ; Binney, Phil. Mag. 1844, 1845, 1847 ; Quart. Journ. Geol. Soc. ii. 390, vi. 17 ; Transactions of Manchester Geol. Soc. i. 178 ; Bowman, ibid. i. 112 ; Brown, Quart. Journ. Geol. Soc. ii. 393 ; Geology of the South Staffordshire Coalfield (Memoirs of the Geological Survey of England and Wales), 2nd ed. p. 216 ; the references in the note to p. 185 ; and Coal, Its History and Uses (Macmillan, 1878). Terrestrial Deposits. 259 Vegetation spread out from the adjoining land, and the materials for a second bed of Coal began to accumulate. The process was repeated many times over, and there resulted a thick body of Sandstone and Shale with seams of Coal at intervals. The great swampy expanses in the delta of the Ganges and Brahmapootra, known as the Sunder- bunds and Jheels, must bear a close resemblance to the marshy flats on which Coal grew.* But as far as we can judge from the short * See Hooker, Himalayan Journals, ii. 262-265. 260 Geology. account of them given by Lady Brassey, some great accumulations of dead trees now forming on the southern coasts of Patagonia would seem to be about the nearest approach to be found nowadays to those huge sheets of vegetable matter which gave rise to seams of Coal. They are thus described. ' : To penetrate far inland was not easy owing to the denseness of the vegetation. Large trees had fallen, and, rotting where they lay, had become the birthplace of thousands of other trees, shrubs, plants, ferns, mosses, and lichens. In fact in some places we might almost be said to be walking on the tops of the trees, and first one and then another of the party found his feet sud- denly slipping through into unknown depths."* Partings and Sundry Irregularities in Coal Seams. The thickness of the mass of strata which separates two consecutive beds of Coal depends mainly on the length of the interval between two consecutive pauses. Sometimes this was so short that there was time for the deposition of only a very thin layer of mud or sand before the growth of Coal began afresh, and sometimes a number of short intervals followed one upon the other. In this way a number of beds of Coal were formed so close together that they form practically one seam, and are spoken of as one seam, the thin bands that separate the different portions being called " Partings." These partings are occasionally very variable in thickness. We h'nd sometimes that a parting, which has been a mere fraction of an inch thick over a large area, swells out in a certain direction till it becomes many feet thick. The Thick Coal of South Staffordshire for instance is in the centre of the field a mass of Coal thirty feet in thickness, and is practically a single seam ; even under this form however it is readily seen to be made up of a number of beds, marked off from one another by planes of stratification, and differing in character and quality ; as we trace the seam northwards, partings come in between the beds and thicken to the north, and in a space of five miles the single searn of thirty feet has become split up into ten Coals, which, with the measures between them, make up a thickness of 500 feet.f The thickening of a parting must have been brought about some- what in the way shown in fig. 97. The lower bed of Coal was first Fig. 97. DIAGRAM TO EXPLAIN THE THICKENING OF A PARTING IN A SEAM OF COAL. formed, and then submerged ; but the sinking gradually increased in amount from the left to the right, so that the Coal was brought into the position ab ; then deposition of sediment levelled over the in- equalities produced by unequal subsidence up to the line cd ; then another seam of Coal grew on the level floor cd. * A Voyage in the Sunbeam, chap. ix. t Geology of the South Staffordshire Coalfield (Memoirs of the Geological Sur- vey of England and Wales), p. 25. Terrestrial Deposits. 261 Again though seams of Coal extend over very large areas without showing any sensible variation in thickness and quality, even the very best and most constant Coals, when they are followed by ^ workings, sooner or later grow earthy and impure, and are split up by wedge-shaped partings which gradually in- crease in thickness till they all but cut out the Coal entirely. It is easy to see that this must be so. The swamps on which Coal grew were very extensive, but they must have ended somewhere, and their boundaries were formed by ridges or tracts rising somewhat above the general level of the swamp. Suppose that a6, cd in fig. 98 are two level tracts, over which Coal-growth is going on, separated by a rising boss of land on which little or no vegetation flourishes. Water running down the flanks of the rising ground will carry sediment on to the levels on either side, but when it reaches the flat ground the water will come to rest and the sediment will be quickly thrown down. Hence banks of sand or mud will form round the boss at the same time that Coal is growing further away from it ; but these banks will thin away rapidly as we recede from the boss, and at a little dis- tance from it the growth of Coal will go on without any ad- mixture of sediment with the vegetable matter. The forma- tion of the seam may be supposed to begin with the growth of the layer e, which thins away and ends off against the boss ; then the bank of sand/ and the layer of Coal g may be formed at the same time ; then there may be an addi- tion of Coal h on the top of g, which extends itself over the top of / up to the boss ; on the top of this another sandbank may be deposited on the right, and more Coal grow on the left. A repetition of this process will give us a thick bed of Coal free from partings on the left, which as we go to the right is split up by partings till nearly all the Coal has disappeared. The same result will be produced over the flat cd ; and when the seam comes to be worked, it will be found clean and good at a and d, but as we approach the middle of the diagram will appear to change into a mass of Shale or Sandstone containing a number of thin layers of Coal. Cases to which such an explanation will apply are of frequent occurrence.* Subaqueous Coal.t That most of our Coal seams have had their origin in the manner described is beyond question, but we do occasionally meet with Coal which has been formed under water out of masses of drift timber and * For instances see Proceed. Geol. and Polytech. Soc. West "Riding of Yorkshire, 1875, p. 68 ; North of England Institute of Mining Engineers, xxv. (1876) 13 ; the Geology of the Yorkshire Coalfield (Memoirs of the Geological Survey of England and Wales), S 20, 228, 300, 383. t These rocks ought by good rights to be placed under Section I. It is how- ever more convenient to consider them here. 262 Geology. plants carried down by rivers and buried among mechanical deposits. Such Coal occurs however rather in lenticular patches than regular beds, and is apt to be impure from a mixture of earthy sediment. In the middle of Sandstone beds little nests of Coal often occur, which must have been formed in this way. The black parts in fig. 83 are examples of these nests. It may well be that these irregular branching patches of Coal were tangled and matted masses of plants, like the Floating Island of Derwentwater, which alternately float and sink in lakes, and that during one of their sojourns at the bottom they were buried in sediment. The bark of fossil trees embedded in rock has frequently been con- verted into very bright and pure Coal. Cannel Coal. One very important variety of Coal, known as Cannel, is probably of subaqueous origin. It invariably occurs in patches thinning away to nothing on all sides, and it seems likely that each patch marks the site of a pool or lake, in which the vege- table matter lay till it was macerated into a black carbonaceous pulp. Many facts lend support to this view. The remains of fish are of constant occurrence in Cannel Coal, and they could not have got there unless the bed was formed beneath water. Beds of Cannel also pass by a gradual increase of earthy admixture into well-stratified black carbonaceous Shale; and we can readily imagine how this would come about, if a stream carried at the same time mud and plants into a lake. The heavier sediment, stained by some vegetable matter, would sink down first, the lighter wood would float to greater distances, and thus near the mouth of the river the deposit would be mainly black mud, further on in the pool the proportion of vegetable matter would increase, and at last, when all the earthy sediment had been strained out, there would be accumulations consisting almost entirely of drifted plants, which continued soaking would reduce to just such a pulp as when compacted would i'urnish a Cannel Coal. The terrestrial deposits produced by the action of ice, which are very extensive and of great importance \vill be treated of in the next section. SECTION V. DEPOSITS OF ICE-FORMED DETRITUS. Much of the waste produced by the action of ice is carried to its resting-place in the same way as the products of other denuding forces ; in some cases however ice itself acts as a carrier, and the accumula- tions thus produced differ in many important points from any we have yet considered. But whether borne away by moving ice or by running water, the deposits due to ice-action have so marked and distinctive a stamp that they may very properly be placed in a section by themselves. Distinctive Characters of Ice-borne Detritus. We will first point out what those characters are which are peculiar to ice-borne detritus, and enable us to distinguish it from that carried by rivers. The latter necessarily undergoes wear and tear and becomes more or less rounded : also running water cannot transport to any great dis- tance blocks of large size. By means of moving ice on the other hand blocks of enormous size may be carried without any rounding at all ; Deposits of Ice-formed Detritus. 263 by land-ice they may be borne away from a mountain-top across valley and hill, and dropped far away from their parent home almost as sharp and angular as when first broken off, and icebergs can float them with as little wear to still greater distances. Also the sediment carried by running water will, when it comes to rest, be arranged to a certain degree according to the size and weight of the fragments : the heavier and coarser will fall down first, and the lighter and finer will remain longer in suspension, and settle down further off from the source of supply. Such deposits will also be arranged in beds or layers. But in the case of the stuff shot over the end of a glacier, or dropped from a floating iceberg, or churned up beneath a sheet of continental ice, there will be no sorting of this kind : big blocks and fine earth will be heaped pell-mell together without regard to size or weight, and the former, instead of sinking on to their broad sides, may be packed on their smaller ends or edges : there will also be little or nothing of bedded arrangement in the deposits formed by land ice. But the most easily recognised and unmistakable sign which ice leaves of its handiwork has yet to be pointed out. The stones frozen into the under surface of a mass of moving ice, and the stones over which it passes, mutually smooth and cut into one another. Two large stones by rubbing against one another become worn flat, and often polished as thoroughly as if they had passed through a lapidary's hands : the harder and sharper stones act as cutting tools, and grind grooves in whatever they pass over, ranging in size from ruts big enough for a cart-wheel to run in, down to scratches as fine as the lines of a steel engraving. Markings like these have a peculiar stamp of their own, and no one who has once seen such can ever fail to recognise them again. As far as we know, the like are made by no agent but moving ice. Whenever then we find a deposit containing far-travelled blocks of large size and but little rounded, the materials of which are heaped together in a confused way without regard to size or weight ; which shows no bedded structure or only rude traces of bedding ; and which contains polished and scratched stones, or stones that retain traces of former polishing and scratching ; we may safely conclude that ice has been concerned in the formation of that deposit, even though the country in which it occurs cannot now nourish large masses of ice. Forms of Glacial Deposits. Deposits formed by the action of ice are called Glacial, and may be considered under the following heads : Till ; Moraines ; Glacial Mud ; subaqueous accumulations containing the droppings of Icebergs or an Ice-foot, which may be distinguished as Boulder Clays;* Erratics ; and deposits formed by the rearrangement under water of any of the preceding forms, known as Rearranged or Modified Glacial Beds. Each of these forms has certain peculiarities of its own which enable us to distinguish it from the others of its class. Till. Typical Till is a deposit of excessively tough dense clay, often gritty, stuck as full as it can hold of stones of all sizes, which are not arranged in any order, but look as if they had been forcibly rammed * This term is often used loosely for any form of glacial deposit. It may be conveniently restricted to the meaning assigned to it in the text. 264 Geology. in anyhow, and are mixed big and little indiscriminately together. Where it has to be cut through, it is more difficult to master than the hardest rock ; it can be neither broken nor blasted, and has to be laboriously worked away by spade and mattock. Every navvy who has to deal with it soon learns to recognise its formidable nature, and becomes as good a judge of what ought to be called Till as the most accomplished geologist. Many of the stones are not exactly angular, nor are they rounded, but they have their edges and angles slightly blunted, and a very large proportion of them show ice-scratching and polishing. The materials of which Till is composed are very largely derived from the rock on which it rests or from rocks in its im- mediate neighbourhood ; thus the Till of a country composed of dark clayey rocks will be dark in colour and very stiff; where the under- lying beds are of red sandstone, the Till will be reddish and lighter in character, owing to an admixture of sand. It is then less tough than if mainly composed of clay. Wherever Till is found there is always independent proof that the country has been covered by a sheet of continental ice. We have already seen that under such a sheet there is probably formed an accumulation of clay and stones known as Moraine Profonde or Grund- morane, and Till resembles exactly what we picture to ourselves that this deposit must be like. There would be weight enough above to give rise to the intense toughness and the close and irregular packing of the stones, and the scratching and polishing would be produced as the mass was pushed hither and thither by the flow of the ice. Such an explanation accords also well with the local character of Till. When an ice-sheet is descending a slope it will push or drag its Grundmorane with it and no Till will accumulate ; but when it reaches flatter ground, the power of traction will decrease, it will begin to override its Grundmorane, and Till will be left behind : and as the heterogeneous mass is swayed to and fro by the varying flow of the ice, portions will be driven into valleys and depressions, and these the ice will pass above without disturbing them further. This agrees exactly with the distribution of Till. Little is found on mountain slopes or steep hillsides ; great sheets spread over the plains : and deep valleys are blocked up by it. Though the existence of the Moraine Profonde is to a certain extent hypothetical, the probability that such an accumulation is formed beneath large ice-sheets is so great, and its character, if it exist, must be so exactly that of Till, that nearly all geologists are now agreed to look upon the latter as having been formed by the grinding and wearing away by an ice-sheet of the ground on which it rested. Another explanation of the origin of Till has been propounded by Mr. Goodchild (Quart. Journ. Geol. Soc. xxxi. 75-98). He thinks that the materials of which Glacial Deposits are composed were origin- ally embedded in the ice-sheet, and that when the ice melted its contents were gradually set free ; some were merely dropped, others more or less sorted and arranged by streams running below the ice ; and in this way he suggests many, if not all, of the various forms of Glacial Deposits may have been produced by the one single operation of the melting of the ice. Whether we agree or not with Mr. Good. OF THE Deposits of Ice-formed Detritus^ IT I V S& SIT child's conclusions, his paper will probably lead us to 1>l(ink that tfjTO^J explanation of the origin of Glacial Deposits is not altogether so simple a matter as some people have supposed, and that there are many points connected with the subject that still want clearing up. Though one of the distinguishing characteristics of Till is the pre- ponderance of stones belonging to the immediate neighbourhood where it occurs, the reader must not suppose that fragments of rock which have come from distant localities are altogether wanting in it. It is by no means uncommon to meet with far-travelled stones in Till, but as a rule they form only a minority of its contents. Further, these strangers are frequently found at much higher levels than the rock from which they were broken off. For instance the Till of the Vale of Eden contains, besides the rocks of the valley itself, many that have come from the Lake Country, and even a sensible proportion of stones that have travelled from the opposite coast of Scotland ; and these foreign materials, with others that have come from the low parts of the valley, can be traced up to the summit of the Pass of Stainmoor.* Both the presence of stones from a distance and the elevation at which they occur are easily accounted for. The gathering-ground from which the ice-sheet started that produced the Till was a long way off, and on its journey the ice picked up samples of the different kinds of rocks it passed over; these travelled on with the ice, carried on its surface or frozen into its mass, and were dropped wherever they were set free by melting or any other cause. Again, ice-sheets we know in many cases pursue their course with but little regard to the shape of the ground, are driven across valleys, and forced up the slopes of hills, and in this way it frequently happens that the stones and boulders they carry are stranded at spots very much higher above the sea-level than the source from which they were derived. Moraines. Moraines resemble Till in consisting of a confused mass of stones and earth : the whole is jumbled together in a pell-mell way without regard to size, shape, or weight, somewhat in the same way as the heaps of rubbish " tipped " to form a railway embankment. The main points of difference are these. In a Moraine the great mass of stones have ridden on the top of the ice, and hence, though they will be mostly subangular, but few will be polished or scratched. Moraine matter too having been merely shot on to or over the end of a glacier, and not pressed down by the weight of the ice, will not possess the characteristic denseness of Till. The external shape of Moraines is also peculiar : they form mounds, arranged in long lines along the flanks of a valley if the Moraine be longitudinal, or stretch- ing in horseshoe-shaped courses across a valley if it be terminal. Both Till and Moraines agree in being perfectly unstratified, and differ in this respect from the other forms of ice deposit. Glacial Mud. From beneath every sheet of ice there issue streams of water loaded with an impalpable mud, called " Flour of Rock " by Swiss geologists, the finer part of the matter ground by the ice from the rock over which it moves. When this sediment is thrown down beneath still water, it forms silty and clayey deposits of unusual * Goodclrild, Quart. Jouru. Geol. Soc. xxxi. 66, 67 ; J. Geikie, Transactions Geol. Soc. Glasgow, iv. 235. 266 Geology. fineness. The water into which glacial streams flow is too much chilled to allow of any but animals that can bear cold existing in it, and hence the fossils that occur in these clays are northern forms. Occasionally too floating ice will drop angular stones and boulders among the fine silt. A deposit then of very fine mud, containing the remains of animals that inhabit northern waters and angular blocks of rock, may safely be set down as having been formed out of sediment deposited by streams discharged from beneath a sheet of ice. Boulder Clays. Under this head we may include deposits formed beneath water, partly out of sediment held in suspension, and partly out of the droppings of floating ice. Such accumulations will be more or less bedded, but they may be distinguished from those stratified deposits, in whose formation ice had no share, by the large angular-travelled blocks which are embedded here and there in them. The beds of these deposits are often bent and twisted into very com- plicated curves. This result seems to have been produced partly by the stranding of icebergs, which, after they had run aground and ploughed into the bottom, were driven on by currents ; partly by the melting of masses of ice buried in the middle of a body of Boulder Clay, by which the beds above were deprived of support and sank down into the cavity produced by the removal of the ice-block. Erratics and Perched Blocks. We next come to those large blocks, often met with lying on the surface, which are known as Erratics or Wandered Stones.* When we can determine by any peculiarity in the rock the locality from which they have come, these Erratics are often found to have travelled far from their home, and in spite of their long journey to be angular or only very slightly rounded. This circumstance, leaving out of consideration their great size, makes it impossible that they can have been brought by water, and ice is the only agent that could have carried them. Some of them have been dropped from icebergs when the ground where they occur was beneath water ; others have been carried on the back of an ice-sheet, and stranded when the ice melted away. The latter are sometimes found in positions in which it seems at first sight quite impossible that they could have come by natural means, delicately poised on their smaller end, or balanced on the top of some projecting crag : these are called Perched blocks. They have been gently let down into their present strange positions by the gradual melting away of the ice beneath them. Fig. 99 shows a Perched Block resting on. a surface of rock that has been smoothed by ice. Rearranged Glacial Beds. All the preceding forms of glacial deposits are liable to be worked up and carried away by denudation, and either considerably modified on the spot, or carried off and rede- posited elsewhere. By such means, specially if they are transported to any distance, glacial formations lose much of their distinctive char- acter. The angular stones become rounded, the large boulders are broken up, and the scratchings and polishings are worn off. But the peculiarities we rely upon as indicating a glacial origin are not always completely wiped out, indistinct traces of ice-scratchings for instance often survive a good deal of wear and tear. By attending to such * The Germans call them Foundlings (Findlinge). How Sediment is compacted into Rock. 26 7 points we can often determine that a deposit, though it may have passed through various changes before it assumed its present form, was originally derived from a mass of ice-formed debris. Fig. 99. PEHCHED BLOCK. Rocks and Deposits of Glacial Origin. There is a large body of deposits, of immense antiquity historically, but young com- pared with the mass of the rocks of the earth's crust, in whose forma- tion ice has been in one way or another concerned. Till occurs abundantly in North Britain, Scandinavia, North America, and many other districts, and shows that these countries were once buried under sheets of continental ice. The Moraines of vanished glaciers are plentiful in many hilly and mountainous districts from which all traces of snow are now cleared off every summer ; and even in those regions, like the Alps, which still nourish perpetual snow, there are Moraines which show that the glaciers were formerly far larger than now. Associated with these indications of a former severity of climate are subaqueous Boulder Clays and deposits of Glacial mud. It is evident that all glacial formations are somewhat restricted in extent, and that the terrestrial forms are very likely to be carried away by denudation, and we should not therefore expect to find very abundant traces of them among the older portions of the earth's crust. Even here however there are Conglomerates and Breccias which there is every reason to look upon as consolidated Tills or Boulder Clays, on account of the close resemblance they bear to these deposits. SECTION VI. HOW SEDIMENT IS COMPACTED INTO ROCK. So far we have found a most perfect agreement, both in broad general character and in the minuter details of structure, between the deposits now forming by the action of denudation beneath water or on dry land, and certain of the rocks of the earth's crust. All the princi- pal kinds of the deposits that are now forming have been passed in review, and to every one we have been able to find a parallel in the class of rocks. In one respect however the two classes will usually be 268 Geology. found to differ. Modern deposits are mostly loose and incoherent, and rocks as a rule hard and compact. To this statement there are many exceptions, but still it is true in so large a number of cases, that we must, if we would make good the original identity of rock and sedi- ment, explain how the latter can be compacted into the former. Any little difficulty that this seeming want of agreement may at first sight cause will vanish if we reflect that if rocks were formed in the way suggested, their formation took place long ago, in many cases very, very long ago indeed, and if we turn over in our minds all that has happened to them since that date. The following are some of the principal causes that have had a share in the conversion of loose sediment into hard rock. Weight of Overlying Masses. In the first place when layers 'of sediment have been heaped one on the top of another till the pile reaches a great thickness, the mere weight of the mass must compress and harden the lower portion. It is this lower part we see now, for the upper beds have been removed by denudation. This cause alone would account in many cases for the solidification necessary to convert sediment into rock. Deposition of Cement. Masses of loose sediment are also traversed by percolating water, which holds in solution substances such as Carbonate of Lime or Silica. These dissolved matters will, if deposited by evaporation or any other means, act as a cement and bind the loose particles together. Chemical Reactions. Chemical reactions too go on among the constituents of sediment, and produce solidification. It is possible for instance that some soft deposits may on drying "set," like mortar or plaster of Paris. Internal Heat. We shall learn by-and-by that the interior of the earth is very hot, and we have already seen that in many cases during the deposition of sediment the mass must have gone on sinking deeper and deeper into the ground. In this way it may be brought within the range of this internal heat, and baked. The same process must go on in the neighbourhood of active volcanoes, though in this case the effect will probably be local. This solidifying cause will be more fully considered under the head of Metamorphism. Pressure. But perhaps the most important agent in the consoli- dation of sediment into rock is one whose action we can only partly explain here. It will be shown by-and-by that beds of mechanically formed rock are seldom found in the horizontal position in which they were originally deposited. They have been tilted so as to lie at all angles with the horizon, and W 7 hat it more especially behoves us to notice in connection with' our present subject, they have often been bent and folded into the most complicated curves.* A change in position like the latter can evidently have been brought about by nothing but forcible lateral pressure ; and we shall also see, when we come to consider these disturbances more fully, that at the time when it was produced the beds were not at the surface, as we see them now, but were loaded above by the weight of a great thickness of overlying rock, which has since been removed by denudation. Now it is just * See Chapter IX. Structures impressed on Rocks. 269 these powerfully folded beds, which have been not only subjected to intense lateral compression, but also pressed forcibly from above, that are the most intensely hardened. In Russia* and North America there are rocks of great antiquity, which are so little changed from the condition in which they were originally laid down that a very slight amount of weathering is enough to reduce them to their pristine state of mud, and which are hence called " Mud Stones." Bat these rocks have been scarcely disturbed at all from the horizontal position in which they ivere formed. On the other hand there are beds, in the Alps for instance, immeasurably younger, which have been solidified and even rendered crystalline to such a degree that they were for a long time assumed, in virtue of their intensely hardened state, to be of very remote date. But these rocks are invariably violently contorted. We may therefore lay it down as a broad general truth that pressure and consolidation go together ; and that where there is an absence of consolidation, there has been also an absence of pressure. Hence pres- sure must be looked upon as one of the most important agents, perhaps the most important, in the conversion of loose sediment into firm rock. It is also possible that when pressure could effect no further com- pression or change of position in a rock, the energy which had hitherto been expended in mechanical work took the form of heat, and so helped on the work of consolidation. These agencies are fully competent to effect the conversion of loose sediment into firm rock, and any sediments that were formed at bygone periods of the earth's history must have been subjected to the action of one or more of them. The only hitch in our line of reasoning has now been removed, and the conclusion is irresistible that the bedded rocks of the earths crust were once of the same nature and origin as these modern deposits, ivhich they resemble in every respect, except occasionally that of hardness. We may safely lay it down as a general rule that in a number of rocks having the same mineral composition, the oldest will probably be the most solidified, because the older a rock is the greater chance will it have had of having been subjected oftener and for a longer time to pressure and other consolidating influences. But exceptional cases, like the two mentioned a little way back, are numerous enough of rocks, on which time and its accidents have wrought scarcely any change whatever, and which now stand before us very nearly as they were spread out on the floor of the ocean untold ages ago. SECTION VII. SOME STRUCTURES IMPRESSED ON ROCKS AFTER THEIR FORMATIONS Besides the different kinds of bedding, which are a necessary conse- quence of the way in which the stratified rocks were formed, there are other peculiarities of structure which have been produced in rocks since their formation. Three of these, Slaty Cleavage, Jointing, and * Russia and the Ural Mountains (Sir R. I. Murchison), i. 78. t In connection with the subject of this section, the student will do well to study Professor Sedgwick's paper, " On the Structure of large Mineral Masses," Transactions Geol. Soc. London, 2nd series, iii. 461. 270 Geology. Concretions, can be understood by any one who has mastered the con- tents of the preceding part of this book, and may therefore be con- veniently treated of here. Others of the structures impressed on rocks subsequently to their formation will be explained further on. Structural peculiarities of this class are not confined to any one class of rocks, but are found in the stratified and unstratified alike. Slaty Cleavage. We will begin with Cleavage, because the cause to which it is due is that Pressure which we have just been talking about as one of the agents that have hardened rocks. In most cases bedded rocks split readily along the planes of bedding : but instances are not uncommon of rocks which are evidently bedded, but which cannot be induced by any means to part along the planes of bedding, while they split readily along a number of other parallel planes, which are often smooth, regular, and close together, so that the rock can be broken up almost without limit into thin plates or laminae. Roofing Slate furnishes the best possible instance, and one which any one may verify for himself. This structure is called Slaty Cleavage, and the planes of division Planes of Cleavage. Fig. 1 00 is a representation of a bit of cleaved rock. The bands of different shade and pattern are layers of different colour, hardness, Fig. 100. CLEAVAGE AND BEDDING. and mineral composition ; and the close resemblance which these show, when the mass of the rock is studied, to layers of stratification, leaves no doubt on the mind that these are the layers in which the rock was originally deposited. We can however no longer separate these layers from one another, they are firmly welded together.* But a set of fine * This, though very generally, is not universally the case : some cleaved rocks may still be split along their planes of bedding. See Jukes, Report on the Geological Survey of Newfoundland, p. 75. Structures impressed on Rocks. 271 lines are seen crossing the face of the block ; these are the edges of planes of cleavage, and along them the rock splits up readily into thin plates, one of which is shown in the drawing. The labours of several observers have given a satisfactory explana- tion of the origin of cleavage. It was noticed that fossils in cleaved rocks were distorted from their natural shape, and that the distor- tion did not take place at random, but always in the following man- ner : they were squeezed flat perpen- dicular to the planes of cleavage, and spread out along those planes. In fig. 101 a, for instance, we have a section of a Limestone containing stems of encrinites, which are cylin- ders, and show their normal circu- lar section; fig. 101 b shows a section of a similar Limestone which has been cleaved ; here the stems are flattened, and the sections of them are ellipses with their longer axes parallel to the planes of cleavage. I have heard Professor Ramsay say that many years ago, before he fully understood the meaning of the fact, he noticed that the pebbles in a Conglomerate in North Wales were flattened in a direction parallel to the cleavage planes of the Slates with which the Conglomerate is inter- bedded : a similar case of flattening of Quartz pebbles is described by Clarence King in North- Western America.* A microscopical examina- tion of thin sections of cleaved rock shows that the minute particles of the rock are flattened in a similar manner. A similar effect is seen in tig. 102, which is a section of a thin 101. SECTION OF ROCK BEFORE AND AFTER CLEAVAGE. Fig. 102. CLEAVED AND CONTORTED LIMESTONE AND SLATE. a. Limestone Band. &. Slate. band of Limestone interstratified with Slate near Ilfracombe, described by Dr. Sorby. All the beds have been bent into a series of sharp folds, and are traversed by cleavage planes, which in the drawing are * Rarnsay, Geology of North Wales (Memoirs of the Geological Survey of Eng- land and Wales), p. 145, fig. 53; Clarence King, Exploration of the 40th Parallel, ii. 399. 2/2 Geology. denoted by the fine parallel steeply inclined lines. The Slates above and below the Limestone are not so sharply bent as the Limestone itself, because being comparatively soft when the folding took place, they yielded to the pressure more than the rigid Limestone. The Limestone band, it will be noticed, varies very much in thickness, and is nipped out altogether at some spots. It was doubtless originally continuous and nearly of the same thickness throughout, and the alteration in its thickness is evidently in some way connected with the cleavage, for the bed is thinnest in those parts which run parallel to the cleavage planes, and thickest in those portions of the bends which are oblique to them. This is just what would happen if the direction of the squeeze which it has suffered was perpendicular to the cleavage planes ; for the pressure will produce its maximum effect when the bounding surfaces of the bed are perpendicular to its direction. In this case too the solubility of the Limestone has helped to produce the result ; as Dr. Sorby has elsewhere shown, mechanical pressure promotes dissolution, and the Carbonate of Lime was dissolved most largely when the Limestone lay directly athwart the pressure.* These observations showed that cleaved rocks had been compressed in a direction perpendicular to the planes of cleavage ; but they did not prove that pressure was the cause of cleavage. That final step in the argument was supplied by Dr. Sorby, and after- wards by Professor Tyndall, both of whom produced cleavage artificially in sundry substances by subjecting them to pressure, and found that the cleavage planes were always perpendicular to the direction of the pressure. Every step in the argument was now complete, and no doubt remained that a, cause of cleavage ivas pressure acting at right angles to its planes. No other means of producing cleavage has yet been hit upon, and therefore we refer all cleavage to this cause, t The reason why pressure gives rise to cleavage is not hard to see. The particles of most rocks are very irregular in shape ; a large number of them are decidedly longer in one direction than in any other, and the longer axes of these lie pointing in all directions. The effect of pressure 'will be twofold. First, it will tend to flatten the particles, and those that yield to it will be squeezed into thin plates with their flat faces all perpendicular to the direction of the pressure, that is to say all parallel to one another. Secondly, if it cannot compress a particle, pressure will tend to turn that particle round till its longest axis and flattest face are perpendicular to the direction of the pressure. The rock will now consist of grains all lying with their broadest faces parallel to one another, and it will evidently split along a plane of broad faces more easily than across them. The cleavage planes are smoother, truer, more regular, and closer together in finely grained homogeneous rocks than in those of coarse composition. Thus in fig. 100 the dotted belts are coarse and sandy, * Quart. Journ. Geol. Soc. xxxv. (1879), Anniversary Address, p. 88. t We must not however altogether, lose sight of the possibility of galvanic currents causing a laminated structure. See the experiments of Mr. R. W. Fox and Mr. T. Jordan, Reports of the Royal Polytechnic Soc. of Cornwall, No. 5, 1838, p. 68, and No. 6, 1838, p. 169 ; and of Mr. R. Hunt, Memoirs of the Geological Survey of England, i. 433. Structures impressed on Rocks. 2/3 the tinted beds fine Slate ; the cleavage planes when they enter the former become irregular, are sometimes deflected a little, and some- times lost altogether, as in the very coarse bed at the bottom. The reason for this is again obvious. If the particles are to be twisted round till their longest axes and flattest faces all lie parallel to one another, the less friction there is, the more thoroughly will re- arrangement be carried out ; and there will evidently be more friction between the rough irregular grains of a coarse gritty rock than between the soft smooth small particles of a clay. The flakiness of puff-paste is, as Professor Tyndall has shown, cleavage; and to make good puff-paste the flour must be well impregnated with butter or some kind of grease, the object being to diminish friction between the grains and facili- tate their arrangement under the pres- sure of the rolling-pin. For a similar reason where rocks have been baked by heat into a hard porcelain-like substance, they do not show cleavage. In them neither flat- tening nor rotation of the particles was possible.* A study of cleavage on a large scale confirms the conclusion that it is due to pressure. Cleavage always goes along with that bending, folding, and puckering up of the rocks, which has been already mentioned ; and it is found that the direction of the planes of cleav- age is always parallel to the axes of the large folds into which the rocks have been thrown. Fig. 103 is a bird's-eye view of a country of contorted and cleaved rock terminated by a cliff in the foreground ; on the face of the latter we see both the folds of the rocks arid also the edges of the planes of cleavage, which are denoted by the fine vertical lines. The trend of the cleavage planes is shown across the country by continuations of these lines, and the directions of the axes of the folds by the range of the several beds as they come one by one to the surface : and these two directions have the same bearing. Now the pressure that produced the folding must have acted per- * Geol. of North Wales (Mems. of the Geol. Survey of England and Wales), fig. 22, p. 97. 274 Geology. pendicular to the axes of the folds, in the direction shown by the arrows at each side of the diagram : and since the axes of the folds and the cleavage planes run parallel to one another, this is the same thing as saying that the cleavage planes are perpendicular to the pressure, to which the bending of the rocks shows them to have been subjected. A good instance of what looks like imperfectly developed cleavage may be seen in the Chalk of Freshwater Bay in the Isle of Wight. The beds are nearly on end. and have been subjected to pressure acting nearly in a horizontal direction. The Chalk is full of thin plates of fibrous Calcite lying with their flat faces parallel to the bedding ; apparently it has been first split up by a number of parallel planes of division, and then these cracks have been filled up by deposition of Carbonate of Lime. The Flints are very much shattered, and many of them show a tendency to split into thin plates whose bounding surfaces are parallel to the bedding. In both Chalk and Flints there is a ten- dency to split along planes perpendicular to the direction of the force which has brought the rock into its present position. The student must not confound slaty with crystalline cleavage. The two have points of resemblance, so much so that it was once con- jectured they might be due to the same cause. But they are essen- tially different, the one being a result of mechanical, the other of molecular forces.* Jointing. Some rocks, which go by the name of Freestones, can be cut with equal ease in all directions perpendicular to their planes of bedding ; and some of these are so valuable that it is desirable in working them to extract the blocks as near as may be of the size and shape that will be wanted, and so save loss in dressing. In this case the quarryman cuts out the pattern of his stone by picking out a shallow groove on a plane of bedding : into this groove he inserts short thick wedges, and by driving these down produces cracks, perpendicular to the planes of bedding, by which the block is detached. But if the only object is to get stone out, without being particular as to the size and shape of the blocks, all this trouble may be saved, for nature has in most cases provided cracks ready to hand of exactly a similar kind. * On Cleavage see Sedgwick, Transactions Geol. Soc. of London, 2nd series, vol. iii. p. 479 ; Rogers, Transactions of the Royal Soc. of Edinburgh, xxi. 447; Baur, Karstens u. v. Dechens, Archiv, xx. 398 ; Phillips, Reports of British Association, 1843, p. 60 ; 1857, p. 386 ; Darwin, Geological Observations on South America, chap. vi. ; Rogers, Edinburgh New Phil. Journ. vol. xxxiii. p. 144; Sharpe, Quart. Journ. Geol. Soc. of London, vol. iii. p. 74, vol. v. p. Ill; Phil. Transactions, vol. clxii. p. 445; Hopkins, Cambridge Phil. Trans- actions, viii. 456; Sorby, Edinburgh New Phil. Journ. vol. Iv. p. 137; Phil. Mag. 4th series, vol. xi. p. 20, vol. xii. p. 127 ; Quart. Journ. Geol. Soc. xxxvi. (1880), Ann. Address, 67, 72; Tyndall, Phil. Mag. 4th series, vol. xii. p. 35 and 129 ; Sir J. Herschel, ditto, p. 179 ; Haughton, ditto, p. 409 ; D. Forbes, Popular Science Review, 1870, p. 8. It will be seen by reference to the above papers that the steps by which a knowledge of the cause of cleavage was arrived at were as follows : First, the discovery of the parallelism between the strike of the beds and that of the cleavage planes ; secondly, the observation of the flat- tening of fossils and particles perpendicular to the cleavage planes ; thirdly, the artificial production of cleavage. The arrangement in the text has been adopted because it seemed to bring out the logical steps in the train of reasoning better than the order of discovery. Structures impressed on Rocks. 275 These planes of division, which are found in all rocks that have been to any extent consolidated, are known as "joints." By means of them, and the planes of bedding if the rock be bedded, it is cut up into ready-made blocks, whose size and shape depend on the arrangement of the bedding and joint-planes. Joints are noticeable in quarries, because in most cases the stone is worked off along these natural cracks, and they come to form the walls of the excavation ; they also often form the faces of natural crags, cliffs, and precipices. Fig. 104 shows a quarry where the joints are very regular and con- F-i Fig. 104. QUARRY IN JOINTED ROCK. spicuous. The nearly horizontal lines are the edges of the planes of bedding. The faces, on which the light falls, are made by a set of joints nearly parallel to one another, which traverse the body of the rock with great uniformity of trend ; another set of joints, also regular and parallel to each other, but at right angles to the first set, form the faces in shadow. Jointing of this regular character is mostly found in hard rocks of homogeneous composition, such as Limestones and thickly bedded Sandstones. There are in such cases usually two sets, the joints of each being roughly parallel to one another, and the bearing of one set is generally not far from perpendicular to that of the other set. As a rule one set is characterized by greater regularity of direction, and by its joints being continuous for longer distances, than the other set. One set generally, in the case of bedded rocks, ranges about parallel to the level line or strike, and the other set to the line of greatest slope or dip* of the beds. The strike-joints are frequently more regular in their trend and run for longer distances without break than the dip- joints. Joints w r hich keep the same direction for long distances and run through a great thickness of beds are called " Master Joints." But in many cases joints show no such symmetrical arrangement as that just described. They cross each other in all directions, change their bearing, and instead of running through a great thickness of beds, * For farther explanation of these terms see Chapter IX. 276 Geology. are confined to one bed, or change their inclination and direction in passing from bed to bed. We also find joints running in more than two directions, which cut up the rock into prismatic masses having a triangular or polygonal section. The faces of most joints are approximately plane, but we occasionally find joints with curved faces, giving rise to masses of rock with an outline like that of the side of a ship. Jointed structure is shown perhaps nowhere so distinctly as in some kinds of Coal. A block of Coal will usually be found to be divided into a number of laminse by planes parallel to the upper and under surfaces of the bed : the bed splits readily along these planes, and the surfaces of the laminse are generally dull, soft, and sooty : but the block will also be found to be cut across by two other sets of parallel planes of division, perpendicular to the bedding, and roughly perpen- dicular to one another, and the surfaces of these planes are brighter and smoother than those of the laminae. The planes of one set are more regular, true, and perfectly formed than those of the other set. In some cases these three sets of divisional planes cause the bed to break up into small cubical blocks of so regular a shape as to give one the idea that the Coal is really crystallized, such Coal is known as " Dicey Coal." There is a limit however beyond which the sub- division cannot be carried, and this is not the case with truly crystal- lized substances ; and the reason why the jointing has been so com- pletely and minutely carried out probably is because the Coal is of fairly uniform composition throughout. The more regular set of joints is known as "the face," " slyne," " cleat," or bord ; " and the other set as " the end." The compass bearing of the face often remains exactly the same over very large areas. This structure is of the utmost assistance in working Coal ; the main roads or galleries are, whenever it is practicable, driven along the " bord," and the cross cuts which connect them along the " end," the first being called " bord gates," the second "endings:" in some cases it is necessary to drive across the " face," but such an operation involves an increase of labour and expense, because the walls of the road are no longer formed by natural planes of division, but have to be hewn across solid Coal. The force that produced jointing must in some cases have been very considerable. In some Conglomerates the hardest pebbles are cut through by joints as neatly as if they had been sliced by a lapidary's wheel ; and this occasionally occurs where the matrix has been very slightly consolidated. Many conjectures have been made as to how joints have been formed. The regular way in which joints cut up a rock into blocks which have approximately the same geometrical shape, causes jointing to bear a superficial resemblance to crystalline cleavage, and it has been thought by some that jointing is a rude sort of crystallization. This theory is on the face of it unlikely. A crystal is homogeneous, all its molecules are exactly alike, and though we cannot explain fully what it is that has caused these molecules to group themselves in a definite order, we can realize that uniformity in molecular arrangement is likely to go along with uniformity in molecular composition. But Structures impressed on Rocks. 277 it is hard to see why a heterogeneous mass of particles, such as usu- ally constitute a rock, should possess a tendency to anything like crystalline arrangement. And there is an essential difference between jointing and crystalline cleavage. There is no sensible limit to the extent to which cleaved crystals may be again and again subdivided ; however small for instance may be the rhombohedron of Calcite we have obtained by cleaving a crystallized mass of that mineral, we can always break it up into similar smaller rhombohedrons, and we can carry on this process of subdivision till the resulting crystals cannot be recognised even by our most powerful microscopes, and then see no reason to think we have reached a limit ; but however close or numer- ous joints may be, we always arrive sooner or later, as we go on sub- dividing a jointed rock, at a piece of finite size with no more joints in it. It is true that in some rocks, which consist largely of a mineral which crystallizes readily, the tendency of that mineral to assume a definite form has influenced the direction of the joints, and caused them to arrange themselves rudely parallel to the faces of that form. Thus in the \vell-known Sandstone of Fontainebleau, which consists of Sand cemented by Carbonate of Lime, the tendency of the latter to crystallize in rhombohedrons has given rise to a series of joints, which divide the rock into rhombohedral masses having the same angles as the fundamental form of Calcite : * but these masses cannot like the Calcite crystal be indefinitely subdivided into similar rhombohedrons, and here as in many other cases similarity does not constitute identity. Again, it has been thought that joints are cracks produced by shrink- ing, " fissures of retreat " as it is sometimes expressed. Certain divisional planes, which are usually included under the head of joints, were formed in this way. The most important occur in rocks which have been in a fused state and have contracted in cooling ; such for instance as the fissures which in the Giant's Causeway cut up the rock into long hexagonal columns. These will be treated of when we come to Crystalline Rocks. Rocks, too, which have been in a state of mud originally, shrink and crack in drying, and it is likely enough that some of the minor joints in clayey rocks may be due to this cause. Perhaps the joints in the recently-formed Limestone of Coral reefs (see p. 229) may have been formed in this way. Further, a passage from a non-crystalline to a crystalline state involves contraction and therefore cracks, and this may be the origin of the hexagortal jointing noticed in beds of the Paris Gypsum, t But contraction cannot have produced all joints. No contraction could carry a joint through the Quartz pebbles of a conglomerate ; contraction would pull the pebbles out of the matrix. And the theory of contraction will not explain why joints run in two sets, and why one set is parallel to the dip and the other to the strike of bedded * See Naumann's Lehrbuch der Geognosie, i. 485, for references ; also Gages, Reports of British Association, 1863, p. 207. t Jukes, Student's' Manual of Geol. 3rd ed. p. 181. This jointing can have nothing to do with crystallization, for Gypsum is not hexagonal. 278 Geology. rocks. It is this last fact which leads us to the cause of many, probably of most, joints. These rocks, which now lie at all angles to the horizon, were not formed in this position. They were originally spread out in level sheets on the bed of the water in which they were deposited. We shall see further on that very powerful mechanical forces were requisite to bring them into the position in which we now see them ; and while they were being brought into this position, they underwent pressure and torsion. Stress and strain would, if carried beyond the limits of cohesion, result in fracture, and the directions of the fissures produced would necessarily bear some relation to the direction of the distorting forces, and therefore to the position in which the beds lie. That this is so and that it is to mechanical forces that we must attribute the formation of a large number of joints, is now placed beyond doubt by the experiments of Daubree. He took long plates of Gypsum and ice, fixed one of the shorter ends in a vice, gripped the other shorter end in a second vice, and gave the plate a twist. The result was the production of fissures which resemble in every respect the Master Joints of rocks. They grouped themselves in two divisions, the cracks of each division were approximately parallel to one another, and the common direction of one group made an angle varying from 70 to a right angle with that of the other group. Even in minor details the resemblance is preserved. In fig. 4, Plate I. of Daubree's work, one set of cracks is very regular in direction, the other has a tendency to form fan-shaped groups ; just the same difference is shown between the behaviour of the two sets of joints which traverse the Carboniferous Limestone of County Clare, described and figured in the Explanation of Sheets 114, 122, and 123 of the Geological Survey Map of Ireland. Daubr6e also submitted prisms of a mixture of plaster, bee's-wax, and resin to pressure, and obtained fractures and a system of minute fissures which followed the same law of arrangement as in the previous experiment.* This explanation also gives a reason why joints are most regular in homogeneous rocks ; the character of the fissures depends not only on the force which produced them, but also on the nature of the material on which that force acted. If this varies from point to point, the effects produced will show a corresponding variation ; if this be constant, the effects will everywhere be the same. It would seem then that at least two very distinct things have been included under the head of Joints. First, Shrinkage-cracks or Fissures of Retreat ; secondly, rents produced by the mechanical strain to which rocks have been subjected as they were brought from the positions in which they were formed into the positions in which we now find them. Probably all the more important and conspicuous joints in non-crystal- line bedded rocks belong to the second class. On the subject of jointing the reader may consult the following papers in addition to those already quoted : Sedgwick : Trans. Geol. Soc. of London, 2nd series, vol. iii. 461. Phillips : Reports of British Association, 1834, p. 654. ,, Trans, of Geol. Soc. of London, vol. iii. p. 1. * Etudes de Geologic Experimentale, i. p. 300 et seq. Structures impressed on Rocks. 279 Phillips : Phil. Mag. and Annals, vol. iv. p. 401. Hopkins : Report of British Association, 1838, p. 78. De la Beche : Geological Observer, p. 718. Harkness : Quart. Journ. Geol. Soc. of London, vol. xv. p. 87. Haughton : Phil. Trans, vol. cxlviii. p. 333. The Geology of North Derbyshire and the adjoining parts of York- shire (Memoirs of the Geological Survey of England), p. 143. Explanations of Sheets 114, 122, 123, and 184 of the Map of the Geological Survey of Ireland. Concretions. Balls, lumps, or nodules, of different composition from the rocks in which they are found, are common in many rocks. They are quite distinct from the pebbles in conglomerates, which were, at the time of the formation of the rock, pebbles just as they are now. The balls we are now speaking of have been formed since the rock in which they are embedded was deposited : we know this, because in many cases the lines of bedding of the adjoining rock can still be traced running through the nodule, as in fig. 105 \ and in the case Fig. 105. CONCRETIONS WITH LINES OF BEDDING PANNING THROUGH THEM. of fine clayey rocks the laminae do not bend up round the nodule, as would have been the case if it had lain as a lump at the bottom of the water from which the sediment was thrown down. That these nodules were formed after the deposition of the rock in which they occur is also proved by the occasional occurrence of fossils, one-half of which lies within and the other half outside the nodule.* Nodules of this kind are of various shapes : t sometimes spherical, at others of fantas- tic forms, but always with a rounded outline ; sometimes they are made up of a number of concentric coats, like an onion ; sometimes they have a radiated structure, i.e. they consist of long slender fibres radiating from a common centre ; sometimes the concentric and radi- ated structures occur together. A very common form, known as a Septarium, shows inside cracks and cavities, largest towards the middle and not extending to the surface, filled up with a crystallized mineral It very frequently happens that in the middle of a nodule there is a shell, plant, fish, or grain of sand : and the shape of this nucleus has evidently determined the external form of the nodule. The arrangement of these nodules generally bears some relation to the stratification, and frequently they are ranged along the planes of bedding, probably because certain beds contained the ingredients necessary for their formation, while other beds did not. * Erdraann, Geol. foren. Stockholm Forhandl. i. 204, 205, figs. 2 and 3, PI. xviii. t For a good group of figures see Dana's Manual of Geology, p. 96. 280 Geology. As instances of concretions we may mention Flints in Chalk, and the balls of Iron Pyrites and of Clay Ironstone which are common in clayey strata. The fact that nodules have been formed since the deposition of the rocks in which they are enclosed, and that they have in many cases been moulded round some body which now forms their heart, leads us to the conclusion that the matter of which they consist was once disseminated through the body of the rock in which they occur, and has been afterwards separated out and gathered into balls. We can even in some cases trace to a certain extent the steps of the process. The early stages seem to be marked by extended lines of flattened nodules, forming broken beds : a further concentration of the segre- gated matter gave rise to lumps more spherical in shape ; and occa- sionally a contraction of the interior, after an outside solid crust had been formed, produced the cracks of the Septaria, in which percolating water deposited a crystallized lining.* It has been observed in laboratory experiments that when different substances in a state of fine division are mechanically mixed together, certain of them do separate out and congregate together into nodular masses, t and it has been noticed that concretionary nodules are being formed in some rocks now in the course of deposition. It is usual to speak of this process as Concretionary Action. There is no objection to be raised to this phrase, and it, or some similar term, may be safely and conveniently used to express the fact that certain matters have been separated out of the body of a rock and collected together in balls, provided always we bear carefully in mind that by giving the process a name we do not get any nearer to understanding the man- ner in which the result has been brought about. If any one asks us what made the nodules, we may if we like say Concretionary Action ; but if the awkward question is put, What is Concretionary Action ? we should frequently be somewhat puzzled for an answer. We know that one of the ingredients of a mixture has been extracted from the sur- roundings and gathered into lumps : how exactly this was done we do not know in many cases. The term in fact is too often only a way of stating our ignorance ; and unless due precaution be taken, a somewhat dangerous way, because to certain minds it looks like an explanation. With Flints however this can no longer be said to be the case. We are, to say the least, well on the road towards definite views as to how they were formed ; and it is highly probable that in the case of the majority of the so-called concretions, the matters of which they are composed were dissolved by percolating water, carried about in solution, and collected at some spot where circumstances tending to promote deposition existed. Concretionary Structure in Rocks. Rocks themselves sometimes put on what may be provisionally called, till we know more about it, Concretionary Structure on a large scale. * De la Beche, Researches in Theoretical Geology, p. 96 ; Penning. Geol. Mag. [2] iii. 218. t Babbage, Economy of Manufactures, 2nd ed. p. 50. Structures impressed on Rocks. 28 1 A classical instance is the Magnesian Limestone of Durham, de- scribed by Professor Sedgwick. This rock, in the neighbourhood of Sunderland, is entirely made up of rounded nodular masses, and when these are loosened by weathering, it has the look of a pile of rudely- shaped cannon-balls. So complete is the separation into nodules that the rock might be mistaken for a conglomerate if it were not that the lines of bedding can still be traced running through the balls and the body of the rock alike.* Fig. 106 shows a case of large concretions in Sandstone, where the process seems to have been imperfectly carried out. l^-v '?.?>''<: Fig. 106. CONCRETIONARY STRUCTURE IN SANDSTONE. We know nothing as yet as to the way in which this structure has been produced ; but in the case of some other rocks we find a structure which closely resembles it in many respects and whose origin we can explain. In this, which we distinguish as Spheroidal Structure, the rock is made up of ball-like masses, and these are often composed of concentric coats. The bounding surfaces of these spheroids and the curved boundaries of the concentric layers are cracks produced by contraction as the rock cooled from a fused state. More will be said on this head further on (p. 307). It may be that the Concretionary Structure of the Limestone and Spheroidal Structure are due to similar causes. The Limestone may have undergone changes in chemical composition which involved a decrease in bulk, and the result may have been the production of curved cracks. It hardly seems likely however that such an explana- tion will apply to the case of Sandstones, which are constantly con- cretionary. In them sometimes the deposition of Oxide of Iron round certain centres has cemented portions of the rock into balls harder than the remainder, and so produced a concretionary appearance ; but this explanation again will not fit all cases. Oolitic Structure. There is another somewhat allied structure which may be noticed here. Many rocks, especially Limestones, are made up of rounded particles varying in size from a pin's head to a pea. There is generally some * Sedgwick, Trans. Geol. Soc. of London, 2nd series, iii. 94 and 465. See also De la Beche, Memoirs of the Geological Survey of Great Britain, i. 43, for another instance. 282 Geology. little foreign body, a grain of sand or a fragment of shell, in the middle of each ball, round which aggregation has taken place. Such rocks, when the granules are small, are called Oolites or Roestones from their resemblance to the roe of a fish : the coarser varieties are called Pisolites, or Peastones. We have already mentioned that this structure has been observed in Limestones now forming on the beaches of Coral islands out of the debris of Coral rock. In some cases it is possible to watch the way in which little grains act as nuclei, and become coated over with successive shells of Carbonate of Lime, and so enlarged into minute concretions.* Another instance of an Oolite now in the process of formation is furnished by the Sprudelstein of Carlsbad. The hot springs of that place come up largely charged with Carbonate of Lime, which is deposited in the form of Aragonite. Small granules are observed tossing about in the water, which gradu- ally increase in size till they become too heavy to be buoyed up and sink to the bottom. Here they are cemented together by further deposition of calcareous matter. Dr. Sorby has ascertained that each granule is formed round a fragment of Quartz or some such nucleus, and consists of a number of fine concentric coats. Each of these coats is composed of prismatic crystals of Aragonite arranged with their longer axes parallel to the surface of the coat. As the Carbonate of Lime is set free from solution it assumes the form of minute prisms of Aragonite ; a grain of sand as it rolls about catches up some of these, and the granule thus commenced catches up more of them, very much in the same way as a snowball rolled along the ground takes up flakes of snow, and grows larger and larger, f In cases like these the aggregation goes on at the same time as the formation of the rock, and the concretions formed differ in their mode of growth from those in which the process of separation and aggrega- tion took place after the formation of the rock was complete. Secretionary Nodules. There is a class of nodules which it is desirable to distinguish from concretions, because they have arisen in a different manner. Like some concretions they are rounded and con- sist of concentric coats ; but when they contain a hollow space inside, as is often the case, its walls are frequently lined with crystals having their vertices or bright faces turned inwards. This last fact shows that the formation of such nodules has gone on from without inwards, whereas concentric concretions were formed in the opposite direction by the successive growth of coat over coat from a central nucleus outwards. Nodules of this class may be called Secretionary or Incretionary : they have been formed by the deposition of mineral matter from per- colating water in hollow spaces in rocks : the first coat was laid down on the walls of the cavity, upon the inner surface of this another coat was deposited, and so the growth of the nodule has gone on in the direction just mentioned. Agates are a common instance of this class of nodules. Secretionary nodules occur most frequently in certain Volcanic Rocks, and further examples will be given when we come to treat of these rocks. * Dana, Corals and Coral Islands, p. 153. t Quart. Journ. Geol. Soc. xxxv. (1879), Ann. Address, p. 74. CHAPTER V. DEFINITION AND CLASSIFICATION OF DERIVATIVE ROCKS: AND HOW FROM A SJ^UDY OF THEIR CHARACTERS WE CAN DETERMINE THE PHYSICAL GEOGRAPHY OF THE EARTH A T DIFFERENT PERIODS OF ITS PAST HISTOR Y. " In these shows a chronicle survives." WORDSWORTH. OUK task in the last chapter was to inquire how the waste resulting from denudation is disposed of. We found that by far the larger part of it is ultimately laid out on the floors of bodies of still water ; and that the deposits now forming in this manner, though they differ from one another in many respects, all agree in possessing a bedded or stratified structure. We have already learned that a large class of the rocks of the earth's crust are characterized by a like bedded arrangement. Here then we had one point of resemblance between certain rocks and deposits now in the course of formation, and when we came to examine the latter more in detail, it was seen that the agreement between the two was not confined to bedding ; in fact as each kind of modern deposit was passed in review, we were able to point to some one or more of the rocks of the earth's crust, from which it differed in no respect whatever, except in certain cases that of consolidation. We were thus irresistibly led to the conclusion that bedded rocks were formed in exactly the same way as those modern deposits from which they differ in no essential respect. Having now learned how bedded rocks were produced, we can substitute for our former three- fold subdivision of them into Arenaceous, Argillaceous, and Cal- careous, a more complete classification, which will have respect not only to what these rocks are made of, but to the way in which they were formed. Derivative Rocks and their Classification. The rocks hitherto treated of owe their origin directly or indirectly to denuda- tion, and hence they may be all classed together as "Derivative." The following table shows in a concise form the conclusions we have come to as to the way in which these rocks were formed. 284 Geology. .5 o 2^ 14 P 2^2 "5 C ;-; a *< y fl -2 s'ic *^T i^ r^ : i 'S o | "t; *S & >,^ c/2 o S PH ^T- S 4 lag* 1 - m o s * S _os i ^ "= 4 i 'ill ^ | * OQ S . ^ 1 1 ^ 'i 2 ^ V; O V T E ROCKS Denudation ; s ^1 53 "y> "S S '^ -MS 1 1 II 1 1 " ^ 1 -s | ERIVA rnishcd t> JB < 2 1 | < ^3 o> )S >> "^ 1 ^ ^ 1 cs" ^ ^ ^i ^-^ Cw c^ ^ o "S3 S S ^2 <5. | S 1 c > 3 ? 'S - ston or OJ "-< O ~ ^ O d O JT" O 03 ^3 Z O O 3 3 a & i I t ,_ CQ O Q ^ (^) - 2C 2C O O H -S P5 f=5 ,JS & ^ & J K s g-s' s J^ S | 3 82 K) .^33 ^ o^? Derivative Rocks. 285 We may divide Derivative Kocks then according to the manner of their formation into 1. Mechanically formed, or clastic. 2. Chemically formed. 3. Organically formed. The first are formed of mechanically transported sediment : the second and third out of the matter carried down in solution, which is sometimes precipitated chemically, and sometimes extracted by the agency of animals or plants. If we look to the circumstances under which Derivative Rocks were formed we may class them as follows. 1. Marine : formed beneath the waters of the sea. 2. Estuarine : formed at the meeting of fresh and salt waters. 3. Lacustrine : formed in inland bodies of water. 4. Terrestrial : formed on land. Under the first head we shall have to distinguish between Littoral deposits, or those formed near the shore : the deposits laid down on parts of the sea-bottom remote from land, but still near enough to it to receive mechanical sediment, which may be called Thalassic ; and those produced at spots so far from land that little or no mechanically carried sediment finds its way to them, which may be called Oceanic. There will be two classes of Lacustrine deposits, those formed in fresh, and those in salt water. Terrestrial deposits are formed mainly by atmospheric weathering, by wind, by vegetable growth and decay, and by the action of ice. The fossil remains of animals and plants preserved in rocks often give a clue to the circumstances under which the latter were deposited. The study of fossils or Palaeontology will form the subject of a future chapter, but we shall point out here the aid they give in the matter of our present inquiries. The foregoing considerations lead us to some such broad general classification of the Derivative Rocks as is given in the following table.* GENERAL CLASSIFICATION OF DERIVATIVE ROCKS. / Littoral. Mechanical. Sandy and coarse. Variable in horizontal range and irregularly bedded. Ex. Conglomerates and Coarse Sandstones. Thalassic. Mechanical, or mixed mechanical and or- ganic. Clayey and fine. M ._ / Constant for large horizontal distances and iVJ. A K 1 .IS .& * 111111 regularly bedded. Ex. Fine Sandstones, Shales, and impure Lime- stones. Oceanic. Organic. Calcareous. Often of great horizontal extent. Ex. Pure Limestone. Altered Organic deposits. Ex. Atlantic Red Mud. h Compare Geological Magazine, v. 503. 286 Geology. Fresh-water. Mechanical. Sandy and Clayey Rocks, and impure Lime- stones. B. ESTUAKINE. I Irregular bedding with frequent changes in mineral composition. Alternations of marine, brackish, and fresh- water beds. Marine fossils often dwarfed. Mainly sandy and clayey beds and impure Limestones of mechanical origin. Organic or semi-organic occasionally. Some chemical precipitates of Carbonate of Lime and Silica. (. LACUSTRINE. Salt-water. Chemical Precipitates, such as Rock-salt, Gypsum, and Dolomite, conspicuous ; occurring in lenticular masses among red sandy and clayey mechanical deposits. Fossils rare, sometimes stunted and de- formed marine forms. Mechanical. From atmospheric weathering, Rain-wash, Screes, Old Soils. From wind D. TKKRESTUIAL. (^Eolian), Blown Sand. Organic. Mainly of vegetable origin, as Coal.* Animal deposits of Guano. N.B. Deposits formed by the aid of ice are omitted from the above table for reasons given below. One word of warning about the last column of the above table. Its object is not to specify every single one of the different kinds of rock which are somewhere or other to be met with in each subdivision ; but to point out those widely prevalent forms, which give to each group its peculiarly distinctive stamp. Thus no mention is made of deposits of a semi-organic character, such as beds of Oyster shells, which occur in the Littoral zone; nor of the rare mechanical and still rarer chemical deposits in the Oceanic area ; because cases like these are of the nature of local and subordinate accidents, which do not from a broad point of view affect the prevailing character of any one of the groups. It is the latter, and not mere accidental accompaniments, that we look to, when we want to find out the circumstances under which any given mass of rocks were formed. Importance of learning the Conditions under which Rocks were formed. The great value of a classification like that just attempted, as compared with an arrangement of rocks depending on mineral composition alone, is this ; it speaks to us of matters of far greater import than chemical and mineralogical constitution, for it asserts that rocks have not always existed as we see them now, and it assigns to each kind of rock the cause and conditions of its formation. And it is not till we have got to this point that we realize what the real aim and end of all geological work is : that it is not merely to * It is very convenient to put these rocks Tiere, though they have scarcely a right to the place, unless we stretch a point and say they are derivative, inas- much as it is denudation that furnishes the soil from which plants draw part of their food. Derivative Rocks. 287 tell us what rocks are like, but to enable us when we look at a rock to say how and where it was formed. When we can do this, Geology becomes not a mere catalogue of dry descriptions, but a history ; and we learn to look upon rocks as the pages of a volume, on which is written an account of what was going on while they were being formed. The student who knows no more of Geology than he has picked up from the preceding pages, will have begun to realize that every rock has a story of its own to tell, and furnishes to any one who can read its tale aright a record of what was the physical condition of the spot on which it is found at the time of its formation. Now in studying the physical condition of the earth at present, we do not confine our attention to any single one of its physical divisions, its land surfaces for instance or its oceans ; but we strive to learn all we can, alike about the dry land, the shallow' parts and more profound depths of the ocean, the lakes and inland seas, and in short about every one of the varied features and modifica- tions of its surface. It is the aim of Geology to furnish us with like detailed information about the earth as it was during past ages ; and as the only documents, so to speak, from which we can draw this knowledge are the rocks that were formed during those times, it is of the first importance we should be able to ascertain under what con- ditions they were formed ; because Terrestrial rocks tell us where the dry land lay, Littoral deposits mark the shore-line, Oceanic beds the depths of the sea, and Lacustrine formations give us the site of inland bodies of water. It is only when we have been able to study a con- temporaneous suite of all these different forms of rocks that we can arrive at a knowledge of the physical geography of the earth at any past epoch. We will therefore give up this chapter to an explanation of the way in which the character of a rock enables us to decide on the conditions under which it was formed ; and so to map out the different distribu- tions of land and water which have existed at different periods of the earth's past history. Teaching of Glacial Formations. Under each one of the above four main subdivisions we might have specified one or more members formed by the action of ice. Thus among the Oceanic group we do find, though but rarely, boulders dropped from icebergs; in Thalassic, Littoral, Estuarine, and Lacustrine beds, Boulder Clays and Glacial Mud are met with; while Till and Moraines are important items in the roll of Terrestrial rocks. But it will be better for our present purpose to look upon ice-formed deposits as constituting a separate glacial class, than to rank them as subordinate members of the classes already mentioned. Our object in the present chapter is to see how far we can make out, from a study of any given rock, what were the physical conditions that prevailed when that rock was formed, at the spot where it occurs. Now there is one fact which all glacial beds, under whatever conditions they were formed, agree in indicating, viz. the prevalence of intense cold ; and this fact is of far greater importance physically than the consideration whether they were terrestrial or subaqueous. While therefore these latter points must 288 Geology. not be lost sight of, the most important truth to be gathered from glacial formations is the existence of conditions that led to the accumu- lation of large masses of ice. These remarks and the description of glacial formations in the last chapter will render it unnecessary to say anything further about them in this chapter. We will now pass to a consideration of the great leading features which distinguish each of the subdivisions of our table. A. MARINE ROCKS. Littoral Rocks. The Littoral zone of any marine area consists of two parts, the belt between the limits of high and low tide, and a tract of shallow water beyond. Over the first the tides and breakers are constantly at work grinding down material detached from the cliffs or brought within the range of their action by rivers ; the bottom of the second is broken up only occasionally during very severe storms. The loose matters lying on the lower of these belts are occasionally transferred from it to the upper, but it is on the latter that the great manufacture of debris goes on ; there the wear of the waves, as they advance and retreat, produces great piles of shingle and accumulations of sand. When these are swept out seawards the finely divided parts travel far before they reach the bottom, but the coarse and heavy materials sink down at once and become heaped up in long banks of shingle and sand ranging generally parallel to the coast-line. Such banks will evidently be thickest on the side nearest the shore, and will thin away in a wedge-shaped form seawards. These materials also will be very irregularly stratified, for the currents traversing the shallow water will give rise to the structure already described as current- or false-bedding. Now among the rocks of the earth's crust we find Conglomerates and coarse Sandstones, which resemble exactly the shore deposits of the present day. In composition they are just the same they have the same wedge-shaped form, for though they may be followed for long distances in one direction, we find when we endeavour to trace them in the direction at right angles to this that they thin out rapidly and become replaced by beds of finer grain. They also invariably show very excessive current-bedding. Rocks answering to this description then give us the position of an old coast-line, and we know that the side on which they are thickest was the landward side, and that the direction in which they thin away led out to sea. The rough usage which the materials of such rocks have undergone has very frequently prevented any remains of animal life being pre- served in them, and they are generally barren of fossils. When they do contain organic remains, these consist of the hard parts of creatures that lived in shallow water, such as molluscs, whose shells in such situations grow thick and hard to enable them to resist the pounding of the shingle. They are also liable to enclose the bones of terrestrial animals and land plants, which have been brought down by rivers and have sunk to the bottom near their mouths. Derivative Rocks. 289 The deposits just mentioned are the most important and characteristic of the Littoral group, but others of a somewhat different nature are formed between tidal limits. In the hollows between shingle- and sandbanks mud and fine sand accumulate, and when the whole be- comes compacted into rock, give rise to lenticular masses of Shale and laminated Sandstone, such as often occur in the middle of bodies of Conglomerate. The surfaces of these finer beds are ripple-marked by the motion of the waves, and stamped with the tracks and burrows of marine animals and the footprints of birds ; when they are laid dry by the retreating tide, they are cracked by the sun ; sometimes too evaporation of pools of sea-water causes the deposition of Crystals of Salt, and these crystals, being afterwards dissolved, leave a cast, which is filled up by sediment, and so models in sand or mud are formed, known as Pseudomorphs.* All these appearances are common in the corresponding rocks, and where they are met with, indicate a Littoral origin. Thalassic Rocks. As we leave the shallow belt which usually fringes a sea-coast and advance into deeper water, the deposits laid down on the sea-bottom become gradually finer in grain, the sandy element, so conspicuous in the Littoral zone, ceases to predominate, and clayey mud replaces it in part : here too mixtures of mechanical sediment with the calcareous remains of marine animals are formed. Such deposits give rise to finely grained Sandstones, argillaceous Sandstones, Clays, Mudstones, Shales, and impure Limestones. These deep-water marine beds will show more regularity in their bedding than those of the Littoral zone, because the currents, to which confused bedding is due, become feebler as the water deepens : they will also spread over larger areas and be more uniform in composition, because the finely divided matter out of which they are formed remains for a long time suspended in the water and is spread over broad spaces before it sinks to the bottom. The area over which finely divided sediment suspended in river-water is distributed is increased by the smaller specific gravity of fresh than of salt water. From this cause the discharge from a river floats on the top of the sea in some cases for hundreds of miles before it becomes fairly mixed up with the salt water, and of course carries along with it its burden of suspended matter. The quietness of deposition of these beds is favourable to the pre- servation of the remains of animals, which live in the water where they are formed ; hence they will be often highly fossiliferous, but these fossils will be almost exclusively marine, and it will be only very rarely that the remains of land animals or plants will have been carried out far enough to sea to have been embedded in them. Limestones are by no means rare among the present class of rocks, but they differ from the typical Limestones of the Oceanic area in being very impure, because they are only partly made up of the calcareous portions of marine animals and contain besides mixtures of muddy or sandy sediment. It seems however possible that where rivers very largely charged with Carbonate of Lime flow into the sea, * Quart. Journ. Geol. Soc. ix. 5, 187; xxiv. 546; and p. 82 of this book. T 290 Geology. chemical precipitation may take place and give rise to beds of purer Limestone : but it is probable that such cases are very exceptional.* Normal Oceanic Rocks. In every large ocean there are bounds past which no sediment however finely divided is carried, and beyond these no mechanical deposit can consequently be formed. In the clear pure water of these regions animals flourish which cannot exist in water fouled by sediment, and by these and other organisms we saw in the last chapter masses of pure Limestone are built up. Many such Limestones occur among the rocks of the earth's crust, about whose origin there can be no doubt, because we can still see that they are almost entirely made up of the hard parts of marine animals ; other large masses of pure Limestone there are which now show little or no trace of fossils ; but to these too we assign an organic origin, because we know of no other way in which they can have been formed, and we suppose that changes, which will be more fully treated of by-and-by, have removed all traces of the fossils which they once contained. In all cases then we look upon great masses of pure Limestone as having been formed by animal agency, and as marking the sites of what were at the time of their formation Oceanic areas far remote from land. In Limestones of this character we almost always meet with siliceous nodules, as for instance Flints in Chalk, Chert nodules in the Carboni- ferous Limestone of England. The origin of these has been already explained, but it is desirable to recall attention to the almost invariable association of the two kinds of rock, because it is a fact in favour of the organic origin of the Limestone. We know that sea-w r ater holds in solution Silica as well as Carbonate of Lime, and that besides the animals and plants which secrete the latter there are others living side by side with them which extract the former ; the intermixture of siliceous and calcareous organisms readily explains the presence of Silica in the middle of an eminently calcareous deposit, a fact which it is not easy to account for in any other way. We must now include in the roll of Oceanic formations deposits like the Red Mud of the Atlantic described in the last ohapter, and rocks which may have been formed in the same way. We find now and then exceptions to the sweeping statement that Oceanic deposits are mainly of organic origin ; these are not numerous enough to upset its general truth, but still require notice. Erratics in Oceanic Deposits. Occasionally travelled boul- ders of large size are met with in the heart of great masses of strata that were formed in still water far away from any land. There are several possible means by which these wanderers may have been carried to their present position. Large stones often get entangled among the roots of trees, and when the latter fall into rivers, are floated down the stream and out to sea, till the decay of the wood drops them to the bottom. Another means of carriage is furnished by sea-weeds, which sometimes grow to a size large enough to float the rocky fragments to which they attach * See Lyell, Principles of Geology, 10th ed. i. 429-431. Derivative Rocks. 29 1 themselves. Lastly, floating ice is another transporting agent, and in all probability the one which has in most cases been employed. Where the fragments are angular they may have formed originally part of the morainic rubbish on the back of a glacier or ice-sheet, and were borne away on icebergs ; where they are rounded, they must have been picked up from the shingle of the beach by coast-ice. The reader will find a description of erratics embedded in Chalk, and a discussion of the way in which they may have been brought, in a paper by Mr. Goodwin Austen in the Quart. Journ. Geol. Soc. xiv. 252. Chemical Deposits in Oceanic Areas. Under certain cir- cumstances too chemical deposits are formed even in the centre of wide Oceanic areas. Thus Dana (Coral Islands, p. 294) gives the follow- ing section of the deposits which fill up the lagoon of an old raised atoll, Jarvis Island, situated in lat. 22' , long. 159 58' W. 3. Guano. 2. Sulphate of Lime, some compact and crystalline, some soft and amorphous, often two feet thick. 1. Fine Coral debris and shells. Here the source of the Sulphate of Lime must have been the sea- water, which holds small quantities of that substance in solution : when the lagoon became closed, evaporation would concentrate the solution till the dissolved salts were precipitated ; if a fresh supply of water were then admitted to the same treatment, and the process repeated often enough, any thickness of Gypsum might be accumulated. Dana further mentions that as far as his observations extend, all elevated lagoons have similar deposits of Gypsum, and that Rock-salt frequently accompanies them (op. cit. pp. 182, 297). Imperfect Dolomite is also formed under similar circumstances. Thus the Coral Limestone of the island of Matea contains* a large percentage of Carbonate of Mag- nesia. This salt does enter into the composition of certain Corals, t but hardly in sufficient quantity to make it possible that they could be the sole source of a rock like this. The Limestone was probably formed out of Coral debris in the drying-up lagoon of an old atoll, which had been converted by evaporation into a strongly concentrated solution of Magnesian Salts (pp. cit. p. 357). Attention has been called to these abnormal forms of Oceanic deposit, because we have already seen that Rock-salt, Gypsum, Dolomite, and other chemically formed rocks are particularly characteristic of forma- tions originating in inland seas. The cases quoted show that we must not jump too hastily to the conclusion that wherever these kinds of rock occur, the beds among which they are found are necessarily Lacustrine. If when we look at their surroundings we find them to * Analysis, Silliman's Journ. 2nd series, xiv. 82. Carbonate of Lime 61 '93 Carbonate of Magnesia Specific Gravity . Hardness . Some specimens gave only 5 '29 per cent of Carbonate of Magnesia. 38-07 2-690 4-25 t Forchhammer found 2'1 per cent, of Magnesia in Corallium rubrum, and 6 '36 in Isis hippuris (Dana, Coral Islands, p. 99). 292 Geology. be merely subordinate patches in the middle of a great mass of rocks evidently of Oceanic origin, we must decide on the conditions under which the group was deposited from the broad general character of the whole, and not from a few local accidents. B. ESTUARINE ROCKS. Everywhere along the coast materials for the formation of submarine rocks are furnished out of the detrital matter brought down by streams or yielded by the destruction of the cliffs. But where a large river enters the sea, an unusual amount of sediment is brought in at a single spot, and the accumulations round its mouth tend in consequence to become specially conspicuous. The distribution of the matters carried down by a great river will depend on the following circumstances : if powerful currents sweep across its mouth, they may bear away the whole or the greater part of the detritus, and little or no deposition may go on opposite the mouth ; but if the sea be free from currents, or if the volume or character of the suspended matter be such that the existing currents are unable to remove it, deposition will take place as soon as the river enters the sea, the latter will be gradually filled up, and a tongue of land constantly growing in size will be pushed out into the marine area. The projections of land formed in this way are known as Deltas. Tides both hinder and promote the growth of deltas. The scour of the ebb tends to sweep away sediment already deposited ; while the pounding back of the river during high tide promotes deposition. Space will not allow of our giving any lengthy description of existing deltas,* but it is desirable that the student should realize the enormous size to which they grow. The fluviatile deposits which form the delta of the Mississippi extend over an area of 12,300 miles, equal to nearly half that of Ireland, and have been proved by boring to be at one spot more than 600 feet in thickness. The delta of the Ganges is not far from twice as large. The nature of the materials of which deltas consist will vary accord- ing to circumstances. Where mountains rise abruptly from the coast, the streams that flow down their flanks will have fall enough to enable them to bring down coarse detritus, and deposits of sand and shingle will be formed around their mouths. But where, as is the case with most large rivers, a broad tract of flat country intervenes between the sea and the mountains on which the sources of the stream lie, the river ceases to be able to carry forward coarse matter as soon as it reaches the low country, and only finely divided sediment reaches the sea. Still, even in the latter case we may expect alternations of beds of different degrees of coarseness corresponding to the seasons when the river is low, and when it is in flood. In the arrangement of their materials the deposits of deltas will bear some resemblance to those of the Littoral zone among marine beds. * For information on this head see Lyell's Principles of Geology, 10th ed. i. chaps, xviii. and xix. ; De la Beche, Geological Observer, pp. 72, 98 ; Carl Vogt, Lehrbuch der Geologic, ii. 114. Derivative Rocks. 293 There will be the same current-bedding, the same interlacing of wedge- shaped masses of beds of different mineral composition, and generally the same prevailing irregularity when the whole is looked at on a large scale. We shall also find the surfaces of the beds ripple-marked, rain- pitted, sun-cracked, crossed by animal-tracks, and dotted over with pseudomorphs of salt crystals. When the surface of a delta has been raised nearly up to the sea-level, the deposits often assume a very complex character ; sand-dunes or shingle-banks are piled up, and by damming back the river- water give rise to lagoons, in which fresh- water animals live and become embedded ; after a while the sea bursts through the barrier and brings with it brackish forms whose remains are preserved in the next series of strata : sometimes the water at one end of a lagoon is salt enough to support brackish-water creatures and sufficiently freshened by the influx of river- water at the other end to allow of fresh- water animals living in it, and thus the beds laid down in it show a gradual passage from one form to the other. In some cases deposits of Rock-salt, Gypsum, and other chemical precipitates are formed by the evaporation of bodies of salt water shut off in lagoons. Occasionally the shutting out of the sea by temporary barriers gives rise to tracts of comparatively dry land, on which marsh-loving plants flourish, and in which land animals that venture on them are liable to get mired, and thus there are produced interstratified terres- trial formations with the remains of the plants that grew on them and the beasts that frequented them. In this way delta deposits show constant alternations of fresh-water, brackish, chemical, and terrestrial formations. This complexity will be vastly increased, if while the deposition of the delta is going on there are changes in the relative level of the land and sea. Suppose the sea-bottom to be sinking slowly, and that the downward movement is interrupted by occasional pauses. During one of the latter the water may be so far filled up that a land surface is produced ; when depression begins again, the terrestrial accumulations become covered- up by subaqueous deposits, and in this way any num- ber of alternations of the two forms of rock may be produced. This is the character of the fluviatile deposits on which Venice stands : they have been bored through to a depth of 400 feet, and at four different levels beds of turf precisely similar to those now forming on the margin of the Adriatic were met with. Alternate elevation and depression of the land will lead to the same admixture of terrestrial and subaqueous deposits. Shape in section of Deltas. If we could make a longitudinal section along the whole length of a delta, we should find the thickness of the deposit increasing for some distance from the mouth of the river, then beginning to decrease, and at last wedging away to nothing. Fossils of Estuarine Beds. The fossil remains preserved in estuarine beds will show a mixture very characteristic of this class of deposits. There will be no deep-sea forms, but the shells and fish that inhabit brackish water will be present ; with these drifted specimens of fresh-water and land plants and shells, and bones of terrestrial and 294 Geology. amphibious animals will occur ; occasionally we shall come upon beds enclosing only a fresh-water fauna, and others which are evidently land growths. We also meet with the shells of Estuarine or Marine mollusca, which are stunted and deformed, as if the conditions under which they lived were unfavourable to healthy growth. These abnormal forms were caused by some sudden increase in the volume of the river, whereby an area, which had for a time been occupied by salt water, became fresh- ened, not sufficiently to kill off the marine inhabitants, but enough to make their surroundings unsuitable to their habits. Deposits formed by the Union of Deltas. The deposits then of the delta of a single river will form a very complicated group, and when, in the course of their growth, the deltas of several neigh- bouring streams come to be united, we get a mass of strata showing still greater irregularity : the sediment brought down by the several rivers may vary very much in character, the prevailing constituent of the detritus of one may be mud, of another sand, and the waters of a third may be so charged with Carbonate of Lime as to promote the abundant growth of calcareous organisms and give rise to beds of Limestone ; and in this way, when the united deposits come to form one great rock mass, we shall find in it beds which at one spot are Sandstone, at another Shale, and at a third Limestone, the three forms passing horizontally into one another by gradual steps. Example of an Estuarine Group. Among the rocks of the earth's crust we find great groups of strata which show all the peculiarities just described as characteristic of delta deposits : a very good instance is furnished by what are known as the Wealden rocks of Kent and Sussex. This formation consists of Clays, Sands, Sand- stones, Calcareous Grits, and impure Limestones : it contains the remains of estuarine and fresh-water shells and crustaceans and fish, which are alone sufficient to decide its estuarine character. We learn further that it was deposited not far from land, because we find em- bedded in it land plants, insects, the bones of birds, and of terrestrial and amphibious animals, specially a gigantic terrestrial lizard known as the Iguanodon, the footprints of which still remain imprinted on the surfaces of some of the beds. We have therefore all the signs by which we recognise a formation of Estuarine origin, and we can determine also the quarter from which the river that deposited it flowed, and whereabouts the sea lay into which that river discharged itself. The thickness of the whole mass of strata in Sussex is at least 1300 feet ; as we trace them westwards along the English coast they fall off rapidly, till, at the last spot where they are exposed, they are less than 200 feet thick. Again the corresponding beds on the opposite coast of France show a still more striking decrease in thickness in the direction of Boulogne. The spot where the beds are thickest was evidently in the middle of the estuary ; the fact that they thin away both to the east and the west shows that the water shallowed in those directions, or, in other words, that the margins of the estuary lay Derivative Rocks. 295 towards those quarters. The estuary therefore in which these beds were deposited stretched across Sussex, and its shores lay to the east and west of that county, that is, its general direction ran north and south. That the sources of the river were to the north, and that the ocean into which it discharged itself lay to the south, we learn from the following considerations. If we cross the Channel and examine the corresponding rocks in France, we find the group to consist there of alternations of beds decidedly estuarine with others undoubtedly marine, of the class we have called Thalassic ; as we go towards the south-east the estuarine portions become fewer and thinner till they at last disappear altogether, and at the same time the marine beds gradually lose the Thalassic type and pass into Oceanic Limestones. We have therefore a gradual and complete passage from beds formed at the rnouth of a river, through alternations of Estuarine and Marine strata, into rocks formed in an open ocean. These broad facts show that land lay to the north and open sea to the south-east ; and by the aid of more detailed observations, which need not be given here, we can restore to a very close degree of approximation the physical geography of the country at the time this group was being formed. To the north lay a tract of land covered with vegetation and inhabited by the Iguanodon and other creatures : one of the rivers draining this continent discharged itself through a long narrow estuary, which ran in a south-easterly direction across the south-east of England towards the centre of France, and opened out there into a broad ocean. The position of this estuary is marked out by the great mass of beds, almost entirely of fresh-water origin from top to bottom, in the south-east of England ; where we find in France marine beds begin to come in, we know that we have passed the mouth of the estuary and are getting out to sea ; and when we find these marine beds gradually losing their Thalassic character and putting on an Oceanic type, we know that we are well in the marine area. One more point calls for notice, the character of the Wealden beds, even where they are thickest, proves that they have been deposited in not very deep water. The only way in which this could happen was by a gradual sinking of the land during the period of their deposition. We have independent evidence that such a sinking did take place. Immediately above the Wealden beds there lies a thin group of strata known as the Punfield Formation, which consists of alternations of fresh-water and Littoral Marine strata : during the formation of these rocks then the sea must have from time to time encroached on the area formerly occupied by fresh water. Above these Punfield beds are others known as the Lower Greensand, which are purely marine ; and from this we learn that by the time these last came to be formed, the sea had permanently overflowed the country. The evidence there- fore all conspires to show that during the formation of all the rocks we have been reviewing the land was going down ; that during the Wealden period the sea was seldom, if ever, brought over the south- east of England ; that during Punfield times it advanced over part of 296 Geology. that district and receded ; and finally that the country was completely submerged during the deposition of the Lower Greensand. LEO 1 1 I 2 K -r K s ^ 'S i |. c I (J l h ^ I- K <; H . K ~ " ~ =. ..- * r The relations to one another of the rocks just described have been thrown into the form of a diagram in fig. 107. Derivative Rocks. C. LACUSTRINE KOCKS. The deposits formed in fresh-water lakes and those have some points in common, which may be considered before we come to the characters which are peculiar to each. The sediment brought into a lake is usually supplied by several rivers, which enter it at different points, which may run over rocks of very different character, and may vary much in their transporting powers. From this cause Lacustrine deposits will show both in a horizontal and vertical direction very marked and often very sudden changes of character. The coarser matter will be thrown down in deltas at the mouth of each stream, and thus fan-shaped masses of Conglomerate will accumulate every here and there along the edge of the water. When we have to deal with a mass of Lacustrine beds, we may, by noting the position of these deltas, fix the boundaries of the sheet of water in which it was accumulated. The more finely divided materials will travel further, and will to a certain extent get mixed together, before they reach the bottom, and thus the central parts of the deposit will be more uniform in character ; but even here there may well be numerous alternations of beds differing in colour, composition, and texture, for it is probable that all the inflowing streams will not be at their fullest at the same time, and the one which has the greatest- volume and velocity will bring in and spread further its own peculiar sediment, and give rise to a layer, which will partake more largely of the character of the sediment of that stream than of the others. When the turn of the next stream comes, it will lay on the top of this a stratum, in which the distinctive character of the sediment which it brings down will prevail, and so on. Many Lacus- trine formations do show numerous alternations of thin beds of different characters, and it is probable that it is for the reason just given. Also we must bear in mind that the streams cannot carry even fine sediment beyond a certain distance from their mouths, and hence the beds will not spread each of them over the whole of the bottom, but will dovetail into one another in a wedge-shaped fashion. The peculiarities just described are well exemplified in the Lacustrine deposits of Auvergne, which the reader will find described in Lyell's " Manual of Geology" (6th ed.), p. 220. When the water of a lake is low, the surfaces of the deposits forming in it are sometimes laid dry, and then become impressed with rain-pittings, sun-cracks, and other such markings, which we have already seen are produced in other deposits under similar circumstances. When the streams which feed a lake are small, each will be able to bring down only a small quantity of sediment ; and if this is finely divided, it will be spread over a large area, and give rise to a layer or stratum of small thickness. A change in the character of the detritus will lay upon the top of this a stratum equally thin, but of different composition. Thus the deposit will be subdivided into a large number of very thin beds, and will contrast strongly with the more thickly- bedded and uniform accumulations of a Thalassic or Oceanic area. Fig. 77, p. 166, gives an instance of this. In the upper fifteen feet there 298 Geology. are no less than nine alternations of rock, each occurring in thin layers, which in some cases are further subdivided into laminae of excessive tenuity ; this portion of the section is of Lacustrine or Estuarine origin. The lower twenty-three feet, which consists of Marine rocks, shows only four subdivisions more massive and blocky in their structure. Fresh- water Lacustrine Deposits. Such being the general character of all Lacustrine formations, we must next consider what are the peculiarities which enable us to distinguish those deposited in fresh- water lakes. All Lacustrine beds resemble those of Estuarine origin in many respects ; they show the same general irregularity both in the composition and arrangement of the deposit, give the same proofs that the surfaces of the beds have been occasionally exposed to the air, and contain the remains of fresh-water and terrestrial plants and animals. But there is this difference between the two : Estuarine formations usually contain beds with brackish water or marine shells interstratified with those in which fluviatile forms alone occur, or beds yielding a mixture of marine and fresh-water forms ; such are of course absent from the deposits of a lake to which the sea never gains access. But in many cases the sheets of fresh water in which Lacustrine formations have been laid down have been from time to time invaded by the sea, and the result has been just such alternations of fresh- and salt-water deposits as we meet with among Estuarine beds. In a case like this, if only fragments of the deposit have been preserved, it will be im- possible to say to which class they should be referred ; but if the formation has come down in anything like an entire condition, the following considerations enable us to decide this point. Estuarine formations will pass in a certain direction into those of a purely marine origin ; we found this to be the case for instance with the Wealden rocks of England. In deposits formed in bodies of fresh water, though there may be marine intercalations, we shall never observe the forma- tion as a whole to pass laterally into one entirely marine. Further, if a Lacustrine deposit be entire, we shall find all round its edges a ring of shore formations, Conglomerates and similar coarse rocks, among which the deltas of the inflowing streams will be specially conspicuous : in a delta the similar rocks will tend to be crowded round one spot, viz. the river's mouth. In the one case the directions in which the deposits tend to become finer in grain will converge towards a centre, the middle of the lake ; in the other they will spread out like a fan from a centre, the mouth of the river. A purely fresh- water formation will contain only fresh-water fossils, but there are every now and then exceptions even to this rule. For instance marine crustaceans * have been dredged up from the depths of the large American lakes, and their remains may well get mixed up with those of fresh- water creatures in the deposits now forming beneath those bodies of water. Besides the ordinary types of mechanical deposits the following * These lakes were probably originally bodies of salt water cut off from the ocean by the upheaval of barriers of land, and have since been freshened by the water poured into them by rivers. Some few marine creatures have been able to accommodate themselves to the change and linger on in their deepest parts. Derivative Rocks. 299 kinds of rock are worth notice as often occurring in Lacustrine beds. Chemical precipitates of Carbonate of Lime and Silica may be formed when springs largely charged with these substances burst out on the banks or beneath the waters of the lake, but the amount held in solution must be considerable in order to produce precipitation ; this is often the case in districts where volcanoes either are or have been active. Semi-organic formations also occur, such as the Shell-marls of some small lakes in Scotland which have been filled up by the deposition of sediment ; these beds are described by Lyell as consisting almost entirely of the shells of fresh- water testacea decomposed into a pulverulent marl. Some lakes swarm with Diatoms, the siliceous cases of which accumulate on the bottom and give rise to the deposits of Tripoli or polishing stone : in other cases Diatoms extract iron from the water and cause the formation of Iron Ores. The Lacustrine deposits of Auvergne furnish a good instance of a purely fresh-water deposit ; while the Molasse of Switzerland, of which the reader will find a full description in Lyell's u Elements of Geology," chap, xv., is a fine example of a formation in the main of fresh-water origin, but containing marine intercalations. There are also very exten- sive Lacustrine formations in the western territories of North America.* Salt-water Lacustrine Rocks. The one conspicuous feature which characterizes deposits formed in inland bodies of salt water is the presence of great masses of Rock-salt, Dolomite, and Gypsum. The mechanically formed Shales and Sandstones associated with these are frequently of a deep red colour. Fossils are rare and restricted to certain areas of limited extent. They show little variety, and belong to comparatively few species, and they are often stunted and deformed. The theory of the formation of these rocks has been so fully treated of in the last chapter, that it will be only necessary here to give an example of a group of beds of this class. Example of Chemically formed Deposits. As an instance of deposits probably formed, in part. at least, by precipitation in an inland sea, we may take the Magnesian Limestone and its associated beds of the north-east of England. This formation consists of Limestones more or less magnesian, Red Marls and Sandstones, and Gypsum : parts of it are fossiliferous, but the number of species is small, and the individuals are many of them puny and show strange variations from their normal form. On these and other grounds we are led to look upon the group as an inland-sea deposit, and when we come to examine its members separately, we can form some notion of the succession of events that led to their formation. The group shows the following main subdivisions, begin- ning from the top. 5. Upper Limestone or Brotherton Beds. 4. Red Marls and Sandstones with Gypsum. 3. Small-grained Dolomite. 2. Sandy Magnesian Limestone. 1. Quicksands, and Marls with thin beds of Magnesian Limestone. * See Sun Pictures of the Rocky Mountains, by Prof. F. N. Hayden, chap. vii. , and the Reports of the U. S. Geological Survey of the Territories. 300 . Geology. The lowest division is mainly of mechanical origin, and seems to have been deposited before concentration had gone far enough to produce general precipitation. The quicksands* occur in local patches of small extent and show marked current-bedding ; they are probably portions of deltas spread out wherever a stream entered the lake ; the marls were formed out of muddy sediment further within the area, and the thin bands of impure Magnesian Limestone that are interbedded with them were probably thrown locally in pools, where the solution became concentrated enough to give rise to precipitation. The second division is an extremely sandy Magnesian Limestone ; the sand must have had a mechanical origin, the dolomitic portion was probably a chemical precipitate. We may therefore suppose that during the formation of this portion precipitation and the deposition of sandy sediment went on together. Some of the beds of this subdivision however show numerous traces of animal remains : these may have been organic Limestones formed when the water became for a time clear enough to allow of creatures living in it, and afterwards altered by percolation when the water became saturated with magnesian salts. It is in this division that most of the fossils occur : they have as a rule the character already mentioned, but it is important for our pre- sent purpose to note that one mollusc, Axinus obscurt/s, forms an exception to this rule : it occurs of great size and in considerable numbers, and would seem to have been a hardy creature that could stand almost anything and live almost anywhere. The third division differs from that below it in containing a much smaller admixture of mechanical matter, in parts probably it approaches very nearly to a true Dolomite. During its formation therefore pre- cipitation must have gone on vigorously. Except a few traces in its very lowest beds this Limestone contains no fossils, and the meaning of this evidently is that the increasing concentration of the solution, which was the cause of the greater purity of the rock, was too much even for the hardy creatures that had struggled on during the deposi- tion of No. 2, and that all animal life was either killed or driven away. During the deposition of No. 4 mechanical action predominated, but the beds of Gypsum show that chemical agency was also at work. The Limestones of the topmost division are in many cases scarcely magnesian at all, and they contain a few sadly stunted fossils. During their formation therefore the water must have been so far free from magnesian salts as to be just habitable. It is curious that the few species of shells found in this division occur also in No. 2 ; these we must suppose escaped destruction when the water became unbearable during the formation of No. 3, struggled on in some sheltered nooks or corners, perhaps some way up the rivers, and came back into the lake when matters began to mend a little. As we should expect, among the survivors is the shell which showed so robust a constitution during the formation of No. 2, Axinus obscurus ; but even this tough * The term Quicksand does not here include certain beds that were originally very sandy Limestones, but are now reduced to sand owing to the dissolution of their Carbonate of Lime by percolating water. Derivative Rocks. 301 fellow had evidently had a hard time of it, for he comes back very much dwarfed in size. The formation of this group of rocks then may be supposed to have taken place in an inland body of water fed by streams which brought in both mechanical sediment and matter in solution. As time went on the water would become more and more strongly charged with dissolved matter, and accordingly as we ascend from lower to higher beds we find the rocks growing more and more chemical in their character : at the same time all traces of life disappear because the increasing concentration killed or drove away such animals as had managed for a time to struggle on. When, later on, the proportion of obnoxious salts decreased, a few of the animals which had found some sheltered spot where they could live came back, but they show by their puny size how hard had been the struggle they had gone through in the meanwhile.* D. TERRESTRIAL ROCKS. We do not propose to add anything here to what has been said on the subject of Terrestrial rocks in the last chapter. It is hoped that the descriptions there given will enable the reader to recognise an old land surface, whenever by a happy accident such a relic has been sealed up among rocks and handed clown to the present day. Application to a particular instance. We will conclude this chapter by an example of the way in which the principles laid down in it enable us to map out the distribution of land and sea that existed, and to determine the changes in physical geography that happened, during a period in the earth's lifetime long past by. The group of rocks known to geologists as the Triassic formation will serve our purpose admirably, they have been traced and identified over a very large part of the world ; as we follow them from place to place we find them continually changing in character, and the form they assume at each spot tells us in unmistakable language what were the physical conditions in that quarter during the time of their formation. In England this formation consists exclusively of Eed Sandstones, Shales, and Marls. It contains thick lenticular masses of Rock-salt and Gypsum. I^o marine fossils have been found in it, but it yields remains of plants and terrestrial reptiles, with fishes and minute crus- taceans. Its beds show plentifully ripple-marked and sun-cracked surfaces, with pseudomorphs of salt, and occasionally the footprints of land animals. All these characters lead us to look upon the English Triassic rocks as having been formed in inland seas, and to conclude that the area they occupy was part of a broad continental tract diversi- fied by large closed sheets of salt water. The beds just described pass upwards into a thin band of Shale, Sandstone, and Limestone known as the Penarth beds, which contain some marine fossils. The character of these rocks shows they were formed in shallow water, and their fossils prove that land was not far * For further details respecting these rocks see Quart. Journ. Geol. Soc. of London, xvii. 287 ; Permian Fossils, Palseont. Soc., vol. iii. (1849). 3O2 Geology. off, for besides the marine forms they include the remains of terrestrial mammals. The group puts on more and more pronounced littoral characters as it is traced westwards. The Penarth beds pass up in- sensibly into a more purely marine formation called the Lias. This group of rocks tells us that after a while the continental area of Triassic times was gently submerged, that the sea stole over it from east to west, not reaching beyond the south-west of England during the formation of the Penarth group, but gradually extending its range as Lias times drew on. When we come to the Triassic rocks of Central Europe, we find them more complicated than their English representatives. A large part of them are Red Sandstones and Shales containing chemical deposits of Rock-salt, Gypsum, and Dolomite, which seem to have been formed in inland bodies of salt water ; but interstratified with these are many beds containing marine fossils ; among the most con- spicuous of the marine intercalations is a thick mass mainly composed of Limestone, called the Muschelkalk. The inference to be drawn from these facts is, that during Triassic times the centre of Europe was in the same condition as England, a continental tract with large salt lakes ; but there was this difference between the two cases : in England we have no proof of the presence of the sea, while here we have evidence that the sea was continually making incursions over the land. The longest and most important of these submergences was that during which the Muschelkalk was formed, and by noting the directions in which that group loses its calcareous character and puts on a littoral type, we can determine how far the sea encroached. Above all these rocks comes a band corresponding in every respect to the Penarth beds of our island. Going still towards the east we find the Triassic beds under yet another form. In the Eastern Alps and Lombardy they consist of thick masses of Limestone, swarming with marine fossils, and it is only in their lowest division that they contain any beds likely to be of inland-sea origin. Here then there must have existed, during the greater part of the Triassic epoch, an open ocean, in which great masses of organic Limestone grew up. It is worthy of note that the fossils found in the marine intercalations of the Trias of Central Europe are also met with in these easterly calcareous equivalents. Above these Limestones come the Kb'ssen beds, a group which corre- sponds to the Penarth beds of England. Putting all these facts together we arrive at the conclusion that during Triassic times the physical geography of what is now Europe was as follows. To the east there was a broad open ocean and to the west a continent with large salt lakes, and the Limestones of the eastern area were accumulated in the one at the same time that the inland-sea deposits of the west were being formed over the other. During the whole of the period there were oscillations of level, in consequence of which the sea from time to time advanced over parts of the land surface and then retreated, but none of these incursions reached as far west as England. Whenever the sea spread itself west- wards, it would bring with it out of the eastern ocean those forms of Derivative Rocks. 303 life which were of a migratory turn, and hence the fossils found in the marine intercalations of the centre of Europe are also met with in the Limestone of the Eastern Alps : other forms, not so ready at shift- ing their quarters, are peculiar to the latter. Finally the tendency of the sea to push westwards culminated in a general submergence, which covered the land as far as the south-west of England with a shallow sheet of salt water in which the Penarth beds were deposited ; and a continuation of the depression resulted in producing the still more widely spread Liassic ocean. CHAPTER VI. LITHOLOGICAL DESCRIPTION OF THE CONFUSEDLY- CRYSTALLINE ROCKS. " In the corner of the hall stood a box of stones. Many pretty eye-catching things were among them." WILHELM MASTER'S TEAVELS. WE will now pass on to an examination of the Crystalline rocks, and we shall best arrive at an explanation of the way in which they were formed by considering first of all the subdivision to which the name Confusedly Crystalline was given. Before attempting any inquiry as to the origin of the rocks of this sub-class it will be necessary to give an outline of the lithological composition of its most important members, and to this we propose to devote the present chapter. We will begin by defining the terms which are used to describe the different kinds of Texture or Grain that occur, then give an account of some structures peculiar to glassy or devitrified rocks, and add a notice of other structures of common occurrence in rocks of the class in general. And it will not be desirable that we should rigidly confine ourselves to pure lithological description. Many of the structures and peculiarities of Crystalline rocks are so intimately related to the conditions under which they were formed that we can scarcely give an intelligible account of the former without at the same time pointing out the connection. We shall therefore assume that all rocks of the crystalline class were once in a more or less fused state, the fusion being due to the joint action of heat and water. Later on we will give the grounds for believing this assumption to be true. Texture of Crystalline Rocks. In some of these rocks the crystals or crystalline particles are large enough to be seen by the unaided eye ; such are called Macro-crystalline, or Coarsely Crystalline. Rocks of this character, when single detached crystals are disseminated in more finely grained paste, are said to be Porphyritic. In the case of other rocks closer scrutiny or the aid of a pocket-lens becomes neces- sary to enable us to recognise their crystals, and these are known as the Micro-crystalline, or Finely Crystalline. Lastly, there are the Crypto- or Obscurely Crystalline members, in which crystals can be detected only in highly magnified transparent slices, and by the aid of optical properties, such as polarization and double refraction. Some Crypto-crystalline rocks are the devitrified forms of glassy rocks. Confusedly-crystalline Rocks. 305 Fig. 108. Crystalline rocks, when decomposed, often put on a loose friable form, and are then said to be earthy. Structures in Glassy and Devitrified Rocks. Perlitic Structure. This structure, a magnified example of which is given in fig. 108, is produced by a number of curved cracks which break up the rock into rudely spheroidal portions. Each spheroid is further subdivided into a number of concentric coats by curved cracks roughly parallel to its boundary. As a rule however the cracks are not continuous all round the spheroid, so that the division is incomplete. The spheroids usually lie packed between rectilinear fissures which traverse the rock in all direc- tions, but which seldom cut through a spheroid. The curved surfaces of the cracks have often a pearly or enamel-like lustre, whence the name. The cracks in Perlitic Structure have been caused by the contrac- tion of the rock as it cooled from a fused state.* This conclusion has been confirmed by the artificial production of Perlitic Structure, t Perlitic Structure is so characteristic of Glassy rocks that when it is found in a rock of a stony character it furnishes strong presumptive evidence that the rock was once a glass, and has arrived at its present state by devitrification. J Microliths, Fluidal Structure. In Crypto-crystalline and Glassy rocks, and in the crypto-crystalline or glassy parts of more largely- grained Crystalline rocks, the microscope often shows minute bodies of various forms. Some of these, which probably represent the first stage in the passage from an amorphous to a crystalline state, 1 are small, rounded, singly-refracting concretions called Globulites. Globulites often occur grouped together in various ways, and so give rise to forms which may be looked upon as the embryos of crystals, and which are hence called Crystallites. Microliths are thin needle-shaped bodies, frequently doubly refract- ing : crystals, but crystals very imperfectly developed. Colourless Microliths are sometimes called Belonites : black, opaque, hair-like forms are distinguished as Trichites. Glassy rocks are often crowded with Microliths, which all lie with their longer axes in the same direction. Crystallization began while the body of the rock was still fluid or viscous and in motion, and the * Bonney, Quart. Journ. Geol. Soc. xxxii. (1876) 149; Geol. Mag. [2] iv. (1877). Rutley, Journal of the Royal Microscopical Society, April 1876. t Fouque and Levy, Comptes Rendus, Ixxxvi. (1878) 771 ; G. Cole, Geol. Mag. [2] vii. (1880) 115. J Allport, Quart. Journ. Geol. Soc. xxxiii. (1877) 449. For an account of these bodies see Rutley, The Study of Rocks, p. 161 ; Die Krystalliten, by the late Hermann Vogelsang, Bonn, 1875 ; Zirkel, Mikros. Beschaffenheit, 88 ; and Rosenbusch, Mikros. Physiographic, i. 23. U 306 Geology. minute crystals were dragged along by the moving mass and arranged with their longer axes in the direction of the flow. The lines of Microliths are seen to bend round any larger crystals or grains that lie in their way. This is designated Fluidal or Fluxion Structure. A very similar appearance is found in the glassy portions of rocks. Layers of glass, which can be distinguished by differences in colour, density, or other properties, lie in long parallel strips, one above another, as they have been dragged out by the slow motion of the viscid rock during cooling. This result is sometimes visible only in microscopic slides ; in other cases it has been set up on a large scale, and has given rise to what is called Ribboned or Banded Structure, A structure which cannot be distinguished from true Fluxion Structure may occasionally be observed in rocks which have never been in a state of fusion. In a rock for instance consisting of grains of sand embedded in fine clayey matter, the grains of clay may sometimes be observed to lie in parallel wavy lines which bend round the grains of sand and resemble most closely true Fluxion Structure. The rock has probably been so far softened by the action of heated fluids that its particles become free to move among themselves. It was further subjected to great pressure, and this would tend to arrange the particles in lines perpendicular to the direction of the pressure. Further, wherever a small obstacle like a grain of sand was found, these lines would be bent over and underneath it. This arrangement may be distinguished as Pseudo-fluxion Structure. The student must be very cautious how he bases any inference as to the origin and previous condition of a rock on what looks like Fluxion Structure alone. Before he concludes that the rock was once fused and probably glassy, he must be satisfied that he is dealing with true and not Pseudo-fluxion Structure. Sphcerulites. In glassy and de vitrified rocks we frequently meet with ball-shaped masses clearly separable from the surrounding rock by their different colour and appearance, which go by this name. They are usually made up of fine fibres radiating from a centre, and in this centre a granule or small crystal is sometimes present ; some- times no such nucleus can be detected. Sometimes they- show a series of concentric coats. It is well known that matter, even in the solid state, has a tendency under certain circumstances to rearrange itself in rounded masses consisting of radiating fibres and concentric coats. What the circumstances are which conduce to this arrangement, arid why it takes place, we do not know, and we cloak our ignorance by giving the name of Concretionary Action to the process. It is doubt- less the collection of portions of the rock into balls by Concretionary Action which has given rise to Sphaerulites. Sometimes the segre- gation has taken place after the rock became solid, as in the case of sphaerulitic rocks possessing Fluxion Structure where the lines of Microliths pass through the Sphaerulites. Sometimes the Sphaerulites are arranged in bands and dragged out in the direction of the flow, when they would appear to have been formed while the rock was still soft and in motion. Confusedly-crystalline Rocks. 307 Usually Sphserulites are scattered irregularly through a rock ; sometimes almost the whole mass of the rock is made up of them, and their mutual pressure has changed their curved outlines into polygonal boundaries. Some of the vitreous rocks of Nevada have been found to contain bodies which are very analogous in structure to Sphaerulites but differ from them in shape. They are all more or less elongated ; the outline of some is a very long oval; in other cases they run in winding strings through the rock; and again they are forked, U-shaped, or assume other irregular shapes. They have been distinguished as Axiolites.* Spheroidal Structure. t Some Crystalline rocks are seen, specially after exposure to the weather, to be made up of rudely spherical or spheroidal balls, and each spheroid consists of a number of concentric coats which scale off as the rock disintegrates. This structure is often scarcely recognisable in unweathered rock, but is brought out by atmospheric decomposition. The upper part of a quarry in such a rock is in many cases occupied by a mass of loose crumbly earth full of rounded blocks, which at first sight looks exactly like an accumulation of coarse gravel. This however gradually passes down into solid rock, on the exposed faces of which it may be clearly seen that the action of the weather is beginning to produce a separation into spheroids. An examination of the loose capping shows that the incoherent part consists of the more easily decomposed por- tions of the rock, and that what might be taken for boulders rounded by running water are really the more stubborn spheroids. Instances may be seen in the Basalt of Antrim and in many dykes of similar rock in Cornwall. The production of these spheroids has been attributed to the imper- fectly understood process to which we have given the name of Con- cretionary Action, in fact they haVe been looked upon as Sphserulites on a large scale. Professor Bonney has however given reasons for thinking that the curved surface of the concentric coats are in most cases cracks produced by contraction while the rock cooled from a state of fusion ; that they are in fact Perlitic Structure on a large scale in stony rock. J I have seen a case where masses of Blast-furnace Slag broke up on cooling into spheroids which resembled in every particular those which are so common in some Crystalline rocks, and here I think there could be no question that the separation into spheroids was due to contraction during cooling. That a subdivision into spheroidal masses has been brought about in Crystalline rocks by Concretionary Action in some cases is however pretty certain. Tabular or Laminated Structure. Crystalline rocks show * Geol. Exploration of the 40th Parallel, vol. vi. Plate vi. fig. 2 ; Plate vii. figs. 2 and 4. t Spheroidal, Tabular, and Columnar Structure belong rather to the domain of Petrology than of Lithology. But it is convenient to place them here on account of their kinship with Perlitic Structure. J Bonney, Quart. Journ. Geol. Soc. xxxii. (1876) 149. 308 Geology. occasionally a tendency to break up into tabular or platy slabs. Where the layers are thin the case may be distinguished as one of Fissile Structure ; where their bounding surfaces are curved gently so that the rock separates into lenticular-shaped blocks the name Curvi- tabular has been used. All these varieties have been shown by Professor Bonney to be in certain instances cases of cracking during the process of cooling ; the heat passed off from a large surface, and the consequent contraction took place and strains were produced perpendicular to that surface. Hence the cracks run parallel to the cooling surface, sometimes with considerable regularity, sometimes only approximately. * It is not unlikely however that a somewhat similar structure has been produced in certain cases in a different way. In some lavas a scaly or schistose structure is caused by the presence of minerals, such as Mica, which crystallize in platy shapes. If these minerals come to be arranged in parallel layers with their flat surfaces all in the same direction, the rock tends to split into plates or laminse bounded by two such layers. Mr. Scrope suggested that the arrangement of the platy crystals was brought about in this way. The crystals were present in the mass while it was in an imperfectly fused state and in motion. As the mixture dragged itself slowly along, the platy crystals would be turned round and placed with their broad surfaces parallel to the direction of the flow.f Columnar Structure. Though not confined to Crystalline rocks, it is in them that this structure is commonest and best developed. It is seen to perfection in the Giant's Causeway and at Staffa, and these examples are so universally known to English readers that they will at once understand that the rock is cut up into a number of long prismatic columns, neatly fitted into one another. The prisms are not all of the same shape, three-, four-, five-, and six-sided examples occur, some even with a larger number of faces, but the six-sided forms are by far the most numerous, and their angles have a decided tendency toward those of a regular hexagon. The prisms are divided by transverse cracks, and the surface of these cracks are alternately concave and convex, so that each column is made up of a number of portions fitting into one another by ball-and- socket joints. There can be no doubt that the subdivision into prisms is the result of contraction. The same structure is seen in the highest degree of perfection in starch which has shrunk in drying, it may occasionally be observed in mud dried by the sun, and it is sometimes produced in bricks and other examples of baked clay. Why the prisms should be hexagonal is not altogether an easy question to answer. Mr. Mallet has endeavoured to show that with this hexagonal arrangement the work done during the shrinking will be a minimum, % but his mathe- * Bonney, Quart. Jouru. Geol. Soc. xxxii. (1876) 142 et seq. t Volcanoes, 2nd ed. pp. 140 and 300. See also Judd, Geol. Mag. [2] ii. (1875) 69. Phil. Mag. 4th ed. vol. i. p. 122. If we subdivide a contracting lamina into regular polygonal prisms with their axes perpendicular to its bounding Confusedly-crystalline Rocks. 309 matical demonstration of this theorem can hardly be looked upon as satisfactory. The explanation of the origin of the ball-and-socket joints presents still greater difficulties. Mr. Mallet's views lead to results which are not in complete accordance with observed facts (see his paper quoted below, and Scrope, Geol. Mag. [2] ii. (1875) 412, 566). Another explanation put forward by Professor James Thomson will be found in the Reports of the British Association, 1863, p. 89. Columnar Structure is most conspicuous in Basalt, in which rock the columns are thick and regular with equidistant joints. The prisms of Felstones are slenderer and less regular, and often stretch to a great length without any transverse joints. The same peculiarity has been noticed in Granite, at the Land's End (Transactions Geol. Soc. of Cornwall, iii. 208), in Algeria (Comptes Rendus, xxvi. 76), and in the rock of Ailsa Craig (Macculloch, Western Islands, ii. 493). Felsitic Matter. The paste of many Crystalline rocks is com- posed of a very finely-grained flinty substance to which the name Felsite has been applied. Some Crypto-crystalline rocks are wholly composed of Felsite. The term was introduced before the microscope had been applied to investigate the minute structure of rocks, and the substances which it originally included, though indistinguishable by the unaided eye, are probably by no means all of the same com- position nor all in the same molecular state. In some cases micro- scopical examination with polarized light shows that Felsite consists of a very closely packed collection of minute crystals, and the minerals can be identified. Thus some Felsite is a very intimate mixture of minute crystals of Quartz and Orthoclase. Such Felsitic matter shows under crossed Nicols a variegated mosaic-like pattern. As the crystals become smaller and smaller, their outlines grow less and less distinct, and it becomes more and more difficult to identify individuals. Several crystals too may overlap even in the thickness of a micro- scopic slide, and successive reflection and refraction at their surfaces and interference may modify the character of the transmitted light ; its colour does not represent the result of passing through a single doubly- refracting crystal, but is the product of all these accumulated modifi- cations. As a consequence the dappled mosaic pattern loses its sharp- ness and passes into a vague hazy light. There is again Felsitic matter in which no individual crystals or grains can be detected and which has no action on polarized light, and of Felsitic matter of this class several very distinct modifications have been recognised. surfaces, and assume that during contraction each particle moves along a line perpendicular to the axis of the prism within which it lies, then it may be shown that the work done in contraction is less when the cross section of the prism is a regular hexagon than when it is a square or an equilateral triangle ; but the difference is far smaller than Mr. Mallet states. I have not yet seen a way to get over the difficulties of the problem in its most general form. Those curious on the history of the question may consult Gregory Watt, Phil. Trans. 1804 ; De la Beche, Eesearches in Theoretical Geology, p. 109 ; Jukes, Manual of Geology, 3rd ed. p. 181 ; Scrope, Volcanoes, p. 93 ; Naumann, Lehrbuch der Geognosie, i. 480; T. P. O'Reilly, Trans. Royal Irish Academy, xxvi. (1879) 641. 3io Geology. We are yet very much in the dark as to the true character and origin of the substances which are included under the name of Felsite. It seems certain that some are as truly crystalline as the most coarsely- grained Granite, only the crystals are very minute ; some indeed are nothing but excessively fine-grained Granite. Others of vaguer char- acter may occupy an intermediate position between truly crystallized and normally glassy bodies, and may represent various stages of devitri- fication ; but this is a point on which we cannot as yet speak with any certainty.* The old-fashioned names Petrosilex and Hornstone and the French term Eurite are equivalent or nearly equivalent to Felsite. Fluid-, Glass-, Stone-, and Gas-Cavities. Crystals often enclose hollow spaces, of all sizes from 1-10,000 of an inch in diameter up occasionally to some large enough to be seen with the naked eye. These cavities contain in some cases liquid, and they are then known as Fluid-Cavities ; sometimes the cavities are entirely filled, sometimes there is a bubble in them which either moves about of its own accord or may be set in motion by the application of gentle heat. The liquid is often water or some aqueous saline solution, but in some cases it has been proved to be liquid Carbon Dioxide and in others a liquid Hydrocarbon. Small cubical crystals of Common Salt and other soluble salts have been observed in some Fluid-cavities. It is known that when substances crystallize out from aqueous solu- tions, small portions of the liquor are caught up and enclosed within the crystal ; if the crystals have been formed at a high temperature, the imprisoned liquid will contract on cooling and the cavity will con- tain a bubble. Water-cavities then containing bubbles show that the crystals in which they occur were formed at a high temperature and in the presence of water. Cavities containing glassy or stony matter, with or without bubbles, also occur in the crystals of some rocks. In their case however the bubbles are immovable. Similar cavities are met with in the crystals of furnace slags, which have solidified from a state of fusion but in which water was not present in the fused mass. Glass- or Stone-cavities indicate that the crystals in which they occur were formed in the middle of a mass rendered liquid by heat, portions of which were caught up and enclosed by the crystal during its growth. Rapid cooling would consolidate the enclosure into a glass, and a Glass-cavity would result. This glass if devitrified would give rise to a Stone-cavity; or a Stone-cavity might arise from cooling sufficiently slow to allow of the enclosure assuming a crystalline state. Other cavities contain vapour or gas, bubbles in fact similar to those which produce the effervescence of soda-water, only enclosed in a solid matrix, f * There is an exhaustive summary of the various opinions that have been held about Felsitic matter in Rosenbusch's Mikros. Physiographic, i. 60 ct seq., and a good abstract of Rosenbusch's views in Rutley's Study of Rocks, p. 167. See also Allport, Quart. Journ. Geol. Soc. xxx. (1874) 544 ; Geol. Mag. (i.) ix. 537 ; Zirkel, Geol. Exploration of the 40th Parallel, vi. 1, 71, 72 ; Mikros. Beschaffenheit, p. 276. t Sorby, Quart. Journ. Geol. Soc. xiv. 453 ; Reports of British Association, 1856, p. 78 ; 1857, p. 92 ; J. A. Phillips, Quart. Journ. Geol. Soc. xxxi. 332 ; Allport, Quart. Journ. Geol. Soc. xxxii. (1876) 413 ; W. N. Hartley, Trans. Royal Mic. Soc. xv. (1876) 173; Rosenbusch, Mikros. Physiographic, i. 27; Zirkel, Mikros. Beschaffenheit, 39. Confusedly-crystalline Rocks. 3 1 1 Vesicular and Amygdaloidal Structure. Some parts of certain Crystalline rocks are full of bubble- shaped holes or cavities, and are then said to be Vesicular. A similar structure is seen in furnace slag which has flowed out and cooled in the open air. It is caused by the boiling up and escape of elastic gases, which as long as the slag was under pressure in the furnace were in a highly-condensed state, but which expanded and blew out the slag into a spongy mass as soon as the pressure was removed. Vesicular structure in rocks has been produced in the same way. Not unfrequently water containing mineral matter in solution has percolated through the rock, and depositing its contents on the walls of the cavities has partly or entirely filled them up. The rock is then called an Amygdaloid, because the kernels bear some resemblance to almonds (Greek, amygdala}. The researches of Professor Daubree have thrown great light on the way in which Vesicular rocks are converted into Amygdaloids. His first observations were made at Plombieres in the Vosges, where warm springs holding Alkaline Silicates in solution rise out of Granite. These springs have been frequented and utilized for bathing at least as far back as the Roman occupation of the country, and the conduits constructed at that date for conveying the water to the baths are still in existence. These conduits are of concrete formed out of broken bricks and stones bedded in lime-mortar. The material is porous and has for centuries been saturated and bathed by a gentle current of the water of the springs. The result has been that the cavities in the bricks have been filled or lined with many of the minerals that are common in natural Amygdaloids, sundry Zeolites, Calcite, Aragonite, Opal, Chalcedony, probably Tridymite, and others. The deposition has extended itself even to cavities and fissures so small as to be visible only under the microscope. The process has gone on at atmospheric pressure and at a temperature not exceeding 70 C. Similar facts have since been noticed by the same observer in other localities.* Subdivisions of the Crystalline Rocks. The classification of the Crystalline rocks is a matter of the greatest difficulty. The variations in their composition are all but endless, and present so many intermediate steps from one form to another that it is scarcely possible to establish subdivisions between which connecting-links may not be found. The anxiety of some observers to elevate every variety that may have come under their notice to the rank of a distinct species, has led to an unnecessary multiplication of names ; and more confusion is introduced by different writers using the same name for rocks of different mineral composition, t Still, if we shake ourselves clear of minute details and take a broad view of the composition of this class of rocks, it seems possible to parcel them out into two main subdivisions, * Quart. Journ. Geol. Soc. xxxiv. (1878) 73; Geol. Experimental, i. 208. t Fortunately these matters are not as important as might at first sight appear. What the geologist wants are not the minute differences insisted on by mineralogists and petrographers, but some broad leading groups in which to arrange the rocks he meets with in the field ; and above all a careful account of those structures which enable him to reason about their origin. 3 1 2 Geology. sufficiently marked in their mineral composition to be clearly dis- tinguishable from each other, and a third class partaking in some measure of the distinguishing characteristics of the first two. The existence of this third class of course makes it impossible to draw any hard lines between the three classes, and in some cases leaves it doubt- ful to which of two subdivisions a particular rock ought to be referred ; but if we neglect for an instant these connecting forms, and fix our attention on typical instances of the two first-named subdivisions, we shall find these so distinct from each other, and find also among the different varieties of rocks so many that conform more or less closely in mineral and chemical composition to one or other of them, that we may usefully group together the rocks that resemble one type in one class, and those that resemble the other type in a second class, even though we know that between these two classes there lies a debatable ground, into which each of them merges by almost insensible grada- tions. ' f Nearly all the Crystalline rocks have a Felspar for one of their principal ingredients, and as the Felspars are divided into the Highly Silicated or Acid and the Poorly Silicated or Basic, so the Crystalline rocks can be divided into two great subdivisions, according as their prevailing Felspar belongs to the first or second of the Felspar families. The third subdivision mentioned partakes in some degree of the char- acters of both of the two first. Acid Rocks. The first of these great subdivisions is known as the Acid, Highly-silicated, Felspathic, or Trachytic class. The Felspar is one of the highly-silicated species, generally Orthoclase, though one of the basic Felspars is frequently present as well ; there is generally also some free or uncombined Silica present in the shape of Quartz or Tridymite. Other minerals may enter into the composi- tion of rocks of this class ; but its two distinguishing characteristics are those just mentioned, the highly-silicated character of its Felspar and the presence of free Quartz. The Acid rocks are poor in Lime, Magnesia, and Iron, and the absence of these substances, which act as fluxes, and their richness in Silica, makes them difficult of fusion. Their specific gravity ranges from 2-3 to 2 -7. Basic Rocks. The second great subdivision is known as the Basic, Poorly-silicated, Magnesian, Hornblendic, Pyroxenic, Basaltic, or Dioritic class. The Felspar may be Oligoclase, but is more frequently Labradorite, Anorthite, or some basic form, and Hornblende or Augite is very generally an important ingredient. No free Silica is present as a con- stituent mineral, but Quartz may occur as an accessory. Compared with the former class these rocks are poor in Silica, and rich in Lime, Magnesia, and Iron. Hence they are the more readily fusible of the two classes. Their specific gravity ranges from 2 '7 to 3*1, so that they are also the heavier of the two. * For a very striking instance of a gradual passage from the extreme type of one of these classes to the extreme type of the other see Leonhard's Jahrbuch (1873), p. 225 ; and the Monzonite of Monzoni, Judd, Geol. Mag. [2] iii. 206. Confusedly-crystalline Rocks. 313 The extreme and typical rocks of the Acid and Basic classes are widely removed and clearly distinguishable from each other ; as has been mentioned however there are many rocks of an intermediate character which form connecting-links between the two, and for the reception of some of these an intermediate class may be established, though it is altogether impossible to say exactly where its boundaries on either side are to be drawn. Perhaps the general composition of its rocks may be thus stated. Highly - silicated Felspar without Quartz, or poorly-silicated Felspar and Quartz. A very useful rough-and-ready test for determining to which class a Crystalline rock is to be referred is furnished by the crust formed on the outside by the action of the weather. The weathered surface of an Acid rock is very usually white owing to the decomposition of its Orthoclase ; the large proportion of Iron in the Basic rocks by its oxidation generally stains their weathered crust, often to a considerable depth, brown or red. They also frequently effervesce with acids in the cracks and crevices of the surface, owing to the formation of Carbonate of Lime out of the constituents of their Lime Felspars. These tests are not infallible, but in a majority of cases they may be relied upon. The table on the following page shows the main distinctive characters of the three classes. It is easy to see to a certain extent how these subdivisions arose, and why the component minerals of the Crystalline rocks do not occur indiscriminately, but are associated together according to broad general laws. In some cases there has been Silica enough to form the most highly- silicated compounds possible, and some to spare besides ; in these accordingly the highly-silicated Felspars prevail, and the superfluous Silica appears as Quartz ; in other cases the Silica was not so plentiful, there was only enough to form poorly-silicated compounds, and all there was, was used up in doing this ; here therefore we find basic Felspars and no free Quartz. Why magnesian Silicates should be so much more largely associated with basic than with acid Felspars is not so easy to explain : the student may consult Durocher's speculations on this point, which he will find very carefully and lucidly explained in Professor Haughton's " Manual of Geology," chap. i. Appendix A. We must warn him however that some of the brilliant Frenchman's facts are, to say the least, doubtful, and others are capable of a very different explanation from that he puts on them. The scheme of classification just described depends mainly on the proportion of Silica in each variety of rock. Other authors have subdivided the Crystalline rocks according to the Felspar which pre- dominates in them. These, and other like systems, rest on a purely mineralogical basis, and it will be found that they all alike lead, when we come to details, to more or less vagueness of definition and con- fusion of nomenclature. To a certain degree possibly this must always be the case. The composition of a large rock mass varies in many cases so much from point to point, that if we trust merely to mineral composition, it is impossible to fix on a name that will be applicable to all parts of it, 3 [ 4 Geology. d CO 00 i "* I 1 1 10 10 c d r^ 1 o I 1 CO r-l S 1 o g 1 x - 1C 00 oo O5 OS 3 O) < i -2 1 i-H 1-H S PH B CO 1 S O SQ 1? p | < '05 o ^ E S C3 3 cc w eg to P '53 1 V ci A 1 02 1 P 00 bO OJ CB 1: c5 JH 'i ce o DQ Alumina CS 1 oT o> Oj 'hJ p od | ej cS Oxides of Ii 'i "^ H 03 OT of fp 1^ K. f 5| |i O 6 i w 'o 1 od S Weathei OQ Confusedly-crystalline Rocks. 3 1 5 and yet there may be satisfactory geological evidence that all the varieties were produced at the same time and fundamentally by the same operation, and that their differences must therefore be from a geological standpoint accidental. Hence arises a constant clashing between mineralogical and geological classification ; as yet we have only the former, but the time may come when an arrangement of the rocks now under consideration on a true geological basis will be possible, and then we may hope that many of the present seeming contradictions will vanish. The right thing seems to be to look upon all the present schemes of classification of the Crystalline rocks as probably artificial, something like the Linmean system in Botany, and to wait patiently till a more extended knowledge enables some one, who shall be at once a great petrographer and a great geologist, to establish a natural system, which shall pay regard first and foremost to the method of formation of the rocks, and look upon their mineral composition as merely subsidiary. One instance will perhaps make these remarks more intelligible. Of the large class of Acid rocks grouped together as Felstones, some have been lava streams poured out in the open air, some have consolidated from a fused state at great depths below the surface, and some are- rocks originally non-crystalline that have been rendered crystalline by heat ; of the differences between these three kinds of Felstone a mineralogical classification takes no note so long as they agree, which they often do, in mineral composition ; whereas a natural system would at once place them in distinct and widely-separated classes. We will now pass to some of the more prominent examples of the three classes of Crystalline rocks. A. ACID ROCKS. The Acid rocks may be grouped under the following heads. A a. The Granite Group. A b. The Quartz-felsite Group. A c. The Felstone Group. A d. The Quartz-trachyte and Rhyolite Group. A e. Glassy Rocks of Acid composition. A a. The Granite Group. Granite. The essential mineral constituents of Granite are Quartz, Felspars, and Micas. The Quartz is usually not in crystals. It either fills up the spaces between the other minerals, or occurs in rounded lumps with irregular outlines. Very frequently the Quartz can be seen to have been moulded on previously-formed Felspar crystals. It then exhibits smooth polished surfaces, which may be mistaken for the faces of crystals, and it may at first sight be supposed to be itself crystallized. A little attention will prevent the student from falling into this mistake. Cavities containing either water or watery saline solutions or liquid Carbon Dioxide are almost invariably present in the Quartz, often in great number. Glass-cavities apparently do not occur. 3 1 6 Geology. Among the Felspars Orthoclase is always present, but there is also in every case some Plagioclase ; the proportions between the two vary very much in different varieties. The Plagioclase has been ascertained to be sometimes Albite,* sometimes Oligoclase, and sometimes Labra- dorite ; Microline occurs in many Granites ; Anorthite has not been observed. The Micas are Biotite and Muscovite, in some cases only the one in others the other alone is present, but the two do occur together. One very distinctive point about Granite is its thoroughly crystalline character. The crystalline grains or crystals are not embedded in a crypto-crystalline or non-crystalline paste, but are in direct contact with one another. The whole mass of the rock is made up of them. Granite then is a rock composed of Quartz, Orthoclase, Plagioclase, and Micas, without any paste or matrix. The Quartz is rarely in crystals and almost always contains water-cavities, but probably never glass-cavities. Besides these essential constituents many accessory minerals are of common occurrence. For an account of them reference must be made to works on Petrology. The very general presence of Apatite may however be noticed. Granite shows all degrees of grain, and not unfrequently the presence of large crystals of Orthoclase renders it porphyritic. In one variety known as Giant Granite^ the crystals of Orthoclase and the plates of Mica reach an unusually large size. Hornblendic Granite.^. This rock differs from Granite proper in containing Hornblende in addition to Mica, or in the place of it. It is usually rich in Plagioclase. When Mica and Hornblende are both present, the Mica is a Magnesia Mica. Magnetite and Spheric are common accessories in Hornblendic Granite. A similar rock, but containing Augite in the place of Hornblende, occurs in the Vosges. Aplite, Haplite, Halb-granit, Granitell, differs from Granite in con- taining no Mica. In one variety of this. rock, called Pegmatite by Haiiy, Quartz and Orthoclase are arranged in somewhat zigzig-shaped alternating plates, and in a section perpendicular to the laminae the Quartz forms figures which bear a fancied resemblance to Hebrew letters, whence the rock has been called Graphic Granite. Greisen, Zwitter, Stockioerks-porphyr \\ is a rock which maybe placed here. It is composed of Quartz and Mica, and is often rich in Tour- maline or Schorl. It is of interest as frequently carrying large * Professor Haughton has given an interesting account of the occurrence of Albite in the Granite of the Mourne Mountains ; see Quart. Journ. Geol. Soc. xii. (1856) 190 ; Geol. Mag. (1) vi. 561. + This has been called Pegmatite by Delesse and Naumann. Though a very striking variety, it is hardly entitled to a distinctive name. J Called Syenite by some English authors. The term Syenite is now however generally restricted to the rock described on p. 324. Rosenbusch, Mineral Physiog. ii. 21. || Greisen has been denned to be a coarsely-grained aggregate of Quartz and Mica ; Zwitter, a finely-grained aggregate of Quartz, Chlorite, Specular Iron, and Cassiterite ; Schorl-rock, Quartz and Schorl. But the three are so intimately associated that they must be looked upon as varieties of the same rock. Confusedly-crystalline Rocks. 317 quantities of Tin Ore (Cassiterite). The Quartz is the more important constituent. The Mica is usually wholly or in part Lepidolite. Greisen frequently contains pseudomorphs of Tourmaline and Quartz after Orthoclase, and this furnishes good ground for believing that it has been produced by the alteration of Granite. The Orthoclase of the original rock has been in part replaced by Schorl ; in part it has been decomposed, and while its other constituents have been carried away, its Silica has remained to increase the percentage of Quartz. Daubree has suggested that the change has been wrought by the action of Hydrofluoric Acid.* Luxullianite. A ground mass of velvet-black Tourmaline in which are embedded grains of whitish Quartz, large crystals of Orthoclase, and small rather irregular crystals of a Felspar. The ground mass on microscopical examination is found to consist of colourless Quartz, thickly penetrated by minute needle-shaped crystals of Schorl ; it contains besides irregular grains of brownish Tourmaline and small crystals of decomposed Orthoclase. The larger Orthoclase crystals are much decomposed, and penetrated by minute needles of Schorl which spring from a grain of clear Quartz. There can be little doubt that this is an altered Granite, and that the Schorl has been produced by the alteration of Orthoclase : possibly the Silica set free during the process may have been the origin of the Quartz base from which the crystals of Schorl spring. Professor Bonney thinks that the brown Tourmaline is altered Mica.t China-clay Rock, Carclazyte. This is a rock which has been pro- duced by the alteration of the Orthoclase of Granite into Kaolin. This change constitutes the main difference between it and Granite, but it also usually contains specks or crystals of Schorl, Lepidolite, Gilbertite (a mineral produced by the hydration of Muscovite), sometimes granules and crystals of Cassiterite, and occasionally Fluor-spar. The frequent occurrence of minerals like Schorl and Lepidolite which contain Fluo- rine in this rock has led to the conjecture that the decomposition of the Orthoclase was effected or helped on by Hydrofluoric Acid.J Granit-porphyr. The rocks to which this name is applied by Zirkel are porphyritic to a marked degree. They agree in their general com- position with Granite ; but the matrix in which the crystals are embedded is of too fine and compact a nature and too decidedly a paste to allow of their being called Granites. On the other hand it is not flinty enough to justify us in calling it Felsitic, and therefore the rocks cannot be placed in the Quartz-felsite group. Such rocks form a connecting-link between Granite and Quartz-felsite. A b. The Quartz-felsite or Elvanite Group. The essential * C. Le Neve Foster, Quart. Journ. Geol. Soc. xxxiv. (1878) 640 ; Trans. Royal Geol. Soc. of Cornwall, ix. pt. 3 (1877) ; Daubree, Etudes de Geol. Experi- mentale, i. chap. i. ; E. Reyer, Jahrbuch der k. k. Geol. Helens, xxix. (1879) 1. t Min. Mag. i. (1877) 215. J The Hensbarrow Granite District, J. H. Collins, Truro, 1878. See also the papers of Foster and Daubree quoted in the last note but one. Cornish miners apply the word Elvan to rocks all of which occur in dykes (see p. 372), but which are not all of the same composition and physical struc- ture. Many Elvans however are true Quartz-felsites, and hence the late Professor Jukes suggested the term Elvanite for the rocks of this group. 3 1 8 Geology. constituents of the rocks of this group are Quartz, Orthoclase, Plagio- clase, and Felsitic matter. The Quartz is not unfrequently in crystals, hexagonal prisms ter- minated by hexagonal pyramids or double hexagonal pyramids. The faces of the crystals are sometimes very unequally developed, and cases may be noticed where two of them are so reduced in breadth that a cross section of the prism seems to a casual observer to be four-sided. In many of the crystals the angles and edges are rounded ; broken crystals are also frequently met with. Quartz occurs besides in lumps without any definite shape. As a rule cavities are not so abundant as in the Quartz of Granite, but they are generally present ; water-, glass-, and stone-cavities all occur, but none have been observed containing liquid Carbon Dioxide. Gas-cavities are also common. When thin slices are viewed under the microscope the Quartz appears constantly to enclose portions of the matrix : it is very difficult to decide whether these are true enclosures surrounded on all sides by Quartz, or whether they are canals penetrating the crystal or lenticular depressions on its faces which have been cut across by the surface of the slice.* In some instances the enclosures have the outline of a double hexagonal prism, and these may be looked upon as true enclosures. The Orthoclase sometimes occurs in large crystals, when the rock is markedly porphyritic : not unfrequently however no Orthoclase can be detected by the naked eye. The Orthoclase crystals are often rendered more or less opaque by decomposition, or the presence of numerous minute foreign bodies scattered through them. Plagioclase is not always present ; and when it is, it cannot in many cases be recognised by the eye alone. The larger recognisable crystals are embedded in a paste, which in typical rocks of this class is of a finely-grained felsitic character ; occa- sionally it is of coarser grain, when the rock approximates to Granit- porphyr. This paste is in some cases known to be a very intimate mixture of Quartz and Orthoclase ; in others its exact composition and molecular state have not been ascertained, and we can merely apply to it the term Felsitic matter in the vague sense in which we have already explained that expression must often be used. Mica is not unfrequently present in Quartz-felsite, but scarcely often enough to allow of its being looked upon as an essential constituent. Hornblende and other accessories also occur. The essential characters then of a Quartz-felsite are the presence of a Felsitic paste, the frequent though not universal occurrence of Quartz in crystals, and the general absence of Mica. The rocks of this group have been called by many different names, as for instance Quart z-porphyr, Quartz-porphyry, Felsit-porphyr, Eurit- porphyr, Eurite-porphyroide, Hornstein-porphyr. The noun Porphyry is now seldom used in Petrological nomenclature in this country, and, wherever it is possible, a rock in which large crystals are embedded in a more finely-grained paste is said to be Porphyritic. Thus we speak not of a Felstone-porphyry but of a Porphyritic Felstone, because * Rutley, The Eruptive Rocks of Brent Tor (Mems. Geol. Survey of England and Wales), pp. 7, 8. Confusedly-crystalline Rocks. 319 the essential point about such a rock is that it is a Felstone, and the porphyritic structure, being a mere subsidiary accident, is most suitably denoted by an adjective. Wherever we can by a slight change of this character make the old nomenclature harmonize with that more gener- ally adopted at the present day, it is desirable to do it. But where this cannot be done, and where we can avoid the use of the word Porphyry only by endeavouring to supersede a well-established name by one newly coined, it is better to retain the noun. Granit-porphyr is an instance which is something different from Porphyritic Granite. A c. The Felstone Group. A typical Felstone is essentially composed of Orthoclase and Quartz, but the crystals or crystalline grains of the two minerals are so small and so intimately mixed up with one another that they cannot be recognised and distinguished by the unaided eye. Many rocks are provisionally placed in this class, which bear a general resemblance to true Felstones, but whose exact composition has not yet been ascertained. Not unfrequently the microscope shows, where polarized light is used, that the rock is a matted mass of minute crystals, and in some cases these are large and well defined enough to enable us to say that they are crystals of Quartz and Orthoclase. Chemical analysis con- firms this view, and in certain cases, where the individual minerals cannot be recognised by their optical and crystallographic properties, leads to the same conclusion as to the mineral composition of the rock. Professor Haughton arrived at this result from the analyses of a number of Felstones, a summary of which is given in the table below.* Max. Min. Mean. Silica . 81 -36 74-88 76-67 Alumina 12-24 6-54 9-99 Ferric Oxide 5-82 0-92 3-37 Potash 5-65 3-09 4-40 Soda . 3-36 2-49 2-92 Lime . 1-81 0-34 0-88 Magnesia Ferrous Oxide 1-28 0-39 0-64 0-04 Loss . 1-20 0-56 0-81 To these analyses the following mineral composition corresponds. Max. Min. Mean. Quartz 45-54 20-51 34-09 Orthoclase . 76-65 54-16 64-44 It by no means follows however that all the rocks whose external appearance would lead us to class them as Felstones have a similar * Trans. Royal Irish Academy, xxiii. 32O Geology. composition. The same uncertainty attaches to many of. them, specially the very compact varieties, as hangs over the question of the composition and molecular constitution of Felsitic matter. When the mass of the rock is not specially finely grained, so that it has a stony look, the rock is called a Felstone ; if the rock be mainly made up of very finely-grained splintery Felsitic matter, it is called a Felsite. Both Felstones and Felsites become occasionally porphyritic by the development of large crystals of various minerals, most commonly of Orthoclase, in them. A porphyritic Felstone differs from Granit- porphyr only in the general absence of Mica ; a Pophyritic Felsite may be separated from Quartz-felsite on the ground that no crystals or lumps of Quartz of recognisable size occur in it. There are how- ever varieties which form connecting-links between all these classes of rock. Mica and Plagioclase are by no means always wanting in Felstones, but the former and perhaps the latter can hardly be reckoned essential constituents. There are cases where it is pretty certain that a Felsite is the devitrified form of a glassy rock : * but we are not in a position to say that this is always the case. A d. The Quartz-trachyte and Rhyoiite Group. Quartz-trachyte. The rocks which bear this name consist of a rough and porous matrix, in which are embedded crystals and crystalline grains of Quartz and Sanidine. The matrix appears homogeneous to the naked eye, and very frequently scarcely any distinct crystals can be recognised in it under the microscope ; less often the microscope shows it to be composed of a mass of interlacing small crystals. It presents many points of resem- blance to Felsitic matter, but differs in being more open and porous. The Quartz show the same apparent enclosures of the matrix as occur in the Quartz of Quartz-felsites. Water-cavities have been observed in it in only a very few instances, and in them they are not numerous. Glass- and stone-cavities occur, but they are not abundant. Glass- cavities have also been noticed in the Sanidine, but they are not plentiful. The following minerals also occur very generally. Plagioclase is usually sparingly present, Albite and Oligoclase seem to be the com- monest species. Biotite and Hornblende are generally present in small quantity, and sometimes Augite in microscopic grains. Magnetite and Apatite are of common occurrence, but not to any large amount. Tridymite has been met with in some Quartz-trachytes, either lining cavities or forming part of the matrix. The more coarsely-grained varieties of Quartz-trachyte in which the crystals predominate over the matrix, have been distinguished as Nevadite. The more Elvanite-like modifications in which there is a marked matrix are by some authors designated Liparite. The term Liparite is however applied by other writers to the whole group. It will be noticed how very closely Quartz-trachyte resembles Quartz- * Rutley, Quart. Journ. Geol. Soc. xxxv. (1879) 508. Confusedly-crystalline Rocks. 321 felsite in all essential particulars. They have been separated mainly on the strength of the fact that the Quartz-felsites are older rocks than the Quartz-trachytes. The inferences to be drawn from this fact cannot be discussed at present, but we shall see after a while that the ' alterations which time necessarily brings about in a rock are just those which would tend to convert the younger rock into the older, and that there is nothing to forbid the notion that the rocks we call Quartz- felsites, or at least many of them, were once Quartz-trachytes. Rhyolite. A Rhyolite differs from a Quartz-trachyte solely in con- taining more or less glassy matter in the matrix. Sometimes the matrix is wholly glass, in other cases glassy and crypto-crystalline matter are intermingled, and there are cases where rounded or angular masses of glass are distributed among the crystals. It is perhaps as well to keep up some distinction between rocks in which no glassy matter can now be detected and those retaining more or less of unmistakable glass. But it will be seen that such distinctions are very artificial ; devitrification would convert a Rhyolite into a Quartz-trachyte, and partial devitrification produces so many inter- mediate forms that no hard line can be drawn between the two species. Sanidine-trachyte. The matrix of this rock is shown under the micro- scope to be composed of small Sanidine crystals with some Plagioclase, and glassy or Felsitic matter : Quartz cannot be distinguished. In this matrix there are embedded larger crystals of Sanidine, Plagioclase, Biotite, and other minerals. Sanidine-trachyte may be looked upon as a variety of Quartz-trachyte in which the Quartz cannot be recognised : that there is Quartz in it is probable, for it contains more than 72 per cent, of Silica. A e. Glassy Rocks of Acid composition. It has been already pointed out that when a substance solidifies from a fused state the cooled mass will be crystalline or glassy according to the conditions under which solidification takes place ; and that among the circumstances which tend to produce a glassy product rapid cooling is one of the most efficient. We can very well understand then that the same fused mass may assume, as it cools, glassy forms at some spots and crystalline states at others. The two products, if looked at apart, are so widely different that at first sight, may be, no suspicion would arise that they were in any way related to one another; but if on further examination we find that a glassy and a crystalline rock agree in chemical composition, and if the constituent minerals of the crystal- line rock occur also in the glassy mass, and if further the two are associated together geologically, better still if we can trace a passage from one form into the other, then it dawns upon us, that, diverse as the two are in molecular condition, they are really only the same thing in two different states, and we speak of the one as the glassy form of the other. The glassy forms of the Acid rocks may be grouped under three heads, Pitchstone, Obsidian, and Pumice. Pitchstone has a greasy, half glassy, or pitch-like lustre, and a fracture more or less conchoidal and sometimes rather splintery. It is generally dark greenish or brown in colour, sometimes red or dark yellow. It always contains water, the amount varying up to 9 per x 322 Geology. cent. The base of the rock is a true glass, but crystals are always disseminated through it : if these are large and numerous, the Pitch- stone becomes porphyritic ; frequently they are so small as to be visible only by the aid of the microscope. Pitchstones frequently show good fluxion structure and are often largely devitrified. The following subdivisions may be recognised. Porphyritic Pitchstone, Vitro-porphyr, Pechstein-porphyr. Differs from Quartz-felsite only in having the Felsitic matrix replaced more or less completely by glassy matter. Devitrification takes place usually in this rock by the production of Felsitic matter. This leads to the sus- picion that many Quartz-felsites are devitrified Porphyritic Pitchstones. Felsite-pitckstone. Differs from Porphyritic Pitchstone in the com- parative scarcity of embedded crystals, but the two pass into one another. It may be looked upon as the glassy form of Felstone or Felsite. Trachyte-pitclistone is related in many cases very closely in chemical composition and the character of the embedded minerals to Quartz- trachyte, of which it must then be regarded as the glassy equivalent, specially in those instances where the two rocks are geologically associated. Rhyolite forms an intermediate step, and there are many connecting-links between it and Trachyte-pitchstone. Devitrification in these rocks is generally caused by the development of Microliths. Fluxion-, spheroidal, and perlitic structures are common both in Felsite and in Trachyte-pitchstone. Some forms in which the perlitic structure is very largely developed have been distinguished as Perlite. They have frequently a shimmering, pearly, enamel-like lustre. There are besides Pitchstones which are probably the glassy forms of other crystalline rocks, Trachyte, Dacite, Andesite, etc., and yet others whose crystalline equivalent cannot be decided with certainty. Where doubt on this point exists, the rock can be called simply Pitchstone. Obsidian or Volcanic Glass is even more decidedly glassy than Pitch- stone, it contains less water, and embedded crystals are not so numerous as in that rock. It is splintery, so that sharp edged fragments can easily be detached from it, and the broken surfaces show very perfect conchoidal fracture. It is mostly black or dark brown. Usually the embedded crystals are microscopic, but occasionally they are large enough to be seen with the naked eye ; they belong to the same minerals as enter into the composition of Quartz-trachyte, with the exception of Quartz. Perlitic structure occurs only rarely in it, the traces of devitrification are slight, and Sphaerulites are not so common as in Pitchstone. In short Obsidian is a more thorough glass than Pitchstone, in some cases perhaps because it has solidified under conditions specially favourable to the production of the glassy state, in many probably because it has been devitrified to a less extent. Its smaller percentage of water is perhaps due to the fact that the con- ditions under which it cooled favoured the escape of the water which was originally present. Pumice, Bimstein, is frothy, porous, stringy Obsidian. It is com- posed of glassy matter, which in some parts is full of bubble-shaped holes and winding canals, and in others is drawn out into long slender rods which unite, interlace, and twist round one another. HE 323 Confusedly-crystalline Rocks. , ^j Pumice has arisen from the solidification of a mass of acid glass permeated by elastic gas while it was in a state of imperfect fusion and in motion. The expansion of the gas as it escaped blew the stuff up into a froth, and the motion dragged the pasty substance out into threads and twisted them round one another. Relation of different forms of Crystalline Acid Rock to one another. The preceding short sketch of the principal varieties of the Acid Crystalline rocks will give the student some notion how numerous they are, and if he refer to any detailed treatise on Petrography, he will find that the different forms under which they present themselves are far more in number than space will allow us to describe here. But a little reflection will show him that a strong family likeness pervades the whole class, which even this manifold diversity cannot entirely hide. The distinctions between the various forms depend mainly on differences in texture or grain ; but when we turn to mineral and chemical composition, we find that these are from a broad point of view remarkably constant throughout the whole group. Now differences in texture depend upon differences in the conditions under which the rock consolidated ; we have already had one instance of this when it was shown that the same fused mass would, if it cooled rapidly, harden into a glass, and, if it cooled slowly enough, form a crystalline aggregate, and we shall learn by-and-by that this is only a particular case of a general truth. Uniformity in chemical composition on the other hand points to a common source from which all the rocks that exhibit it sprang. If we bear this in mind we shall come to the conclusion that it is almost allowable to say that all these kinds of rock, diverse as they are in many points, are no more than modifications of one fundamental type. And the notion gains strength when we find that passages from one form into others have been actually observed. Thus M. Ebray (Bull. Soc. Geol. de France [3] iii. 290) describes rocks which are at one spot Granites, at another Elvanites, and at a third Felsites. A similar case is thus described by Mr. Gilbert : " From a Trachyte composed of a rough lithoid paste with embedded crystals of Orthoclase, Oligoclase, Horn- blende, Mica, and Quartz there is a perfect gradation to varieties with a compact and even vitreous paste, and thence to Obsidians, and vari-coloured glasses, and to Pumice " (Geol. of Western Territories, p. 133). Again it may occur to the student that even in the limited list we have given there is an unnecessary multiplication of names, that the differences between Quartz-felsite and Quartz-trachyte for instance are very small indeed. This is true ; we may arrange the varieties to which we have given distinct names in two parallel columns in the way shown in the following table, so that for most of the varieties in the one column there is a variety closely resembling it in the other. Macro-crystalline Micro-crystalline Glassy j Granite | Quartz-felsite Felstone Felsite-pitchstone Nevadite. Liparite. Trachy te-pitchston e. Obsidian. ., 324 Geology. When this is done we shall find that the rocks on the left-hand column are as a rule comparatively the older and these in the right- hand column of comparatively more recent date. The distinction is a valid one, but too much importance must not be attached to it. We must not infer because differences in age go along with certain differences in character that the agencies by which crystalline rocks were produced during the earlier part of the earth's lifetime differed from those which operated during later times. We must take into account that time works changes in rocks as in everything else, and when we come to know more of the nature of these changes we shall see that they are fully competent to account for nearly all, if not all, the differences which distinguish the older from the newer crystalline rocks. And we can on similar grounds explain the absence of a counter- part to Granite among the newer and to Obsidian among the older rocks. It is not because no Granite was formed during the epochs nearer to our own day, but because we have not had an opportunity of getting a sight of it. Why we shall know by-and-by. And pro- bably the reason why no such distinctly glassy rock . as Obsidian has been detected among the older rocks is because in the lapse of time devitrification has more or less destroyed the glassy character of the older Obsidians. Whether it was worth while giving distinct names to rocks which practically differ only in age, is questionable. The time may come when our nomenclature can be considerably simplified, but it has hardly come yet. B. INTERMEDIATE ROCKS. The principal rocks of this class may be grouped under the following heads. B a. The Syenite Group. B b. The Trachyte Group. B c. The Mica-trap Group. B a. The Syenite Group. Syenite. The essential constituents of typical Syenite are Orthoclase and Hornblende. The whole rock is made up of crystals or crystal- line grains in contact with one another, and there is no paste or ground-mass. Sphene is an accessory so very commonly present as to be almost characteristic. The microscope shows that there are also present Plagioclase, Biotite, Apatite, Specular Iron Ore, and occasion- ally a little Quartz. Fluid-cavities have been noticed in the Orthoclase, and also in the Quartz when it is present. It requires only an increase in the percentage of the Quartz and Mica to convert Syenite into Hornblendic Granite, and rocks do occur which form connecting-links between the two. Auyite-syenite, Monzonite. The largest constituents of this rock are Orthoclase, Plagioclase, and Augite. At the typical locality of Monzoni the proportions of Orthoclase to Plagioclase are so variable that some specimens have a percentage of Silica that would entitle Confusedly-crystalline Rocks. 325 them to a place in the Acid class, and in others from the same rock- mass the Silica ranges so low that they might be placed among the Basic rocks. The rock contains besides as accessories Hornblende, Biotite, Magnetite, Sphene, Pyrites, and Apatite. Elceolite-syenite, Foyaite. A number of rocks somewhat varying in composition but closely allied to one another may be included under this head. Their essential constituents are Orthoclase or perhaps Microline, Plagioclase usually, but to a very variable amount, Horn- blende, and Elseolite. The last is a mineral which differs little if at all from Nepheline, the main points of distinction being that it tends to form stouter and larger crystals, and that it has a very conspicuous greasy lustre. As accessories the rock contains Magnetite, Sphene very commonly, Iron Pyrites, and Biotite. A variety containing Zircon and little Hornblende has been called Zircon-syenite. The Zircon-syenite of Norway is remarkable for the large number of rare minerals which it contains.* Miascite is another variety containing Sodalite ; Ditroite another which contains besides Sodalite a mineral Cancrinite, closely allied to Nepheline, but con- taining Water and Carbonates : it can be distinguished from Nepheline by its easier fusibility, by giving water in the closed tube, and by effervescing with acids. Quartz- freier Porphyr, Quartz-freier Orthoclas-porphyr, Syenit- porphyr. Some rocks of not very common occurrence, for which we have no name in English, are thus designated by German petro- logists. The minerals of which they are composed are the same as in Syenite, but instead of being directly in contact as in that rock, they are porphyritically embedded in a micro- or crypto-crystalline paste. Quartz is very often present, and the paste in some cases contains felsitic and glassy portions. The rock therefore stands somewhat in the same relation to Syenite as Quartz-felsite to Granite, or in many cases rather as Granit-porphyr to Granite. B b. The Trachyte Group. Trachyte. The most important constituents of Trachyte are Sani- dine and in many cases Plagioclase, though instances are not rare in which Plagioclase is absent altogether ; Magnetite and Apatite are also generally present. One or more of the minerals Hornblende, Augite, and Biotite come in to a subordinate degree. Sphene, Olivine, Sodalite, Hauyne, and Nosean occur as accessories, Quartz very rarely, Tridymite frequently. These minerals are embedded in a matrix of very variable character, frequently a mass of Felspar Microliths mixed up with granules and spiculse of Magnetite and Hornblende. In many cases a certain amount of isotropic matter can be detected in the matrix, sometimes as a mere skin incrusting the crystals, sometimes entering more largely into the composition of the matrix. Gas- and glass-cavities are met with in the Sanidine, fluid-cavities are of the rarest occurrence. The Sanidine crystals are often shattered in such a way as to convey the impression, that when the rock was * Zirkel, Lehrbuch der Petrog. i. 593. 326 Geology. poured out in a fused state it contained crystals of Sanidine ready formed, and that these crystals were broken up by being dragged along by the mass when it had been reduced to a pasty state by cooling but was still in motion. A peculiar rock, Domite, which may be placed among the Trachytes, occurs in Auvergne. The mass of it consists of a light-coloured, porous, friable, and brittle matrix, in which are embedded small glassy crystals of Sanidine and Plagioclase, plates of Mica, and some- times prisms of Hornblende. The matrix consists of granules and strings of Felspar, Microliths of Hornblende, grains of Magnetite, and some glassy matter : Tridymite also occurs somewhat plentifully in it. Pkonolite. The most important constituents of this rock are Sani- dine and one or both of the minerals Nepheline and Leucite. Hauyne and JS T osean are also very commonly present, and Sphene is frequently met with. Apatite and Magnetite come in in many cases, but to a comparatively small and very variable amount. Plagioclase is fre- quently but not invariably absent. Hornblende,* Augite, and Biotite also occur in some Phonolites, and among the accessories we may specially mention Tridymite and Olivine. Under its commonest and most typical form the rock consists of crystals of Sanidine, Nepheline, and very frequently Hauyne, embedded in a compact paste. The crystals may be either of macroscopic or microscopic dimensions. The paste is crypto- crystalline, and sometimes Sanidine sometimes Nepheline preponde- rates in it. Phonolite very generally shows a strong tendency to split into slabs, sometimes thin enough to be used for roofing. It bears the name Clinkstone on account of the clear ringing sound which these slabs give out when struck. This structure may be in some cases due to contraction during cooling : in some cases it may be produced by the arrangement of a large number of tabular Sanidine crystals in parallel layers with their flat faces all in the same direction. Andesite. The most important constituents of this rock are Plagio- clastic Felspars, Andesine, Oligoclase, and Labradorite ; Hornblende ; Biotite ; and Augite. The varieties which contain little or no Augite are distinguished as Hornblende-andesites, those in which Augite pre- ponderates over Hornblende as Augite-andesites. The typical forms differ from Trachyte in containing no Sanidine, but an Orthoclastic Felspar is present in some varieties, occasionally in sufficient quantity to make them approximate to Trachyte. Magnetite and Apatite are very generally present. Among the commoner accessories are Specular Iron Ore, Ilmenite, Garnet, Tridymite, and Iron Pyrites. Hauyne occurs in an Andesite from the Canaries. There is also very often but not always more or less of an amorphous paste. A majority of the Andesites are more or less porphyritic in texture, the crystals being embedded in a paste which is either crypto-crystalline or a mixture of crypto-crystalline and amorphous matter. Granular forms * Zirkel looks upon Hornblende as one of the principal constituents of all Phonolites. Confusedly-crystalline Rocks. 327 in which the crystals run about the same size throughout do however occur. Dacite. This rock differs from Andesite in containing Quartz. The Quartz presents itself both in well-formed crystals and in crystal- line grains : in some examples it contains fluid-cavities, in others numerous glass-cavities. The line between Andesite and Dacite is by no means sharply defined. Andesites are met with containing a small amount of Quartz, and as the percentage of this mineral increases we have intermediate forms which lead on to Dacite. For this reason Dacite is often called Quartzose-andesite. B c. The Mica-trap Group. The establishment of this group may be justified perhaps more on the ground of convenience than of sound classification. The rocks that we place in it however are very much alike in general appearance, and they further resemble one another in the fact that they all occur in dykes, so that it is very handy to treat of them all at the same time. Their most important constituents are Mica and Felspar, and they fall into two classes, Ninette when the felspar is Orthoclase, and Kersanton when it is Plagioclase. The Mica is so very plentiful that it is the most conspicuous element macroscopically, sometimes indeed the rock seems to be made up of little else but Mica. Minette. Numerous plates of Magnesian Mica and occasionally a few crystals of Orthoclase embedded in a paste which under the microscope is seen to be an entangled mass of Orthoclase and Mica. Plagioclase is usually absent, and where it is present it occurs only in small quantity. Hornblende and Augite occur in some Minettes, and Apatite, Magnetite and Iron Pyrites are common accessories. Minettes have often suffered much alteration and various secondary products have been set up by decomposition of their original minerals. Among the more important of these are Calcite, Limonite, Chloritic minerals, and Quartz. Typical Minette is evidently closely allied to Syenite, and by some authors the rock is called Mica-syenite. It differs from Micaceous Felstone in the general absence of Quartz, but if any of the Quartz that occurs in some Minettes be an original constituent, there are connecting-links between the two. Kersanton, Kersantite. Numerous plates and crystals of Magnesian Mica in a crypto-crystalline paste which is either composed of Plagio- clase (Oligoclase it is said) or is an intimate mixture of Plagioclase and Mica. Some varieties contain a little Hornblende, and these have been distinguished as Kersantite. By other authors the term Kersan- tite has been applied to those forms of the rock in which the porphy- ritic structure becomes very marked. Kersanton has often suffered from alteration with very much the same results as Minette. The affinities of Kersanton are with the Diorites, and it has been classed with the Mica-diorites.* * For a description of some English Mica-traps see Bonney, Quart. Journ. Geol. Soc. xxxv. (1879) 165. 328 Geology. C. BASIC-CRYSTALLINE ROCKS. The Basic-crystalline rocks may be ranged in the following groups. C a. Diorite Group. C b. Diabase Group. C c. Basalt Group. C d. Norite and Hypersthenite Group. C e. Nephelinite and Leucitite Group. C f. Glassy Rocks of Basic composition. C (/. Peridotite Group or Ultra-basic Group. C a. Diorite Group. The typical constituents of the rocks of this group are Plagioclase and Hornblende, and according to their grain they may be subdivided into two minor groups. The first may be distinguished as Granular Diorite, but is often simply called Diorite. In it there is no notable approach to porphyritic structure, the crystals or crystalline grains run pretty much of a size and form a granular aggregate. In the second subdivision there is a more decided tendency to porphyritic structure : crystals are set in a paste which is either micro- or crypto-crystalline. Such a crypto-crystalline paste plays somewhat the same part among Basic rocks as Felsitic matter among those of the Acid class, and is often described as Aphanitic. The rocks of the second subdivision may well be called Porphyritic Diorites; the term Porphyrite is also very generally applied to them, but if we use the latter we must add some qualifying prefix to it to distinguish the Porphyritic Diorites from the Porphyritic Diabases, and accordingly we style the rock Diorite-porphyrite. Diorite. The main essential constituents are Plagioclase and Horn- blende. Augite does occur in some varieties, but only to a small amount ; it is always associated with the fibrous form of Hornblende, and may be a product of the alteration of that mineral. Orthoclase, Apatite, and Magnetite are nearly always present as well, but only to a subordinate degree. Among the commonest accessories are Ilmenite, Biotite, and Iron Pyrites. The grain of the rock varies very much ; sometimes large crystals of Hornblende are embedded in a paste composed of small crystals of Plagioclase ; sometimes tabular crystals of Plagioclase are buried in a finely-crystalline mass of Hornblende ; sometimes the rock is made up of an interlacing mass of small crystals of both minerals. In some Diorites the Biotite becomes an important ingredient. These are distinguished as Mica-diorites. When, as happens occasion- ally, there is a marked decrease in the percentage of Hornblende, we get rocks which form connecting-links with Kersanton. Quartz-diorite. In addition to Plagioclase and Hornblende this rock contains Quartz as an essential constituent. Biotite is also usually a more prominent ingredient than in Diorite. The Quartz occurs as in Granite, filling, up the spaces between the other minerals, very rarely crystallized ; it contains fluid-cavities. Forms occur in which the Biotite increases, and the Hornblende decreases to such an extent that it sinks to the rank of an accessory, Confusedly-crystalline Rocks. 329 and in this way we have a passage into a rock called Mica-quartz- diorite. Some forms of Mica-quartz-diorite contain Augite and Calcite and no Hornblende. It is perhaps questionable whether the Augite and Calcite are not wholly secondary products, and whether these varieties are anything more than the result of the alteration of the Hornblendic type. It is in favour of this view that there are rocks in which the distinctive characters of the Hornblendic and Augitic varieties are united. Orthoclase is more plentiful in some forms of Quartz-diorite than in Diorite, and by its gradual increase connecting-links are formed between Quartz-diorite and Hornblendic Granite. Tonalite and Banatite are local varieties of Quartz-diorite. Porphyritic Diorite, Dior ite-porphy rite. The constituent minerals of this rock are the same as those of Diorite, but it differs from that rock in possessing more or less of a porphyritic texture. The paste is very variable in its composition and character ; micro- or crypto- crystalline in some instances, with more or less glassy matter in others. The glassy forms frequently show well-marked fluxion structure, and contain Spha3rulites and Trichites. The same rock-mass shows in certain cases a stony paste in some parts and a paste entirely glassy in others, with a gradual passage from one form into the other. But though the Porphyrites show this tendency towards a glassy condition, and approach very nearly in some cases to the state of simple glass, a perfectly glassy form, which would hold the same relation to them as Pitchstone to Felstone, has not been observed. All the different varieties depending upon mineral composition which we noticed among the Diorites recur among the Diorite-porphyrites. These are Quartzose and Quartzless forms, and others rich in Mica which correspond to the Mica-diorites. C b. Diabase Group. The rocks belonging to this group vary very much in composition and character, but they all contain as essential constituents Plagioclase and Augite. Like the Diorites they fall into two subdivisions according as they are granular or porphyritic in texture. The granular forms are called simply Diabase ; the porphy- ritic are distinguished as Porphyritic Diabase or Diabase-porphyrite. Alteration has frequently gone on in rocks of this group to such an extent that they are rich in Calcite, Limonite, Chloritic minerals, and other secondary products. So often is this the case that some authors have looked upon these minerals as essential constituents, and have proposed either to confine the term Diabase to rocks in which they occur and to consider it as merely an altered form of Dolerite, or to abolish the name altogether. But Diabases are known in which these secondary minerals are not found, and on this ground it is maintained that the species is a good one. If this view be taken, it is, it must be confessed, difficult to see where the difference lies between Olivine- bearing Diabase and those more coarsely-grained forms of Basalt which have been distinguished as Anamesite and Dolerite. Those who advocate the retaining of the name Diabase would confine it to the older Plagioclase-augite rocks, and would use Dolerite for rocks of the same composition and texture but of younger age. 33O Geology. The practice of giving different names to two rocks which differ only in age is perhaps of questionable utility, but as it is still adhered to by nearly all of the most eminent petrologists, we hardly like to take upon ourselves the responsibility of departing from it. And we have the less hesitation in retaining the present nomenclature, because, when we come to inquire into the way in which the different kinds of crystalline rock were formed, it will be shown that there are really substantial reasons for separating the Basalts from most, if not all, of the Diabases. Diabase. This rock is a granular mixture of Plagioclase and Augite, with nearly always more or less Magnetite which is often titaniferous. Some Orthoclastic Felspar is usually present as well ; Professor A. Geikie has described rocks, that are more nearly related to Diabase than to any other rock, in which the Orthoclase sometimes exceeds the Plagioclase in amount.* Apatite is also very generally present, but often to a small amount. Other constituents of less importance and occurring irregularly are Horneblende, Biotite, Enstatite, Quartz, and Olivine. The Plagioclase has been found to be in some cases Oligoclase, in some Labradorite, and in some Anorthite. Cavities containing water and liquid carbon dioxide have been noticed in the Augite of some Diabases. Quartz is rarely an original constituent ; when it is, it occurs as in Granite filling up the spaces between the other minerals, but it is not so rich in fluid-cavities as the Quartz of Granite. The Hornblende is in some instances an alteration product ; where it is an original constituent it is less in quantity than the Augite. As in Diorite, there are varieties of Diabase, such as Quartz-diabase, Quartzless Diabase, depending on the presence or absence of some of the less important minerals. The Diabases are a very numerous and varied class, and different forms have received various names. Augite-porphyry and Uralite- porphyry belong to this group, and also many of the rocks which have been called Melapliyres. Porphyritic Diabase, Diahase-porphyrite, differs from Diabase in possessing a more or less porphyritic structure. No hard line can be drawn however between the granular and the porphyritic forms. The constituent minerals of the two rocks are the same, and the only difference consists in the presence of a paste. This is sometimes glassy, sometimes glass partly devitrified, sometimes crypto-crystalline, and sometimes micro-crystalline, or it may present some one of the intermediate forms which exist between two of these states. In some cases the amount of glass is so large that the rock becomes entitled to the name of Diabase-pitchstone. Olivine-diabase. Olivine is present to a small amount in many Diabases ; when it is sufficient in quantity to become an essential constituent the rock receives this name. Olivine-diabase is liable to very great variations in the relative proportions of its constituent minerals. Melaphyr. There is perhaps no name which has been so ill used by * Trans. Royal Soc. Edinburgh, xxix. (1879) 487. Confusedly-crystalline Rocks. 331 petrologists as this. It has been employed by so many authors in so many different senses that if used without some qualification it has ceased to bear any definite meaning. We will here follow Kosenbusch in restricting it to the porphyritic form of Olivine-diabase. But even according to this author the porphyritic structure is very imperfectly developed in some of the rocks which he classes as Melaphyres, and in their case we confess we are at a loss to know exactly what constitutes a Melaphyr. Taking the term in the restricted sense mentioned, we have to note that glass- and stone-cavities have been noticed in the Plagioclase. Hornblende and Biotite are not common, but Biotite occurs more frequently than Hornblende, and in some varieties it is plentiful. The paste may be glass, or glass with Globulites and crystalline products of devitrification, or crypto-crystalline, or some intermediate form, or finely granular. Fluxion structure is sometimes well developed in it. Gabbro. This is generally defined to be a rock whose main essential constituents are Plagioclase and Diallage. We have already pointed out that it is very questionable whether Diallage has any right to the rank of an independent mineral species, and have given reasons for maintaining that it is no more than a peculiarly aggregated form of Augite. If this view be correct, Gabbro can hardly be looked upon as an independent rock-species, and must be considered as no more than a variety of Diabase, which differs from the generality of Diabases merely in containing Augite of the Diallage type in the place of Augite of the ordinary type. We may however provisionally retain the name, leaving it an open question whether it will be discarded or not when the day comes for reform in our petrological nomenclature. The Plagioclase is Labradorite or Anorthite or some allied form. The Diallage in some cases fills in the spaces between the felspar crystals, in others it occurs itself in well-outlined crystals. Magnetite, often titaniferous, is usually present, and Apatite is frequently plenti- ful. Other less important constituents are Orthoclase, noticed only in a very few cases, Hornblende, Biotite, Enstatite or Bronzite, and very rarely Quartz. The rock is often very coarsely grained, and contains crystal two inches or more in length ; but more finely-grained varieties are also met with. Cases occur where the Plagioclase is altered into a substance known as Saussurite and the Diallage at the same time into a product known as Smaragdite ; and the rock is then distinguished as Saussurite- gabbro. Olivine is by no means an uncommon constituent of Gabbro, indeed by some authors it is looked upon as essential. Where it is plentiful the rock is sometimes distinguished as Olivine-gabbro, but no hard line can be drawn between the varieties rich and those poor in Olivine. The relative proportion of the constituents is very variable in Olivine- gabbros, and they sometimes contain scarcely any felspar. C c. The Basalt Group. We have already explained that the name Basalt is applied to the more modern equivalents of the older Olivine-diabases and Melaphyres, and expressed an opinion that it is 33 2 Geology. decidedly an open question whether difference in age is a sufficient ground for the double terminology. Basalt. The most important essential constituents of Basalt are Plagioclase, Augite, and Olivine ; Magnetite, Ilmenite, and Apatite are very generally present as well. Sanidine arid Nepheline occur, but only to a small amount, and Leucite has been recorded in at least one instance. Specular Iron Ore is met with, but rarely. Many Basalts also contain Native Iron. Well-developed fluxion structure is common. The Plagioclase may be Oligoclase, Andesite, Labradorite, or Anorthite. It contains glass- and gas-cavities, but fluid-cavities are very rare. The crystals are not usually large, often a closely-inter- woven mass of small individuals. The Augite is in many, but not in all cases, the predominant mineral ; it occurs not unf requently in crystals of some size. The Olivine is mostly in crystals or crystalline grains of recognisable size. Hornblende and Mica do occur, but not often, though some varieties are rich enough in Mica to have been distinguished as Mica-basalts. There is besides a paste which varies very much both in amount and character. Sometimes it is a micro- crystalline aggregate of uniform grain ; sometimes a glass or a glass full of Trichites and minute dark globules ; or it may be a mixture of glassy and micro-crystalline matter. The Basalts have been divided according to their grain into (1) Compact crypto-crystalline varieties commonly called Basalt. (2) Micro-crystalline varieties, called Anamesite. (3) Macro-crystalline varieties, called Dolerite. Tephrite. In the rocks which we have classed as Basalts Nephe- line does occur occasionally, Leucite it would seem very rarely, and Olivine is an essential constituent. There is another class of rocks closely allied in many respects with the Basalts, but differing from them in these points. Olivine is either absent or present only to a small extent, and Nepheline and Leucite, or both, are among the most important essential constituents. To these the name Teplirite has been given. They may be thus defined. The most important essential constituents are Plagioclase, Augite, Nepheline, and Leucite. Hornblende, Biotite, Apatite, and Magnetite are very generally present as well. As accessories Sanidine occurs very frequently and Hauyne occasionally. When the Plagioclase is in large crystals, these are very rich in glass-cavities. Sanidine is usually present to a small amount only, but varieties occur comparatively rich in this mineral which form connecting-links between Tephrite and Phonolite. Nepheline is often porphyritically disseminated in well-formed crystals, but this is not always the case. Both granular and porphyritic forms occur, and glassy matter enters into the composition of the paste of the latter, sometimes in thin films between the crystals, sometimes to a larger extent. In both forms there is well-developed fluxion structure. The Tephrites have been divided into Nepheline-tephrites, Leucite- Confusedly-crystalline Rocks. 333 tephrites, and Nepheline-leucite-tephrites, according as one or both of the minerals Nepheline and Leucite occur in them. There is a rock which has been separated under the name of Tesch- enite, which seems to differ very little, if at all, from Nepheline- tephrite, the only point of distinction being that it contains Olivine. The belief that it is of older date than the Tephrites seems to be the main reason for giving it a distinctive name. C cL The Norite and Hypersthenite Group. The essential constituents of the rocks of this group are Plagioclase, and one or more of the minerals Enstatite, Bronzite, and Hypersthene. These three minerals are very difficult to distinguish from one another, especially when they have to be examined microscopically : it is indeed questionable whether they are distinct species, and not rather varieties of one mineral which differ only in the extent to which Magnesia is replaced by Iron. Consequently as yet no subdivisions have been established in this group. liosenbusch applies the term Norite to all its members ; these rocks in which it is supposed that Hypersthene can be certainly identified have been called Hypersthen- ites : but the mineral taken for Hypersthene in many of the rocks that were formerly classed as Hypersthenites has been shown to be Diallage, so that they are really Gabbros. Olivine, Diallage, Biotite, and in at least one case Orthoclastic Felspar have been noticed as accessories in Norite. The typical Hypers- thenite of St. Paul's Island, Labrador, for instance contains both Olivine and Diallage. The rocks of this group are of very limited occurrence. C e. The Nephelinite and Leucitite Group. In the rocks of this group one main essential constituent is Augite, and the other is one of the three minerals Nepheline, Leucite, and Hauyne. Magnetite and Apatite are usually present as well, and sometimes Sphene. In many members Olivine is an important ingredient : others contain little or no Olivine. Among the accessories are Hornblende and Biotite ; Orthoclastic and Plagioclastic Felspars are also present ex- ceptionally, but to so very small an amount that the rocks may be looked upon as practically free from Felspar. The members in which Nepheline is an essential constituent, while Leucite and Hauyne sink to the position of accessories, are called Nephelinites, and when they contain Olivine to any extent Nepheline- basalts. Where Leucite is the essential constituent, the name Leucitite is used ; and the forms rich in Olivine are styled Leucite-basalt. The name Leucitophyr is also applied to these rocks. The comparatively rare rocks in which Hauyne is the essential con- stituent are distinguished as Hauyninite or Hauynophyr. C /. Glassy Rocks of Basic composition. Glassy forms of rock are far less common in the Basic than in the Acid class, and the name Tachylite is generally used to include all that are known. Tachylite is a good deal like Obsidian, but when struck with the hammer it usually splinters up into small irregular chips, and does not 334 Geology. yield broad flakes and large surfaces of conchoidal fracture like that rock. Also Tachylite is easily fusible under the blowpipe, while in the case of Obsidian it is only the edges of thin splinters that can be rounded. Tachylite often shows under the microscope Globulites, Crystallites, Trichites, and other devitrification products, also minute crystals of many of the common constituents of the Basic rocks. As these embedded crystals increase in number and size passages may take place into a semi-glassy form of Basic rock. In the majority of cases Tachylite can be hardly said to constitute rock-masses ; it usually occurs in the form of a thin crust which coats these portions of a crystalline rock that have been suddenly solidified by contact with a cold substance. But in Kilauea in the Sandwich Islands there are vitreous basic lavas which form extensive bodies of rock.* Some Tachylite is certainly the glassy form of Basalt, for passages have been traced from one rock into the other. Thus Professor A. Geikie, in describing the Basaltic dykes of the island of Eigg, says, " Towards the edge of the vein the grain of the rock is usually very close, passing sometimes through various stages of flinty Basalt into bright, black, lustrous Tachylite." t That other Basic rocks also put on glassy forms is probable, but an absolute passage from the crystalline to the glassy state does not seem to have been observed except in the case of Basalt. Eosenbusch divides the basic glasses into Tachylite which is soluble and Hyalomclane which is insoluble in acids. C g. Peridotite Group. One essential constituent of the rocks of this group is Olivine, also called Peridote, whence the name. The other most important constituents are one or more of the following minerals, Augite, Enstatite, Bronzite, Hypersthene, Magnetite, Chrome Iron Ore, Ilmenite, and Picotite or Chrome Spinel.^ Hornblende, Biotite, and Apatite occur less frequently as accessories, and there is sometimes a little Felspar. It is evident that by ringing the changes through the above rather numerous list of essential constituents a great array of rocks could be formed, and the varieties that occur in nature are numerous. The following are some of the more important. Olivine-diallage-enstatite-peridotite or Lherzolite. The commonest variety. Olivine ; a mineral of the Pyroxene group, often Diopside ; a mineral of the Enstatite group ; with Picotite, sometimes in the rock, sometimes enclosed in the Olivine ; and Magnetite. Diallage-peridotite or Olivine-dicdlage Rock. Sometimes contains Garnet, when it has been called Garnet-olivine Rock. Enstatite-peridotite, where the main component besides Olivine is * Cohen, Neues Jahrbuch, 1876, p. 744. t Quart Journ. Geol. Soc. xxvii. (1871) 299. The Spinels are typically Aluminates of Magnesia (MgO.Al 2 3 ), but the Magnesium is often replaced by Calcium, Iron, and Manganese, and the Alumina by Ferric and Chromic Oxides. H. 8. Isometric, generally in octahedral forms. Infusible. Confusedly-crystalline Rocks. 335 either Enstatite, Bronzite, or Hypersthene. Garnets are abundant occasionally. These are comparatively rare rocks. Serpentine is a rock very largely composed of the mineral Serpentine ; it often contains Bronzite, Enstatite, Chrome Iron Ore, and many other minerals ; it is soft but tough, with a slightly soapy or unctuous feel, often of a dull green or greenish-grey colour, but frequently beauti- fully mottled red, green and purple, and veined with Steatite or Chrysotile. Serpentine has in all cases been produced by the alteration of some other rock. Now in the three rocks last named all the minerals and especially the Olivine are usually transformed to a large extent into Serpentine or Serpentinous products, and in some cases the alteration has gone so far that none of the original constituents remain and the rock has been wholly changed into Serpentine. Many Serpentines then are certainly altered forms of Peridotite, but we shall afterwards see that Serpentine may have arisen from the alteration of other rocks as well. Pikrite or Olivine-augite-peridotite. The Olivine is usually largely altered into Serpentinous and the Augite into Chloritic products. Magnetite and Picotite occur in the Serpentine. A small amount of Plagioclase is sometimes present, which is usually more or less altered into Saussurite and occasionally to Zeolitic minerals. A porphyritic form of Pikrite (Pihrit-porphyr) occurs in which the constituent mine- rals are embedded in a glassy paste, which often contains Globulites, Trichites, and other products of devitrification. Limburgite, Magma-basalt, is a name which has been applied to the more recent forms of Porphyritic Pikrite. Both the Augite and Olivine contain glass-cavities, fluid-cavities have been rarely observed in the Augite. The paste is glass or partially-devitritied glass, passing into crypto-crystalline matter. Dunite consists of Olivine and Chrome Iron Ore, with traces of Diallage and Enstatite. The average percentage of Silica in the rocks of this group is so low that they are sometimes distinguished as Ultra-basic rocks. Summary. The two following tables show the most essential points in the mineral composition of the chief varieties of the Crystal- line rocks, and the average chemical composition of the leading members. MINERAL COMPOSITION OF THE CRYSTALLINE ROCKS. Only the more important essential constituents are enumerated. ACID ROCKS. ( Granite. Quartz (seldom in crystals), Orthoclase, } Plagioclase, Mica, no paste. GRANITE GROUP. ) Haplite. Quartz, Orthoclase, Plagioclase. ( Greisen. Quartz, Mica, Schorl. ELVANITE r / Quartz (often in crystals), Orthoclase, UP ' 1 Plagioclase, in felsitic paste. 336 Geology. FELSTONE GROUP. QUARTZ -TRACHYTE GROUP. GLASSES. Orthoclase and Quartz. Quartz-trachyte. Quartz, Sanidine, in a stony paste. Rhyolite. Quartz, Sanidine, in a glassy paste. fitchstone, Obsidian, Pumice. SYENITE GROUP. TRACHYTE GROUP. \ MICA-TRAP GROUP. INTERMEDIATE EOCKS. Syenite. Orthoclase, Hornblende, no paste. Augite-syenite. Orthoclase, Plagioclase, Augite. Elceolite-syenite. Orthoclase, Hornblende, Elseolite (Nepheline). Phonolite. Sanidine, Nepheline, Leucite. Porphyritic, Syenite. Orthoclase, Hornblende, and paste. Trachyte. Sanidine, Plagioclase, Augite, Hornblende, Biotite, paste. Andesite. Plagioclase, Hornblende, Augite, Biotite, and paste. Dacite. Andesite with the addition of Quartz. Minette. Orthoclase and Mica. Kersanton. Plagioclase and Mica. BASIC ROCKS. ( Diorite. Plagioolase, Hornblende, no paste. | Quartz-diorite. Diorite with the addition of Quartz. DIORITE GROUP. [ Diorite-porphyrite. Plagioclase, Hornblende, in a paste. | Qaartz-diorite-porphyrite. Diorite-porphyrite with the I. addition of Quartz. Diabase. Plagioclase, Augite, Magnetite, no paste. Gabbro. Plagioclase, Diallage, Magnetite. Quartz-diabase. Diabase with the addition of Quartz. Diabase -porphy rite. Plagioclase, Augite, Magnetite, in a paste. Quartz-diabase-porphyrite. Diabase- porphyrite + Quartz. IJIABASE AM \ niton-no rlinhno* \ BASALT GROUP. NORITE AND HYPERSTHENITE GROUP. Melaphyr. Plagioclase, Augite, Olivine, Magnetite, in a paste. Olivine-gabbro. Gabbro + Olivine. Tephrite. Plagioclase, Augite, Nepheline, Leucite. I Plagioclase, Enstatite, Bronzite, Hypersthene. PERIDOTITE GROUP. GLASSES. > NepheHne - Peridotite. Olivine, Augite, Enstatite, Bronzite, Hy- persthene. Porpliyritic Peridotite, Limburgite, the same minerals in a glassy paste. Tachylite. Confusedly-crystalline Rocks. 337 O O Oxides of Iron and Manganese. a 8 8 g b b g g g O 00 8 fc x. 30 b b 8 S 53 S S 3 S S fl PPP S ? 000 00 S 9 b b 8 S ?? 5? g 00 O P ? 8 S b b b S S S r-l O O b b co cj rt i>i g s s & s CO 1 j it was subsequently truncated and gutted by a paroxysmal outburst, and its northern half almost en- tirely blown away. The point of discharge then shifted to 2, a cone was piled up around this vent, and in its turn ruined and broken down ; what remains of it forms the ridge B. C and D are the remnants of later cones which were formed when the eruptions burst out from the points 3 and 4. A further trans- ference of the vent to the north gave rise to Vulcano itself of which only the ruins remain. This mountain was in frequent eruption down to 1786 ; since then its - energy has declined, and it is now in the Solfatara stage. | Its fumaroles give off enormous quantities of Sulphur, g Sal-ammoniac, Boracic Acid, and other vapours which are collected and utilized for commercial purposes. To the north of Vulcano is a little cinder cone, and beyond that the small volcano of Vulcanello. All the successive positions of the vent lie nearly on a straight line, which doubtless coincides with a fissure that communicates with the underground reservoir : whenever from time to time a part of this fissure became so weighted with the accumulated discharges ^ of the eruptions that the steam could not break through the pile, the pent-up vapour sought an easier exit further to the north. Of Puys we find excellent instances on the flanks of Etna. That mountain is now so lofty and well-built that lava is seldom pumped up to its summit, and it can only release itself by tearing openings in the flanks ^ of the mountain. Each of these is the origin of a minor or parasitic cone, hundreds of which are sprinkled over the slopes of this volcano. The formation of dykes has also been witnessed on Etna. In the eruption of 1669 "a fissure six feet ^ broad, and of unknown depth, opened with a loud crash, and ran in a somewhat tortuous course to within a mile of the summit of the mountain. It emitted a most vivid light," from which we may conclude that oj it was filled with melted lava, which on hardening would give rise to a dyke. " Its length was twelve miles. Five other parallel fissures of considerable length afterwards opened one after the other."* * Lyell, Principles, tenth ed. ii. 21. 352 Geology, Submarine Eruptions. Volcanic outbursts take place from the sea-bed as well as on dry land. We have not the same oppor- tunities of examining the former class of eruptions as of the latter, but enough is known to warrant the belief that the two differ in no essential respect whatever. The flow of lava beneath water, though at first sight an unlikely occurrence, has been repeatedly noticed. At first contact a certain amount of vaporization takes place, but a crust rapidly forms round the lava, which prevents the water coming in contact with the molten interior of the stream, and in consequence of its low conducting power checks the escape of heat. In this way a constantly lengthening tunnel is formed, within which the molten matter pursues its course often to large distances : and if the discharge take place in deep water, it is conceivable that the pressure of the overlying fluid will check the escape both of elastic vapour and of heat from the lava, and so keep it fluid for a longer time, and cause it to spread out in wider and more regular sheets, than if it had flowed out in the open air. Dispersion of Ash and flow of Lava beyond the Cone : Prolonged Dykes. So far we have considered only that part of the products of an eruption which go to the building up of the vol- cano : these however in many cases form only a small part of the whole discharge. The ashes and other ejected materials are often cast forth far over the surrounding country, and when finely divided are carried by the wind to enormous distances. The lava streams extend far beyond the base of the cone, covering in some cases hundreds of square miles of country, and even pursuing their course over the sea- bottom. We have not the same opportunities of tracing the extension of dykes as we have in the case of subaerial products, because they are formed beneath the surface, but there is every reason to believe that they are not confined to the cone itself. In many cases volcanic matter has been observed to be emitted from a number of vents ranged in a straight line for miles across the country ; and it seems reasonable to suppose that such points of escape lie on a great underground fissure injected with lava, the cooling of which, on the cessation of the vol- canic activity, will give rise to a prolonged dyke, similar to those which may be observed on a smaller scale in the walls of the cone itself. A case occurred in the eruption of Skaptar Jokul in 1783. "Lava was emitted consecutively at several points on a linear range of two hundred miles. No doubt an underground fissure of this length at least was injected with lava by that eruption, and remains now as a dyke traversing the substrata."* Fissure Eruptions. The streams of lava poured out by active volcanoes frequently flow far enough to cover enormous areas. In 1783 Skaptar Jokul sent out a stream fifty miles in length with a breadth in places of fifteen miles. But there are sheets of lava com- pared with which this is a mere pigmy. In the Western Territories of the United States one may ride for days and see nothing but lava, and in the sides of the ravines which the rivers cut across the country, * Scrope, Volcanoes, p. 52. Volcanic Products. 353 the sheets are seen to attain enormous thicknesses and to consist of bed piled upon bed to a depth of hundreds of feet. No single cone could have been the parent of such enormous spreads, but it might be conceived that the flows from numerous vents had so far run to- gether as to become undistinguishable and produce the impression of a single sheet. It is said however that as yet no signs whatever of any cones have been detected in the district. To meet this difficulty it has been suggested that dykes are numerous, and that the discharges have taken place not from isolated openings, but along the whole length of numerous fissures, and the name Fissure Eruptions has been given to outpourings of this kind.* An explanation of this nature, if admissible, would account for the observed facts very satisfactorily, but till something of the nature of a Fissure Eruption has been wit- nessed, it cannot rise above the rank of an hypothesis. When the distribution of volcanoes is studied on a map of the world, one is at once struck by the way in which they range themselves along lines approximately straight, and there is good reason to think that these lines mark the direction of fissures in the earth's crust. But, as far as observation goes, it seems to be the habit of pent-up volcanic products to relieve themselves at certain outlets only, and not along the whole length of a fissure. Mr. Clarence King describes some ridges in the Truckee Range whose structure is well accounted for on the supposition that they were formed by fissure eruptions. Each ridge consists of a number of thin beds of Basalt, and the beds curve over the summit of each ridge and dip down on either side in the same direction as the slope of the ridge. He thinks that the Basalt is the overflow of a dyke, and that the lava discharged during each eruption ran down over the sloping surface of the sheet that had been poured out by the eruption that immediately preceded.! SECTION III. VOLCANIC PRODUCTS. The products of Volcanic Action may be subdivided into 1. Molten or Lavas. 2. Fragmental or Ashes. 3. Gaseous. 1. LAVAS. Fluidity. The degree of fluidity of lavas varies considerably, but it is rarely if ever the case that they are in a state of perfect fusion. J * Geikie, Nature, xxiii. (1880) 3. t Geol. Explor. of 40th Parallel, i. 673. % "In the case of a metal in a state of perfect fusion, it is conceived that no particle of finite dimensions retains its solidity. In the case of fluid lava, it is supposed that a large portion of the mass consists of small but finite particles, which retain severally their solidity, while their relative mobility is maintained by the remaining portion of the mass intervening in a more perfect state of fluidity between the solid particles, and consisting partly of internal elastic vapours, which by their ascending movements keep the component particles in a constant state of ebullition " (Hopkins, Report on Elevation and Earthquakes, British Association, 1847). Z 354 Geology. Some lavas which have hardened into a glass have probably made the nearest approach to this state, but in the majority of cases fluid lava seems to consist of a mass of crystals and other solid particles enveloped in a pasty, imperfectly-fused mass which is permeated throughout by superheated steam and other gases. It is the presence of these elastic fluids that gives to most lavas their apparent liquidity, and the ability to flow in streams ; they move indeed somewliat for the same reason as a mixture of sand and water, which will run down a slope on which dry sand would lie. Mr. Scrope has aptly compared lava to " certain stages of the manu- facture of sugar, when the matter consists of a soft mass, or ' magma ' of granules or imperfect crystals enveloped in a liquid (syrup), which being subsequently dried by evaporation or drainage, consolidates into a hard substance formed of interlacing crystals." * Of the presence of water in lava there is abundant proof. The column of vapour which rises from the crater during a volcanic erup- tion, when not carried away by wind, condenses and falls as rain ; great clouds of steam rise from the surfaces of cooling lava-flows, and large bubbles can be observed making their way to the top and bursting out in steam-jets with incessant explosions. It is the bub- bling up and escape of this vapour that fills lava with the cavities and vesicles that are so characteristic of it, specially over its upper surface. Some difficulty may at first be experienced in understanding how water can exist in the middle of a mass so intensely heated as lava ; a much lower temperature, it might seem, would be sufficient to vaporize it and expel it as steam. But it must be recollected that the temperature at which water is converted into vapour depends on the pressure to which it is subjected ; increase the pressure, and you raise the boiling-point. Deep down in the volcanic focus the weight of the overlying lava will evidently exert an enormous pressure on any water that may be present. Possibly this pressure may be sufficient to keep the water still liquid, in spite of the intense temperature ; possibly it may not be able to prevent the formation of steam, but only to check its escape. In the first case, if the pressure be sufficiently relaxed, the water will flash suddenly and with violent explosion into vapour ; in the second, under similar circumstances, a mass of high-pressure steam will rush out. When the struggles of the imprisoned vapour to get free have raised the lava sufficiently near the surface to produce the requisite decrease in pressure, one or both of these results is produced, and the explosions from the crater, or the outbursts from the surface of the lava stream, are the result. We must also bear in mind, in connection with the present subject, that there are circumstances under which water, even under an absence of pressure, may be exposed to high temperatures without being vaporized. If water be dropped on highly-heated surfaces, it is not converted into vapour if the tempera- ture be sufficiently high, but rolls about in a spheroidal mass and slowly evaporates without boiling. Possibly water may exist within heated lava in this spheroidal condition. It is also known that, when a fluid is shut up in a closed space and heat is applied, the pressure of * Volcanoes, p. 45, and pp. 121, 122. Volcanic Products. 355 the vapour produced is able to keep a portion of the fluid still liquid up to temperatures far higher than that at which it boils under atmos- pheric pressure. Under such circumstances M. C. de la Tour raised water to a temperature of 773 F. before it became wholly converted into vapour. A consideration of these facts tends to remove the difficulty which must at first be felt in attempting to realize how water can exist in large quantity in the volcanic focus, but whether we can be fairly said to have any knowledge of the physical state of water under the cir- cumstances of the case is a very open question. The tension of aqueous vapour has been determined experimentally for comparatively small temperatures and pressures, and an empirical formula connect- ing the three has been deduced which is correct within the limits of experiment, but which cannot safely be extended much beyond them. How enormous are the pressures and temperatures with which we have to deal a single instance will show. During the eruption of Cotopaxi in 1877 the lava was pumped up to the very summit of the mountain, more than 19,000 feet above the sea and flowed over the rim of the crater. Supposing the lava to have been generated no lower down than the sea-level, this column would exert a pressure of more than 1300 atmospheres, and the temperature must be high enough to liquefy lava under this pressure. ]S T o experience which we can command will enable us to form an idea of what would happen to water placed under these conditions. The fluidity of any lava stream will depend on its composition and the temperature to which it is subjected : cceteris paribus, lavas of a basic are more easily fusible than those of an acid composition ; hence the former often issue in a state of sufficient fluidity to allow of their flowing to considerable distances, and it is lavas of this class that as a rule form the longest and most extensive streams. The acid lavas on the other hand furnish the most striking instances of imperfect fluidity ; in some cases they ascend in such a thick pasty state that they consolidate in dome-shaped mounds immediately around the vent ; given heat enough however, even the most highly acid lavas attain a high degree of fluidity and give rise to flows of considerable dimen- sions. Motion of Subaerial Lava Streams and Texture of their different Parts. On account of its pasty condition and the rapid cooling of its exterior surfaces, the motion of a lava stream differs considerably from that of a perfect fluid. The surface hardens into a crust of slags and cinders, which on account of its low con- ducting power checks the escape of heat from the interior of the mass, and allows the central part of the flow to preserve its liquidity and capability of motion sometimes for years after the upper portion has become hard enough to allow of its being walked over with safety. The diagram in fig. 114 illustrates the variations in texture which different parts of a lava-flow exhibit. At the surface, where there is nothing to check the escape of the steam, it boils up freely and blows the pasty matter out into numerous and large vesicles. These are pulled out by the motion of the stream and elongated in the direction 356 Geology. in which it is flowing. Lower down the weight of the overlying lava hinders the discharge of steam, the vesicles become smaller and less Fig. 114. IDEAL SECTION OF SUBAKRIAL LAVA-FLOW. numerous, and at last we reach a level where they disappear altogether. In the upper part of the stream the heat escapes by radiation or by conduction through a thin layer of lava; cooling is therefore rapid and unfavourable to the production of crystals ; the texture is glassy or stony and crypto-crystalline. But nearer the interior, where the heat has to travel through a large thickness of lava of low conducting power before it can make its escape, the rate of cooling is slower, and the lava when it hardens becomes more largely crystalline. Crystalline texture will be best developed in a belt about the centre of the stream ; below this the chilling influence of the ground will make itself felt, the lava will cool somewhat more rapidly and its texture will be more stony. The base has often a characteristic structure : the crust formed on the surface is dragged on by the motion of the still fluid portion underneath and is broken up into ragged cinder-like fragments. These fall over the front and are piled up there in a heap ; they are pushed forward by the advance of the fluid interior portion, so that the face of a lava-flow frequently gives no indication of the real com- position of the stream, but shows only a pile of jagged fragments rolling one over the other with a loud crackling grating noise. While the upper part of the pile is thus being constantly shoved forward, the fragments which lie at the bottom are prevented by the roughness of the ground from being pushed ahead, they lie comparatively undis- turbed and form a rugged pavement, and the fluid lava running in between them binds them into a solid mass. From their resemblance in shape to cinders the fragments are often denoted by that term, or they are called Scorice : since they have come originally from the surface of the flow, they will be vesicular, and hence the base of the stream is frequently Cindery or Scoriaceous and Vesicular. A subaerial lava-flow then under its most typical form possesses a Vesicular Top, the vesicles being dragged out in the direction of motion ; a stony interior becoming more distinctly crystalline towards the centre, and a Cindery and Vesicular Base. Slaggy, Ropy, and Scoriaceous structures and glassy matter are also common on the top. Volcanic Products. 357 The crust too is sometimes broken up into slabby masses, which are heaped up arid jammed against one another till they form a surface of almost indescribable ruggedness. Characters of Lava which has hardened under pres- sure. Where lava is forcibly intruded through other rocks, the pressure to which it is subjected prevents the formation of vesicular, slaggy, or cindery texture. The bounding surfaces of the intrusive mass cool too rapidly to allow of the formation of crystals and are usually dense and compact, sometimes glassy ; the centre, which cools more slowly, may show a more crystalline texture. Formation of Crystals in Lava. There can be little doubt that in many cases it is the rate of cooling which determines whether a lava shall consolidate into a coarsely-crystalline, a crypto-crystalline, or a glassy rock, but the student must not jump to the conclusion that every coarsely-crystalline rock he meets with necessarily owes its texture to its having cooled slowly. There are lavas containing large crystals whose superior coarseness of grain cannot be accounted for in this way, and in which the crystals must have been present in the lava, ready made so to speak, at the time of its emission. And such a supposition presents no difficulty. We have seen that most, if not all, lavas consist of a greater or less proportion of solid particles enveloped in a mixture of imperfectly-melted matter and steam, and in the lavas we are now dealing with, some of the solid particles are well-formed crystals. The lava of Stromboli must be full of such crystals, for the mountain showers out not only blocks of rock and ragged fragments of lava but also perfectly-formed crystals of Augite in great numbers. There are two ways of accounting for the presence of these crystals. It is possible that, as the lava rises from the intensely-heated depths of the volcanic focus into higher and somewhat cooler regions, portions may consolidate and crystallize while the mass remains in a more or less molten condition. Crystals formed in this fashion are often cracked and shattered and have their edges and angles rounded off as they are rolled along by the advancing flow. These larger crystals too often show that they have been formed under pressure by containing fluid-cavities. It is possible in some cases to determine the order in which the different minerals crystallized out from the fluid magma. Crystals of one mineral occur enclosed in those of another mineral. This is sometimes due to alteration, but where no signs of rearrange- ment can be detected, we may safely infer that the enclosed mineral crystallized before that which surrounds it. Again one mineral often occurs not in well-shaped crystals but in crystalline masses which are moulded on the other crystals, and bear the impression of the striae which cross their faces. Such a mineral has been the last to solidify. It requires some little care occasionally to distinguish between true crystalline faces, and the smooth impressions of the faces of embedded crystals. Again crystals met with in lava may have come there in the follow- ing manner. We shall see in a subsequent chapter that many rocks have been subjected to a process known as metamorphism, by which crystals have been developed in them, though they were previously 358 Geology. non-crystalline. We are very much in the dark as to the exact manner in which the change has been produced ; but there is little doubt that three necessary and powerful agents in bringing it about are heat, pressure, and the presence of water. All three of these are at work in the depths of a volcano ; and there can be little doubt that the manufacture of crystals by metamorphic action must be constantly going on in the rocks surrounding a centre of volcanic activity. When those rocks are shattered by the throes of an eruption, we can readily imagine that the crystals will be shaken loose, and falling into the more fluid portions of the lava, will be carried out when it is discharged, and embedded in it when it consolidates. Crystals formed in this way are more likely to occur in Ashes than Lavas. Many of the rare minerals for which Vesuvius is noted, Idocrase for instance, are found in blocks of Limestone which have come from the walls of the volcanic reservoir. The Limestone has been subjected to Meta- morphic Action, and blocks of it have been torn off and thrown out of the vent and embedded in the Breccias of the mountain. It may not be always possible to say whether the crystals in lava have been formed during consolidation or previously to emission, but we must be careful not to lose sight of the possibility of coarsely-crystalline texture having been produced in more than one way. Bedded Structure. When one flow has had time to harden before another has been laid down upon it, lava streams assume a rude sort of bedding ; and here we find one exception to the generalization that Crystalline rocks are unstratified. The beds however will be wedge-shaped and ill-defined, and more like the irregularly-stratified accumulations of shallow water than the evenly-bedded rocks pro- duced by slow and regular deposition of sediment. 2. FKAGMENTAL PRODUCTS. The fragmental products of volcanic action are of two kinds : first, the ejected masses of solid pre-existing rock through which the erup- tion has burst its way ; secondly, portions of the melted lava, which have been torn off and tossed up into the air in a melted or pasty condition, and hardened as they fell. Deposits of volcanic ash generally consist of a mixture of both kinds of materials, and vary in grain from the finest and most impalpable dust to coarse accumulations containing angular blocks up to several tons in weight. Thus, to speak first of the coarser ejectments, " by the eruption of Cotopaxi in 1533, the plain around the foot of the mountain was strewed through a radius of fifteen miles and more with great frag- ments of rock, many of which measured as much as nine feet in diameter" (Scrope, Volcanoes, p. 55). Similar ejected masses occur in rocks of older date but of the same origin. Fig. 115 shows a group of such blocks embedded in stratified ash on the shore near North Berwick. The waves have stripped off the finer and softer parts of the deposit by which these masses were originally surrounded, but boulders equally large and angular may be seen hard by, still enclosed in ashy layers that have not yet undergone denudation. I Volcan ic Products. 359 know of nothing which brings home so forcibly to the mind the enor- mous energy of volcanic forces as the sight of these huge missiles, and Fig. 115. LARGE ANGULAR BLOCKS EMBEDDED IN VOLCANIC ASH, NORTH BERWICK. an attempt to realize the power necessary to tear them from their bed and hurl them forth broadcast over the country. A large portion however of the matter thrown out of volcanoes is tossed up over and over again till it becomes broken and ground down to stones, dust, and powder. Out of this finer material, together with the portions torn off the surface of the liquid lava, the greater part of ashy formations are made up ; it forms the paste in which the large blocks, when present, are embedded, or it constitutes the whole body of the deposit. The following are the chief terms used in describing volcanic frag- mentary productions. Scoriae are the ragged fragments of lava which have hardened into cinder-like forms. Bombs are portions of lava which have been thrown out in a liquid state, and owing to a rapid rotation in their path, have assumed a rudely spherical shape : they are often hollow. The smaller fragments go by the general name of Lapilli or Volcanic Stones. Puzzolana is a name given by the Italians to still finer volcanic powder ; and the finest dust of all is called in the same language Ceneri, or Ashes. Structure of Subaerial Ashy Deposits. The structure and arrangement of deposits formed out of volcanic ejections will vary according as the latter fall on the land or into water. In the first case there will usually be an absence or an imperfect degree of bedding, and a tendency more or less pronounced to a con- fused grouping of the materials. When a subaerial accumulation of ejected materials contains many large blocks, and when its components are huddled together in a pell- mell way without regard to size, shape, or weight, it forms a Volcanic Agglomerate ; and if the larger fragments are markedly angular, this 360 Geology. peculiarity may be denoted by the qualifying term Brecciated. A Volcanic Breccia is a rock of similar character, in which the confusion and absence of arrangement is less marked. The finer fragmentary productions, which float through the air often for long distances, settle down more gently and regularly, and give rise to deposits with some appearance of bedding, occasionally with very pronounced and regular bedding. The name Tuff is applied to such deposits. The finer ashes in their passage through the air often become mixed with water, either rain or condensed steam issuing from the volcano ; in this case they cohere more readily when they fall, and give rise to a harder and more compact rock than that formed by accumulations of dry ash. The beds that cover Herculaneum are of this nature. Structure of Subaqueous Ashy Deposits. The frag- mentary productions that fall into still water will produce rocks differ- ing but little from the more gently formed of the subaerial ashes ; but streams, tides, and currents will round and distribute the fragments they receive, and will give rise to Conglomerates and gritty rocks of various degrees of coarseness and with well-marked bedding. These may be distinguished as subaqueous Tuffs. It may well happen that water into which volcanic ash is being showered receives at the same time sedimentary materials from rivers that empty into it. In this way there will arise rocks of a mixed origin : when the mechanical sediment is mainly Sand, we shall get an Ashy Sandstone ; if Limestone is being formed in the water, there will result a Calcareous Ash ; and so on.* All these fragmentary deposits are capable of enclosing and pre- serving the remains of plants and animals which lived at the spots where they fell, and such remains are occasionally found in them. The term Pyroclastic is sometimes used to denote fragmental rocks of volcanic origin. Torrential Accumulations of Ash. Intermediate in character between true subaerial and true subaqueous deposits of ash, are the accumulations produced when ashy matter that has already settled on the ground is torn up and swept away by heavy rain or floods. Violent floods are common accompaniments of volcanic outbursts ; they are produced by the sudden condensation of the emitted steam causing torrential rains, by the bursting of lakes which have gathered in the craters during periods of rest, or by the sudden melting of snow when the volcano rises above the snow-line and the rapid development of heat within the mountain or the outpouring of lava liquefies instan- taneously its icy mantle. In the eruption of Cotopaxi in 1877 the lava boiled over the rim of the crater, not in several streams, but in a continuous sheet all round, spread over the snow and ice of the summit peak, and melted them with almost inconceivable rapidity. Great sheets of water tore in torrents and cataracts down the steep sides of the mountain and rushed through the gullies that trench its slopes, sweeping with them stones, ash, and dirt, and forming huge deluges of pasty mud. Towards * For a beautiful example of a volcanic stone embedded in sedimentary deposits see Geol. Mag. i. p. 24. Volcanic Products. 361 the lower part of the mountain, where the ravines shallow and open out into broad valleys, the mixture of mud and water spread itself out, and rushed on with such violence that large buildings were utterly swept away, and the whole country was buried under a great ocean of mud tossing about in the wildest confusion. A ravine 100 metres broad and 60 deep was filled to the brim and the contents overflowed on either side, and a comparatively flat tract a square mile in extent was covered to a depth of from 20 to 30 metres. The materials swept down were such as are found on the flanks of a volcano, blocks of lava, pumice, and sandy ash : when the debacle reached the limits of vege- tation it caught up plants, trees, and turf which were entombed as the mud came to rest below. The term Moya, is applied in the district to accumulations of mud and debris produced in this manner. The Italians call these mud streams Lave di fango, the molten lava being Lave di fuoco. Accumulations produced in a similar but perhaps less violent manner fill up many of the valleys among the extinct volcanoes of the Eifel. They consist very largely of fragmentary pumice, and contain blocks of Basalt, Phonolite, burned Clay-slate, and the carbonized trunks and branches of trees. The deposit is known as Duckstein or Trass, and is employed in the manufacture of cement that sets under water. The Peperino of Central Italy is another deposit of this class, formed largely of Trachytic fragments ; their mixture with water has formed a substance that has set into a rock hard enough to be used for building. Other fragmental Volcanic Accumulations. When a lava- flow is suddenly advanced into water, or is ejected beneath water, it sometimes happens that the lava is instantly broken up into fragments. Fragmental matter thus produced may often be distinguished from ejected ashes in this way : each grain of the latter is more or less glazed on the outside by the heat of the volcano, the fragments of a lava stream disintegrated by the sudden action of water show no such glazed coating. The crust that forms on the outside of a lava-flow is of a brittle glassy nature, it is frequently so completely shattered, as it is dragged on by the motion of the underlying fluid portion, that it falls into a mass of angular fragments. These sink down into the pasty stuff below, and the whole solidifies into a Breccia, which might be easily confounded with a Breccia formed of ejected fragments. Such Breccias are the constant associates of Rhyolitic Lavas. Volcanic ash may also be showered over the surface of a lava stream while the latter is still soft, and a rock is then formed of lapilli em- bedded in a paste of lava. Volcanic ashes and tuffs are sometimes described and classified according to the rocks that predominate in their composition. Thus an Agglomerate mainly made up of blocks of Basalt may be spoken of as a Basaltic Agglomerate ; a Tuff composed of lapilli of Trachytic lava may be called a Trachytic Tuff. Some of the rocks of this class however are composed of such a diversity of materials that such a system of nomenclature could not be applied to them. Interbedded with those old Crystalline rocks, which we have learned 362 Geology. to look upon as lavas, are numerous deposits, many of the components of which are readily seen to be fragmental ejections from volcanoes. In some cases the ejected matter has fallen on land, and rocks have resulted, resembling in every respect Volcanic Agglomerates, Breccias, and ashy accumulations of finer grain. In other cases the deposit has been of subaqueous origin. Instances will be given in the next section. 3. GASEOUS AND SUBLIMED PRODUCTS. Steam is the vapour most largely discharged by volcanoes, and is given off in greatest abundance when they are at the height of their activity. They evolve also other vapours and gases, and these become most conspicuous in the Solfatara stage. Among gases Carbono- Dioxide is the commonest : it often continues to be given off long after all other indications of volcanic action have ceased. It may be produced by the underground calcination of Limestone. Sulphur Dioxide (SO 2 ) is very common. It gives rise to incrustations of Sulphur : it also by oxidation is converted into Sulphuric Acid (H 2 S0 4 ). Sulphuric Acid acting on the surrounding rocks forms Sulphates of Alumina, Lime, Magnesia, Soda, Potash, and Iron. Trachyte is thus converted into Alunite, and other transformations of a similar kind are brought about. The rock is sometimes so thoroughly corroded that all the bases are removed and there remains only a white, chalk-like mass of Silica. According to Bunsen, Sul- phuric Acid is converted into Sulphuretted Hydrogen (H 2 S), the decomposition being promoted by the presence of water and the adjoining porous rocks. The further decomposition of Sulphuretted Hydrogen gives rise to deposits of Sulphur. This may happen thus H 2 S + = S + H 2 O, or Sulphuretted Hydrogen and Sulphur Dioxide may mutually act on one another, and Sulphur and Sulphuric Acid be produced. Hydrochloric Acid is also a frequent product, derived very likely from the salt of sea-water, which we have seen is a probable source of the water of the volcanic foci. Hydrochloric Acid corrodes and decomposes the adjoining rocks and forms manifold new compounds out of their constituents. Oxygen, Nitrogen, Hydrogen, and Ammonia have also been noticed. Among the substances sublimed are Ammonium Chloride or Sal- ammoniac, Chlorides of Iron, Copper, Calcium, Magnesium, and Sodium, and the volatile metals Arsenic, Antimony, and Mercury. The Ferric Chloride gives rise by decomposition to Specular Iron Ore,* crystals of which cluster in the chinks of the lava. Chlorides of Calcium and Magnesium when acted on by Sulphuric Acid give rise to Calcium and Magnesium Sulphates, t Boracic Acid is carried up Remnants of Old Volcanoes. 363 in a finely-divided state by steam in large quantity at Voltura in Tuscany and in the crater of Volcano, and is obtained at both places for commercial purposes. Among these substances the Sulphates of Lime and Magnesia have a special interest for the geologist because they are the materials out of which Magnesian Limestones and Gypsums may be manufac- tured. Specular Iron Ore too is the colouring matter of the red beds which are so constantly associated with the chemically-formed mem- bers of these classes of rock. It is, as has been already mentioned, to a volcanic source that we must look for the materials of the rocks formed by precipitation. Hot and Mineral Springs. These are plentifully produced during the final stages of a volcano's existence, and they have been in many cases the sources of the chemically-formed siliceous and cal- careous rocks described on pp. 242 and 243. For an account of the phenomena of Geysers the reader may turn to Lyell's " Principles," chap, xxxiii., and the Keports of the United States Geological Survey of the Territories for 1872 and 1873; and for an explanation of the mechanism of their action to Tyndall's " Heat a Mode of Motion," p. 122. SECTION IV. REMNANTS OF OLD VOLCANOES. The main features and chief products of the volcanic action of the present day have been now described, and we have seen that certain of the older Crystalline rocks and their accompanying fragmental accumulations resemble the latter so closely that there can be no doubt that they were produced by similar agencies in bygone times. We wdll now inquire whether an examination of the geological record will enable us to point out the site of former volcanic discharges, whether any ancient volcanic cones are still recognisable, and generally to what degree of detail we can read the history of volcanic action in the past. Ancient Volcanic Cones. We can scarcely expect to find cones of any antiquity often preserved : modern volcanic mountains are already becoming scarred and seamed by subaerial denudation, and a long continuance of this action must at last wear down entirely the loose materials of which the greater part of a volcano consists, while submergence beneath the sea would sweep it away altogether. As a matter of fact however the number of volcanic hills that still remain in a fair state of preservation, to attest the former presence of volcanic activity in districts where neither history nor tradition give any hint of its existence, is larger than would at first sight be expected. In Auvergne, for instance, the Eifel, New Zealand, and elsewhere, we have not only sheets of solid lava and beds of fragmental ejections, but there are still standing the cones which were piled up round the vents from which these issued, worn and withered indeed by long exposure, but still retaining enough of their characteristic structure and outline to make it evident to the most casual observer that they 364 Geology. owe their origin to the same processes which are piling up our modern volcanoes. In all these cases however the date of the eruptions, though historically most distant, falls into a late period of geological chronology. As we go back in geological time, we find the remains of the easily-destroyed volcanic cone to become rarer and more and more fragmentary and very soon to vanish altogether. Remains of Central Plug of Lava. But even where this has happened, we can still sometimes fix the site of an old volcanic vent : the plug of lava, which hardens in the chimney when a volcano dies out, is of tougher stuff than the cone which surrounds it, and the lower portion of this central core often remains standing up, like a massive column, when all the rest of the volcanic hill has dis- appeared. Other Proofs of Old Volcanic Action. The records that remain of periods in the earth's history still more remote are seldom perfect enough to enable us to determine the exact spots on which eruptions have burst forth, but among them we find distinct proofs of volcanic activity in the presence of rocks, both molten and fragmentary, of undoubtedly volcanic origin. Thus, though the evidence, as we go further back, becomes less and less complete, there is always enough to assure us that during no epoch, of which any history has come down to us, was our earth without volcanoes. Example of Arthur's Seat. A very good instance of a fairly well-preserved volcano, of much older date than those mentioned a little way back, is furnished by Arthur's Seat at Edinburgh ; and as this is riot only an excellent specimen of its kind, but furnishes an admirable instance of the methods employed by geologists to decipher the records of ancient eruptions and to determine the mode of origin of different kinds of volcanic products, we will describe it at some length. Fig. 116 is a section of this hill, and fig. 117 a rude diagrammatic view of its western face. The body of it consists of Shales, Sand- stones, fine Conglomerates, and impure Limestone of marine origin, (1) in the figures. Along with these are beds of lava and volcanic ash, formed contemporaneously on the floor of the same sea which received the sedimentary materials of (1). The lowest of these is a sheet (2) of Doleritic rock ; then follow Sandstones formed partly of mechanical sediment and partly of ash, which was showered into the water at the same time as sand was brought into it by running streams, and some cherty Limestone (3) ; above these come well-bedded strata of ejected materials, the components of which were shot out into the water and then arranged in stratified layers, forming a volcanic tuff (4) ; the series is continued by flows of Basaltic and Felspathic lavas (5) and (6) ; and these last are covered by purely sedimentary Shales and Sandstones (7). That the beds of igneous rock in this part of the section were streams that flowed over the bed of the sea, or possibly on land sur- faces produced by its temporary elevation^ while the deposition of the sedimentary beds with which they are associated was going on, and were not injected among the latter subsequently to their formation, is clear on several grounds. They conform perfectly to the bedding of Remnants of Old Volcanoes. 365 the rocks above and beneath them, and nowhere cut in the slightest degree across the stratification. Their structure tells the same tale ; ad 'm i their upper surfaces are vesicular, showing that when they cooled the contained gases were free to bubble up and escape : this would not 366 Geology. have been the case if they had been covered at that time with over- lying rocks as they are now, and they must have hardened before the beds above them were deposited. As we descend into the interior and lower parts of each flow, where there was weight enough of lava overhead to check the ebullition, the cellular structure gradually dis- appears, and the bed becomes close-grained and com- pact, and more decidedly crystalline as we approach the centre. In fig. 117 the outcrops of the hard lava streams are seen to form steep craggy cliffs, ranging across the hill ; while the softer sediment- ary beds and ashes occupy the more gentle slopes below these and the hollows between them. The reasons for this will be given in Chapter XII. At some time after their deposition the beds just j, described were traversed by sheets of intrusive . Dolerite, marked a, b, c. The structure and lie of i these contrast forcibly with those of the contem- ' co poraneous flows. Their course conforms to a certain g degree with the direction of the bedding, and in g some places they have for a certain space been jg actually thrust in between the beds, so that a hasty "^ examination of a detached section might lead to the H o idea that they were interbedded and contempor- g aneous. But a more careful inspection soon dispels this notion ; if followed for a short distance they are ^ ^ found every now and then to eat their way up or 5 down into the strata above or below them. Their 2 intrusive character is also forcibly brought out when H r * the structure of the whole hill is viewed from a little g distance. On account of their hardness they crop out | in steep craggy cliffs, which on the north side form * oo Salisbury Crags and Heriot Mount ; these, as we trace | them southwards, instead of keeping the same dis- J C tance apart, like the interbedded traps, gradually draw together, and at last unite into the single mass of Samson's Ribs. This union could be brought about only by the sheets cutting obliquely across the bed- ding. That these rocks are intrusive is also shown by the baking of the beds both above and beneath them : the heat of a contemporaneous flow may alter the rocks underneath it ; but those above were not laid down till after it had cooled and they cannot therefore be affected by it. Lastly, in their structure there is a marked distinction between these and the contemporaneous lavas ; they do not possess a vesicular top, but are compact throughout, the cause of this difference evidently being that they consolidated under pres- sure sufficient to prevent the expansion and escape of their elastic vapour. The beds described so far were tilted from the horizontal position Remnants of Old Volcanoes. 367 in which they were laid down, and denuded,* and a hill produced whose outline must have been something like that shown in fig. 118, and here ends the history of the older portion of Arthur's Seat. There are however other rocks, shown in fig. 116 by a darker tint, which overlie those already described ; these are volcanic, and consist of a mass of very coarse volcanic Agglomerate (A), a central pipe of Basalt rising through the Agglomerate (), and a hummock of Basalt (C) known as the Lion's Haunch. These are the products of a later eruption. After the course of events which gave rise to the lower part of the hill had been completed, a period of repose followed, and then at a subsequent epoch volcanic action broke out afresh. An orifice w r as torn through the heart of the hill, and from it blocks of the rocks through which a way had been burst were showered out, and piled in a huddled manner round the vent. The Agglomerate is what remains of this accumulation. From the angular nature of its enormous blocks, the pell-mell way in which they lie, and the absence of any trace of bedding, it is clear that it was heaped up in the open air, and the contrast between it and the subaqueous tuff (4) in these respects is most marked and instructive. The Basalt (B} is lava, which boiled up in the chimney of the volcano ; it can be seen most distinctly running down through the Agglomerate in the form of a rudely-cylindrical column. At some time when there was not pressure enough to drive the lava over the lip of the crater a tunnel was torn open in the flank of the hill, and through it a flow of lava took place, a portion of which now forms the Basalt (C) of the Lion's Haunch. We have thus still preserved part of the cone of ejected materials and the central plug of this old volcano, but the portion remaining is probably only a small fragment of the original hill. Much of the loose Agglomerate has been carried away by denudation, and the part that is left has escaped mainly because the heat and heated vapours of the vent have baked and hardened it, and enabled it to hold out against atmospheric wear and tear better than the outer portions which underwent no alteration. In fig. 117 the conical central mound is formed of the Agglomerate (A), and at the summit the Basalt plug (B) is seen to rise out of the middle of it. The various members of the older group of rocks disappear beneath the newer formations, and come out again in a corresponding series on the south. On the ground the abrupt trun- cation of the bed (2), where it has been torn through by the later eruption, can be most distinctly made out.t * Owing to this denudation we have left only fragments of the old lava-flows, and cannot trace them up to the vent from which they issued. It is likely enough however that this lay where Edinburgh Castle now stands, and that the Castle-rock is a plug of lava which fills up the lower part of the orifice. t The explanation of the structure and history of Arthur's Seat given in the text is substantially that put forward by Maclaren, and adopted with some modifications by Professor A. Geikie. Its correctness has been called in question by Professor Judd (Quart. Journ. Geol. Soc. xxxi. 131). Anything coming from such an authority is entitled to the most respectful attention, but I cannot say that to my mind Professor Judd has succeeded in showing that the older inter- pretation is untenable, or even less probable than his own. Professor Judd's paper however should certainly be consulted by the student. 368 Geology. Ancient Volcanoes of North Wales. As an instance of volcanic products of still older date we may take the lava-flows, ashes, and intrusive rocks which make up so large a portion of the mountains of North Wales. We have evidence of active volcanoes at two dis- tinct epochs during the formation of the old rocks, known as Silurian, of that district. The earliest volcanic products consist of lava-flows and beds of ash interbedded with the sedimentary strata of the Arenigs, the Arans, and Cader Idris. These intercalations show in their upper part volcanic ashes and conglomerate, then a great mass of bedded lava streams, and beneath this a thick body of ash and volcanic conglome- rate : the whole reaches a maximum thickness of between 5000 and 6000 feet. The volcanic products of a somewhat later date have given rise to the rocks which now form the summit of Snowdon. They consist of the following members : Columnar Felstone ..... 200 Ashy Beds . . . . . .1200 Slaggy and Brecciated Porphyritic Felstones . . 1700 All these rocks furnish most undoubted proofs of a volcanic origin, and of their being of the same age as the sedimentary beds with which they are associated. The Felstones are sometimes vesicular and scoriaceous, and some- times show a decidedly slaggy structure, the lines of viscous flow being as apparent as in the cooled slags of a blast-furnace. They are also often most markedly columnar. They conform precisely in their bedding to the sedimentary rocks above and below them ; and when they produce any alteration in the associated strata, it is the beds underneath alone that are baked and hardened. Interbedded with the thick masses of Felstone are bands of sedimentary strata, showing that the former are not the results of a single flow, but that they were built up by a succession of streams poured out over the sea-bed, and that sediment was laid down in the same water in the intervals between their emission. The ashy strata, like the lavas, were accumulated beneath water. They contain a sufficient proportion of slaggy and scoriaceous frag- ments, and occasionally volcanic bombs, to leave no doubt on the mind that they are formed in large measure of volcanic ejections ; but these have in many cases been largely mixed up with sandy, clayey, or calcareous sediment, and in some cases they may have been formed of ashes dropped on land and afterwards carried into the sea by running water. Their submarine origin is further proved by inter- stratified beds of purely sedimentary origin, and by the occurrence in them of marine shells. In the case of each of the groups of volcanic rocks just mentioned we find them thickest round a certain centre, and as far as the remnants left of them will enable us to judge, thinning away in every direction from that centre till at last they disappear altogether. We thus get a clue to the quarters in which the vents lay from which the eruptions took place, but all traces of the piles of materials that must Petrology of Volcanic Rocks. 369 have once surrounded these vents have been swept away by denudation. There are however great intrusive masses of dioritic rock, which, from their distribution, seem to be connected in some way with both the groups of interbedded felstones and ashes, and these perhaps are the hardened contents of the reservoirs of molten matter from which the volcanoes were fed.* It is worthy of note however that while all the undoubtedly contemporaneous volcanic rocks, both molten and frag- mental, are felspathic, these deep-seated intrusive masses are markedly hornblendic ; and it seems strange that if the volcanic foci possessed so abundant a supply of hornblendic lava, no sheets of a hornblendic composition should ever have flowed from the craters. The explana- tion may be that, in accordance with Durocher's ideas, the heavier hornblendic matter sank to the bottom of the reservoir, while the lighter felspathic lavas floating above, came to be poured out in currents. Did space permit, these instances of beds of volcanic origin among the rocks of the earth's crust might be largely added to ; the examples given will however suffice as illustrations, and the reader may refer for an exhaustive sketch of the volcanic productions of different ages in Britain to Professor Geikie's Address to the Geological Section of the British Association at the Dundee Meeting of 1867. He should also make an attentive study of Professor Judd's restoration of the old volcanic area of the Western Isles of Scotland (Quarterly Journal of the Geol. Soc. xxx. 220) ; of Mr. Ward's description of the volcanic rocks of the Lake district (ibid. xxxi. 388) ; and of Professor A. Geikie's account of the volcanic rocks of the basin of the Firth of Forth (Trans. Royal Soc. Edinburgh, xxix. (1879) 437). SECTION V. PETROLOGY OF VOLCANIC BOCKS. The comparison which has been now made between the phenomena and products of modern volcanoes and certain members of the Crystal- line class of rocks will, it is hoped, have conclusively demonstrated that the latter must have owed their origin to a volcanic source. Our next task will be to look at these rocks on a large scale in the field, to study the different shapes and forms under which they present them- selves in mass, to examine their relations to the beds by which they are surrounded, and to inquire how far such observations will guide us to a knowledge of the circumstances under which they were formed. Distinction into Intrusive and Contemporaneous. The molten products of volcanic action are in the first instance forced up through orifices torn through the surface of the ground, or are forcibly injected into cracks and rents. When the contents either of an orifice or a crack cool and harden, bodies of Crystalline rock will be formed ; these will evidently assume different external shapes in the two cases, but they will both agree in one point they will both give * For details respecting the Igneous Rocks of North Wales see Professor Ramsay's Geology of North Wales, Memoirs of the Geological Survey of England and Wales, vol. iii. 2 A 3/0 Geology. proof of having burst through the rocks which surround them ; hence they are classed together as Intrusive. The columns of cooled lava which fill up an old volcanic chimney are known as Necks; the hardened contents of cracks form Dykes or Veins, the term Dyke * being restricted to those fissures which are approximately rectilinear and vertical, while the more irregular tortuous or branching rents give rise to Veins. Where the fissure into which lava has been driven is approximately horizontal and of con- siderable breadth, so that the hardened stuff forms a flat tabular mass, we have an Intrusive Sheet. It is clear that any mass of intrusive rock must be of later date than the rocks it traverses, hence intrusive volcanic rocks are also spoken of as Subsequent. But the molten matters poured out from volcanoes also present themselves to us under another form. Great sheets of lava flow from the vent and spread themselves over the surrounding district. These stream over the pre-existing rocks instead of bursting through them, and they therefore stand somewhat in the same relation to the rocks beneath them as an overlying stratum of sedimentary material, and are in this respect clearly marked off from rocks of an intrusive character. Lava sheets poured out on dry land may be lowered beneath bodies of water and sedimentary rocks laid down above them ; or it may happen that, after a mass of derivative rocks has accumulated beneath water, a subaqueous stream of lava may flow over them, and be in turn itself covered by further depositions of sediment. In these ways there will arise formations consisting of alternations of beds of molten and sedimentary origin, the rocks of both classes being truly interstratified and perfectly conformable in their bedding to one another. An instance of a formation of this sort is furnished by the older rocks of Arthur's Seat, shown in the section on fig. 118. Sheets of volcanic rock formed under these circumstances are called Interbedded, or, because they were formed at the same time as the sedimentary beds among which they are found, Contemporaneous. The terms Eruptive and Irruptive may be also used for the two cases. An eruptive rock has been forced up to the surface and has flowed over it ; an irruptive rock has been forced up through other rocks but has not succeeded in reaching the surface. Now in studying any mass of volcanic rock we may come across in our investigations, the first thing to be made out is, to which of these two classes does it belong. Is it Intrusive or Contemporaneous ? In many cases there is not room for a moment's hesitation on this point : a cylindrical pipe drilled through bedded sedimentary strata, or a rent torn across them at right angles to the bedding and filled up with hardened lava, furnish conclusive proof that the latter is intrusive. And at first sight it might seem equally safe to pronounce that a lava sheet, lying between two masses of sedimentary strata with its upper * " Dyke " means in north country language a wall. The contents of igneous dykes are usually harder than the rocks they traverse ; hence the latter are more easily and largely removed by denudation than the former, and a rib of the more durable rock is left standing up, running like a wall across the country. Petrology of Volcanic Rocks. 371 and under surfaces parallel to the bedding of the rocks above and beneath it, was undoubtedly contemporaneous. But such a conclusion would not necessarily hold good, because there are cases when lavas have been forcibly intruded between the planes of bedding of pre-existing sedimentary rocks, and where they have confined themselves so strictly and for such long distances to the space between two consecutive beds, that a delusive appearance of true interbedding is produced. In the case therefore of a lava sheet much care is often needed before we can decide whether it is intrusive or contemporaneous ; some of the tests which decide the question have already been hinted at in the description of Arthur's Seat, and the subject will be more fully treated of further on. Alteration of Neighbouring Rocks. One fact which will aid us in clearing up such doubtful cases, and which we shall also have occasion to refer to in connection with other branches of our subject, is the alteration wrought by the intense heat of lavas in the rocks they come in contact with. Soft Clays are baked into hard, flinty, porcelain-like rock ; Sandstones are hardened and rendered crystalline ; Limestones are turned into Marble ; Coal is converted into cinder or soot. Owing to the low conducting power of rocks, such changes seldom extend to any great distance from the margin of a mass of lava ; and sometimes the rocks immediately in contact with the once molten matter are perfectly unaltered. The absence of alteration is easily explained, if we recollect that the chilling effect of the surrounding rock would rapidly cool the layer of fused matter immediately in contact with it, and a crust of low conducting power would thus be formed which would separate the still liquid central portion from the rocks on either side and prevent the free passage of heat from one to the other. Included Fragments. Intrusive rocks too frequently tear off and carry along with them fragments of the rocks through which they have forced their way ; such detached portions vary in size from mere stones up to large masses. They are usually baked and hardened. Fig. 119 shows a case of this sort seen in the castle rock at Dunbar. Fig. 119. SECTION AT DUNBAR CASTLE, SHOWING AN INCLUDED MASS OF ALTERED SANDSTONE IN DoLERITE. A mass of Dolerite (1) has burst through Sandstones and Shales (2) ; the former are altered along the line of junction into a hard crystalline Quartzite (3), and in the very middle of the Dolerite is a wedge-shaped mass of similar Quartzite (4), evidently a large fragment of the Sand- stone through which the intrusive rock was ejected, caught up by the liquid lava, baked by its heat, and retained in its midst when it cooled. 372 Geology. It sometimes happens that subaerial lava streams contain included blocks. If they flow over ground covered with loose fragments, some of these may be caught up and embedded in them ; or blocks of rock may have been thrown out of a volcanic vent, fallen on the surface of a lava while it was yet soft, and sunk down into the body of the rock. Fragmental Interbedded Rocks. The fragmental products of volcanic action give rise to rocks which are necessarily for the most part of an interbedded character; deposits namely of stones, cinders, and lapilli shot into water and arranged in layers alternating with beds of purely sedimentary origin, or strata formed of a mixture of volcanic ash and sediment. Necks of Agglomerate. But it sometimes happens that the chimney of a volcanic vent has been filled in, not with hardened lava, but with a confused mass of volcanic ejections which have fallen back into it after they were hurled into the air. We thus get Necks of Volcanic Agglomerate, where the latter stands to the surrounding rocks in the same relation as a plug of intrusive molten rock. Such masses, though they can hardly perhaps be called intrusive in the proper sense of the word, are clearly distinguishable from the truly interbedded deposits of volcanic ejections. Instances of the Modes of Occurrence of Volcanic Rocks. We will now consider a little more in detail the different forms under which we have seen that large bodies of volcanic rock present themselves. These may be classified under the following heads : Dykes and Veins, Necks, and Sheets. Dykes and Veins. It has been already mentioned that modern volcanic cones are frequently torn across by rents approximately vertical, which are afterwards filled in with lava ; ribs of igneous rock, exactly similar in character, are constantly met with cutting across the older rocks of the earth's crust. Sometimes these can actually be seen to be offshoots from a large igneous mass ; this is the case in fig. 123, where many dykes are seen running out from the main body of the trap into the surrounding country. More frequently however the connection of these older dykes with the parent mass can no longer be traced ; they have been either severed from it by denudation, or they descend deeper into the earth than we can follow them before they reach it. Dykes are found of all dimensions, both in length and breadth ; they vary in thickness from less than a foot upwards ; and while in Fig. 120. DYKES CUTTING THROUGH SHALES AND LIMESTONES, WEST OF DUNBAK. some cases the portions of them that appear at the surface do not extend longitudinally for more than a few yards, in others they can be traced continuously for miles. Petrology of Volcanic Rocks. 373 The section on fig. 123 shows dykes with remarkably true and even walls ; fig. 120 is a sketch of a group of dykes (G) more irregular in outline, cutting through Shales, Sandstones, and impure Limestones, and sending tongues out into them. f Necks. A good instance of a plug of hardened lava ; filling up an old volcano vent has been already given in the section of Arthur's Seat in fig. 116. In this case the neck is still surrounded by part of the cone which it traversed ; more frequently however the softer and friable surrounding of f ragmental matter has been carried away by denudation, and the central column of hard lava, or the lower portion of it, alone remains to fix the position of the orifice. These remnants of the plugs that once filled volcanic vents are very common among the old volcanic rocks of the central valley of Scotland. Some good instances may be seen on the Berwickshire coast, where isolated bosses of Felstone stand up in low hills above the softer rocks of the surrounding country.* When these are examined they are found to be rudely cylindrical columns which descend vertically through the surrounding strata. There are also in the same dis- trict lofty conical-shaped hills Traprain Law and North Berwick Law for instance composed of a similar rock. We cannot prove by actual inspection in these latter cases that the eminences are the summits of great cylinders that pierce through the beds around them ; but such a supposition is highly probable ; and we may very fairly look upon these prominent peaks as probably portions of the column of lava that rose through a volcanic orifice, their size being due partly to the fact that the volcano was a big one, and partly to their having formed a portion of the column where the orifice was roomy. These necks usually project above the surrounding rocks for the same reason that dykes stand up in walls, because they are the harder of the two. But we have already mentioned that a volcanic vent is sometimes filled in, not with hardened lava, but with a mass of Volcanic Agglomerate. A capital instance in Ayrshire, described by Professor A. Geikie,t is illustrated by fig. 121. The section runs across a basin consisting of the following rocks. At the bottom are Shales and Sand- stones (a) 'j upon these rest some beds (b) which consist of irregular alternations of volcanic ash and brick-red Sandstone. Then come sheets of lava (c), consisting of Labradorite and Specular Iron Ore, with sometimes a * The Geology of East Lothian (Memoirs of the Geological Survey of Scotland), p. 40. t Geol. Mag. iii. 243 ; Memoirs of the Geological Survey of Scotland, Explanation of Sheet 13, paragraph 5, and of Sheet 14, paragraph 26. 374 Geology. little Augite. These are covered by beds (d) similar in composition to (b). Above all there is a group of brick-red Sandstone (e), containing in its lower part occasional nests of volcanic lapilli and single stones. From this section we learn that the district was free from volcanic activity during the deposition of the beds (a). The next beds (b) show the commencement of eruptions ; they were formed partly out of sandy sediment, and partly by showers of volcanic dust and lapilli which fell into the water where the silt accumulated. Then came the emission of the lavas (c). These present all the peculiarities which we have described as characteristic of contemporaneous flows : they are divided into beds with slaggy or cindery upper and under surfaces, and some- times parted from one another by layers of brick-red Sandstone. Running across these lavas in a direction perpendicular to the stratifi- cation there are what look like veins of horizontally-bedded red Sand- stone, which sometimes radiate from a centre forming star-shaped figures. These were doubtless produced in the following manner. As the lava cooled cracks opened in its surface, and sand was washed into the cracks and filled them up before they were covered by the next flow. The absence of lava sheets in the beds (d) shows that volcanic activity was on the decrease during their formation, and its final cessation is indicated by the gradual disappearance of volcanic products as we pass from the lower to the upper part of the group (e). In the country occupied by the beds (a) there are a number of small rounded hills or hillocks, two of which are shown at (/) and (g). These consist of a very coarse red volcanic Agglomerate, unstratified and tumultuous in appearance, made up of fragments of lava similar to (c), of all sizes up to masses a yard or more in length, angular, subangular, and rounded, embedded in a gritty, felspathic, ferruginous paste. These hillocks rise conspicuously above the neighbouring ground, and might at first sight be taken for the remnants of a deposit which once extended over the beds (a), the greater part of which has been removed by denudation ; but such is not the case. Though surrounded by the Shales and Sandstones, they do not lie on the latter ; on the contrary they descend vertically through them, like so many huge pipes. In short they are true volcanic necks, each repre- senting a former focus of eruption. In one case the sides of a neck were found coated with a lining of rough slaggy lava, which had clung to them during the discharge of a stream of molten matter. The close neighbourhood of these vents to the volcanic rocks of the section, and the similarity between the blocks they contain and the lavas (c), makes it almost certain that it was from these orifices that the volcanic products of the adjoining rocks were discharged. For other instances of necks of volcanic Agglomerate see the Geology of East Lothian (Memoirs of the Geological Survey of Scotland), p. 44. Sheets. The only remaining form under which igneous rocks occur is that of sheets, and these are, as has been already mentioned, of two kinds, intrusive and contemporaneous. Intrusive sheets, like dykes, are merely hardened masses of lava, which were forcibly injected in a molten state into rents traversing pre-existing rocks ; but while we restrict the term dyke to those cases Petrology of Volcanic Rocks. 375 where the fissure is vertical or cuts across the bedding at large angles, we use the word sheet where the cleft is inclined at small angles to the bedding, or where the actual space between two beds has furnished a channel for the molten matter. The difference in appearance between the two forms is quite enough to justify us in using the two names ; and it is besides desirable to keep up the distinction, because, whereas no one could suppose a dyke to be anything but intrusive, those sheets which have been thrust in along the planes of bedding are liable to be mistaken for contemporaneous flows. A contemporaneous sheet is formed by the flow of a lava stream over the bottom of a body of water on which sedimentary deposits have previously been laid down, and by the subsequent formation on the top of the stream, after it has cooled and hardened, of other accumulations of sediment. Where an intrusive sheet cuts across bedding planes, its true char- acter is at once recognised ; but where a sheet occupies the space between two beds, it requires some care to determine to which class it belongs. The tests to be applied in such a case have been some of them already mentioned and may be summarized as follows. 1. Alteration of the adjoining rocks. A contemporaneous sheet can produce alteration only in the beds below it, because those above were not laid down till after it had cooled. An intrusive sheet can alter the beds above as well as those below it. We must recollect how- ever, in applying this test, that the absence of alteration proves nothing, because igneous rocks sometimes produce no change in the beds they come in contact with. If the beds above as well as those below an igneous sheet are altered, we are sure it is intrusive ; if those below only are affected, it yields a presumption that it is contemporaneous, but other tests must be applied before we can be certain it belongs to this class. 2. Texture. The rocks which now overlie a contemporaneous sheet wei-e not there at the time when it was poured out, and there was no pressure on it from above at the time when it cooled. It will therefore have all the characteristic traits of a subaerial lava-flow. The top will be vesicular or amygdaloidal, the vesicles being all pulled out in the same direction ; and it may show slaggy or ropy texture. The middle portion will be compact and perhaps towards the centre distinctly crystalline. The base may be cindery. But when an intrusive sheet was injected, the lava was weighted above by over- lying rock, and the pressure so produced prevented the contained steam from boiling up and producing vesicular structure; the upper and under surfaces may, on account of their more rapid cooling, be finely grained, while the slower hardening of the interior may have allowed of the formation of more largely crystalline products ; but though for these reasons different parts of it may differ in grain, the whole of it will be dense and compact; it will rarely show slaggy portions; and if it be vesicular at all, the cavities will not usually be either numerous or large. We do occasionally find vesicular structure on the top of an intru- sive sheet.* It seems to have been produced where a crack ran * I have noticed vesicles move than an inch long on the top of an intrusive sheet in Glen Eisdale in the island of Arran. 376 Geology. through the overlying rocks up to the surface and allowed the escape of steam from the portion of the lava in its own immediate neighbour- hood. But even in such a case there is not much risk of our being misled as to the character of the sheet. If it be intrusive, any vesicu- lar structure that may be present occurs in detached patches here and there, and is obviously exceptional. In a contemporaneous sheet the whole top is vesicular. 3. Relations to the bedding of rocks above and below. A con- temporaneous sheet evidently cannot cut across or eat its way into the beds above it, because they were not there when it was poured out. It is possible that, in its passage over the underlying strata, it might disarrange and perhaps work down into them, but this could take place only to a very small extent. We may say of such sheets that their upper and under surfaces are everyivhere parallel to the bed- ding of the rocks above and beneath them. This will not be found to be the case with intrusive sheets ; they have sometimes pursued their 'path with wonderful constancy along planes of bedding, so much so that the examination of a limited exposure might lead to the belief that they were interbedded ; but if followed out they will invariably be found sooner or later to cut up or cut down into the rocks above or beneath, or to send out tongues or spurts into them, and some- where to break across the stratification in a way that at once reveals their true character. 4. The presence of included blocks may sometimes decide whether a sheet is intrusive or contemporaneous ; if these have come from the overlying rocks, the sheet must belong to the former class. The third test is perhaps the one of most universal application. Texture alone is not an absolutely certain criterion, and alteration of the adjoining rocks may be wanting; but an intrusive sheet will seldom fail to show its real nature somewhere or other by pursuing a course transgressive to the bedding of the rocks among which it occurs. In the account of Arthur's Seat a few pages back, the application of these tests to a particular instance was pointed out, and fig. 122 is a section on the coast west of Dun- bar which further illustrates the subject. Let us confine our atten- tion first to the left-hand half of the figure. We have there a sheet of Dolerite shown by the dark colour, apparently interbedded with a group of Shales, whose bedding is shown by the fine parallel lines. The upper and under surfaces of this sheet are so far parallel to the bed- ding of the Shales, that if we saw Fig. 122. DYKE AND INTRUSIVE SHEET OF DOLERITE. only the left-hand side of the section, it would not be apparent at first sight to which class it belongs. Let us apply the tests just given to determine its true character. In the first place the beds both above and below are markedly altered ; the Shale, which a few feet off the I HE Petrology of Volcanic Rocks. (f IT N I VsfiR S I V /v. Crystalline rock is so soft and clayey that it yields ros^jjy tPl$& \ \S impress of the finger, is baked into a hard, flinty, felsitic rock, wMch will scratch glass and has here and there a slightly crystalline tex- ture. Hand specimens of this altered Shale could not be distinguished by the unaided eye from a compact Felstone. Next, in spite of its rough conformity to the bedding, we find, when the junctions are closely scanned, that the rock has thrust out bosses and strings, giving it rough and irregular boundaries very different from the even planes of bedding of the Shale. The floor of the stream too gives unmis- takable proof of rough usage ; the beds of Shale have been puckered and crumpled, and the uppermost of them broken up into small bits which are tilted on end against one another, so that the line of junction has a jagged, saw-like edge. By all these signs then we are sure that the sheet is intrusive, and our conclusion is confirmed when we trace it towards the right, for we then find it springing out from a dyke which cuts vertically across the bedding. This dyke has caused great alteration of the rocks on either side ; they are baked into a Felsite, such as lies above and below the sheet, and their bedding becomes gradually effaced as they approach the Dolerite and replaced by a rude jointed or platy struc- ture with divisional planes parallel to the walls of the dyke. The above is a happy case, where we find all the tests which dis- tinguish an intrusive from an interbedded stream exhibited. The observer must not expect to be always equally lucky, some one or more of the marks of distinction are often absent ; the volcanic rock, for instance, may not have given rise to any alteration of the beds it traverses, and w r e may thus miss one of the easiest ways of determining its nature. He must then fall back on another test ; and there are very few cases indeed in which, when all the circumstances have been taken into account, any uncertainty will remain. Circumstances under which Intrusive Sheets "were injected. The injection of a sheet of lava of the intrusive class involves the lifting up of the whole mass of the overlying rock through a space equal to the thickness of the sheet. Intrusive sheets some- times extend over areas of hundreds of square miles in extent and reach considerable thicknesses, and if at the time they were driven in there was any considerable thickness of rock above, we may well hesitate before we allow that even the mighty energy of volcanic forces would be equal to such a feat. Where we find vesicular structure in patches on the top of a sheet manifestly intrusive, it seems certain that there must have been fissures at the time of its injection running through the overlying rocks up to the surface, and that in these cases therefore the overlying rocks could not at that time have been very thick. Likely enough what is true in the case of these sheets is true of all ; they were injected after the deposition of the beds next above them, but not very long after, so that no large thickness of sediment had been laid down above their level. Where we find intrusive sheets we almost always find contemporaneous lava- flows in the same rock group ; the two probably came from the same reservoir and were formed during the same period of volcanic activity. 37 s Geology. In the one case the fissure which gave passage to the lava upwards reached the surface ; in the other case the lava having travelled verti- cally upwards for some distance met with a check to its further progress in that direction and spread itself out horizontally, a plane of bedding furnishing a line of weakness along which it could most easily insinuate itself. SECTION VI. LITHOLOGICAL VARIETIES OF VOLCANIC ROCKS. In the preceding chapter the Crystalline rocks were arranged in classes according to their mineral composition, and all rocks which had the same mineral composition were called by the same name. Such a classification has its advantages ; but manifestly it cannot be looked upon as final, for it altogether ignores the one item, which from a geological point of view is of paramount importance, viz. the way in which a rock has been formed. And the nomen- clature used in this classification is not unlikely to lead the beginner astray. He finds a group of rocks all called Dolerites ; he forgets that they are placed together in the same class simply and solely because they have all the same mineral composition, and he is apt to jump to the conclusion that they were all formed in the same way, and generally that they agree in every particular. This is not so ; many, perhaps most Dolerites are Lavas ; but it does not necessarily follow that all are. It would be better, if we found a rock which was a Dolerite in composition and was also a Lava, to call it a Doleritic Lava ; but this nomenclature has not yet come into general use. Retaining then the nomenclature in vogue, we will point out which of the classes of Crystalline rocks established and named in the last chapter include members that belong to the Volcanic subdivision. None of the Granites and Syenites, and I believe I may add none of the Mica-traps and Diorites are volcanic. The majority of the Elvanites are riot volcanic, but some may be de vitrified Rhyolites. Some Quartz-trachytes are volcanic. Of the following classes the majority of the members, in some cases all, occur as Lavas, or in dykes or necks intimately connected with Lava : Rhyolites, Pitchstones, Obsidians, Pumices, Phonolites, Trachytes, Andesites, Melaphyrs, Porphyrites, Basalts, Dolerites, Tachylites, Tephrites, Nephelinites, Leucitites, Pikrites, and Limburgites. Identity of modern and old Volcanic Rocks. Before concluding our sketch of this class of rocks, we would again call attention to the perfect agreement in every essential respect between the volcanic products of all the periods of the earth's lifetime of which any record has come down to us. The old fashion was to confine the term " Volcanic " to these rocks that had been formed during the present and the later geological epochs. The rocks which bore considerable resemblance to these but belonged to more remote geolo- gical epochs it was the habit to denote by some different name, such as Trappean, on account of certain differences from the volcanic Lithological Varieties of Volcanic Rocks. 379 products of more recent date which they were supposed to present. The growth of our knowledge is tending more and more every day to upset such a notion, but the ill-effects of it are still felt in our nomen- clature. As was pointed out several times in the last chapter, two rocks of absolutely identical composition still go by different names because one happens to be of later date than the other. The following are some of the old-fashioned notions on this subject. It was stated that there are certain mineralogical differences between the two classes. The Potash Felspar of the Trappean rocks was said to be Orthoclase, that of the Volcanic rocks Sanidine. If true, a distinction almost without a difference ; and it is doubtful if it be universally true. I have a slice of one of the very oldest known Volcanic rocks in Britain in which is a crystal of Orthoclastic Felspar very much decomposed, but still showing internal structure closely resembling that which characterizes some Sanidines. Hornblende was supposed to be specially characteristic of Trappean, Augite of Volcanic rocks of the Basic class. Microscopic examination has shown this to be very wide of the truth. A speculation has also been put forth that the older Igneous rocks are mainly Acidic and the newer mainly Basic in composition.* It is difficult to resist the suspicion that the wish has here been father to the thought. If certain theoretical views are correct, it ought to be so ; but facts are directly in the teeth of it.t It is mentioned here mainly because geologists of high repute have been found to counte- nance it. Some authors hold a similar opinion, but under a modified form. Thus Gotta says, "We are almost justified in holding it for a universal law that whenever Igneous rocks rich in Silica occur together with Basic Igneous rocks of the same period of eruption, the latter are of somewhat later origin than the former.'^ This notion has an ^ priori probability in its favour, for it is likely that the lighter Acidic pro- ducts would be discharged before the heavier Basic, and cases may doubtless be found where the rule is true ; but there are also exceptions enough to it to prevent our accepting it as a universal law. (See Scrope's Volcanoes, pp. 125 and 347.) Other attempts have been made by chemists and petrographers to discover a connection between the mineral character of an igneous rock and the date of its formation, but the rules laid down have either been too vague to be of practical value, or they have broken down when subjected to the test of geological examination in the field. Perhaps the nearest approach to success in this line has been made by v. Richthofen, who found a certain succession among the Volcanic rocks of Hungary and Transylvania. A very similar succession has been observed to obtain among rocks of the same age in the Western Territories of the United States. The latest eruptions in both * Duroclier's Essay on Comparative Petrology. There is a translation in Professor Haughton's Manual of Geology. t See Scrope's Volcanoes, p 128 ; Allport in Geol. Mag. vol. x. 196. Rocks classified and described (English translation), p. 187. Geological Exploration of the 40th Parallel, vi. 9-11. Geology, cases gave rise to Basaltic Lavas, but there is nothing to distinguish these Basalts from Basalts of far earlier date. The oldest emissions formed rocks which the author calls Propylite ; but his definition of this rock is excessively vague, and it is quite impossible to say wherein Propylite differs from Hornblende-andesite, so that this part of the generalization seems to be of doubtful value. Much confusion has been introduced into this part of the subject from not distinguishing between these products of igneous action which are truly volcanic, namely those which have cooled and hardened on the surface or not far below it, and those Igneous rocks which will be treated of in the next chapter under the name of Plutonic, that have consolidated below the surface. It has been asserted that the older Igneous rocks are mainly Plutonic, and those of later date mainly Volcanic in their character. This is true enough, but it by no means follows from it that the igneous action of the earlier geological epochs differed in any respect from that of the present day. Why this is so is not far to see. That Plutonic or deep-seated rocks should be more numerous among the older deposits, and that Volcanic or subaerial rocks should prevail among those of younger date, is only what is to be expected ; for subaerial products of any antiquity could only by a lucky chance escape the destructive action of denudation, and must therefore be rarer than those deep- seated products which have been better protected from its action ; while in the case of the newer formations denudation has not yet had time to carry away the lavas, ashes, cones, and other external products of volcanic action, and, by the removal of these and the masses of rock underlying them, to lay bare the formations that have hardened far down in the bowels of a volcanic area. In short, the generalization amounts to no more than this : the older Igneous rocks are mainly deep-seated, because denudation has largely carried away the subaerial igneous formations of distant epochs ; newer Igneous rocks are mainly subaerial because denudation has not yet worked its way down to the deep-seated formations of recent periods. The effect of alteration in modifying old Volcanic rocks and pro- ducing a difference between them and similar rocks of a more recent date has already been noticed. It is a factor whose importance has been frequently overlooked. It is to the late Mr. Poulett Scrope that we are indebted for a true explanation of the mechanism of volcanic action, and the student should study with the utmost attention his works* on the subject. Professor Judd's more recent work, " Volcanoes " (International Scien- tific Series, vol. xxxv.), gives much additional information. The reader may also usefully consult a paper by Professor A. Geikie " On the Carboniferous Volcanic Rocks of the Firth of Forth," Trans. Royal Soc. Edinburgh, xxix. 437 (1879); and the Geology of the Northern Part of the English Lake District (Memoirs of the Geological Survey), chaps, iv. and v. * Volcanoes, the Character of their Phenomena, etc.; Longmans, 1862. The Geology of the Extinct Volcanoes of Central France ; Murray, 1858. Mud Volcanoes, 38 SECTION VII. MUD VOLCANOES. SALSES. Certain volcanic outbursts are accompanied by discharges of mud, and all the volcanic hills which show this behaviour are usually classed together as Mud Volcanoes. But it seems likely that two very distinct classes of eruptions are included under this head. Both agree in pouring out bodies or streams of mud ; but the accompanying gases and other products, and probably to a certain extent the cause which produces the eruption, are different in the two cases. One class of Mud Volcanoes is found in districts where volcanic action is still going on and in the immediate neighbourhood of active volcanoes. They are often ranged along straight lines like volcanic vents. The gases which they give off are those which are usually met with in Solfataras, and there is nothing except the muddy nature of their products to distinguish them from ordinary volcanic fumaroles. They are merely small vents which happen to be situated in neighbour- hoods where clay, or volcanic products easily reduced to a clayey state, abound. The clay is rendered plastic and is forced out by the ebul- lition of steam and gas. Mud Volcanoes of this class occur in Iceland, Celebes, Luzon, New Zealand, and other volcanic districts. The other class of Mud Volcanoes are conical-shaped mounds made up of dark-coloured, dried and hardened mud, with a funnel-shaped or basin-shaped orifice or crater at the summit. They are usually of no great height; often rising only a few feet above the surrounding ground. Macaluba in Sicily is 150 feet high : some Mud Volcanoes in the neighbourhood of the Caspian range from 100 to 150 feet: Agh Sibyr in that district is said to reach 500 feet. In the bottom of the crater are small openings from which gases and vapours escape. The crater usually contains water which reduces some of the mud of the cone to a plastic or half-fluid state. This softened mud is spurted up by the outbursts of gas, and occasionally when the discharge is large and the mud sticky enough to impede its free exit, streams of mud are forced over the edge of the crater and flow down the slopes of the cone and add to its size. Sometimes the streams of mud are large enough to spread over the country round. The water discharged is often salt, hence the name Salses. The main points in which the eruptions of these volcanoes differ from ordinary volcanic outbursts are these. The temperature of the vent and of the substances emitted is very little higher than that of the surrounding air, or it rises above this point only for a short time during the actual eruption. The chief products too are markedly different from those of ordinary volcanic outbursts. The gas given off most largely is Carburetted Hydrogen, and steam appears in quantity only during those short periods of high temperature which accompany eruption. Petroleum, Naphtha, Asphalt, or some Bituminous* sub- * A number of natural inflammable pitchy or oily substances are included under the general term Bitumen. They consist of various hydrocarbons with variable quantities of oxygen and nitrogen. The solid varieties go by the general name of Asphalt or Mineral Pitch. The liquid forms are called Naphtha when 382 Geology. stance are constant and most characteristic products. Sufficient heat is sometimes generated to fire the Carburetted Hydrogen, and then flames are emitted. The following are some instances of this class of Mud Volcanoes. In the island of Trinidad there is a small group of Mud Volcanoes, the discharges from which always contain Petroleum. Mr. Wall has shown that the substratum of the district is clay containing beds of lignitic matter, and that the lignite is now being converted into Petroleum or Asphalt : during the change Carburetted Hydrogen and Carbon Dioxide are given off. He therefore refers the eruptive phe- nomena solely to chemical changes which are going on at small depths below the surface.* In the much larger Mud Volcanoes of Tanan (West Caucasus) and at Baku (East Caucasus) the evolution of inflammable gas and naphtha is so large that Eichwald says they ought to be called Naphtha-vol- canoes. Macaluba in Sicily stands upon Blue Clay containing Gypsum, Salt, Sulphur, Sulphides of Iron and Copper, and Sulphates of earthy bases, interstratified with Marls and Limestones. It is a truncated cone with a number of smaller cones perched on its flat summit : each of these has at the top a funnel-shaped cavity filled with a mixture, a little above the temperature of the air, of water, mud, and bitumen. Gas is given off, 96 to 99 per cent, of which is Carburetted Hydrogen, the remainder being Carbon Dioxide. During eruptions the mud is thrown up to a height of 200 feet and there is a strong smell of sulphur. Dr. Daubeny was inclined to refer these phenomena to the slow com- bustion of beds of sulphur ; the heat thus produced would decompose any organic matter in the adjoining rock and give rise to Petroleum and Carburetted Hydrogen ; Sulphuric Acid might also be formed, and acting on the Limestone might set free Carbon Dioxide, t It seems then that there is a valid distinction between ordinary volcanoes and such Mud Volcanoes as have been just described, and there also seem to be good grounds for believing that the eruptions of these Mud Volcanoes are produced by the escape of elastic gases, mainly Carburetted Hydrogen, which are generated by the chemical decomposition of organic matter at no great depth below the surface. The further question arises, Where does the heat necessary for produc- ing this decomposition come from 1 Now all the districts in which Mud Volcanoes occur are either in the neighbourhood of active volcanoes or have been once the scene of volcanic action. In the first case the answer to our question is obvious ; in the second it seems likely that there may yet be enough heat remaining to set going the chemical decompositions which are the direct causes of the outbreaks. This view is further supported by the fact that even the second class of Mud Volcanoes occasionally break out into eruptions differing only they are thin and slightly coloured : Maltha or Mineral Tar when they are very viscid ; and Petroleum when they are intermediate between these extremes. Other names applied to them are Stcinol, Erdol, Bitume liquide. * Report on the Geology of Trinidad (Memoirs of the Geol. Survey), p. 149. t Volcanoes, p. 266. EartJiquakes. 383 in degree from those of true volcanoes. They are preceded and accompanied by slight earthquakes, and there is considerable elevation of temperature. Steam and Sulphuretted Hydrogen are the gases most largely given off, and there is scarcely any Carburetted Hydrogen. It would seem then that all Mud Volcanoes owe their activity ulti- mately to deep-seated volcanic heat. Some are merely miniature volcanoes of the ordinary type situated in clayey ground. In the case of others the heat is too small in amount to generate high-pressure steam and produce the terrific manifestations which accompany ordinary volcanic eruptions, but indirectly it causes outbreaks of a less violent kind by promoting chemical decomposition and generating elastic gases and inflammable compounds. In the one case volcanic heat is the direct, in the other an indirect cause of the eruptions. I am not acquainted with any rocks in the earth's crust which can be certainly said to be the products of old Mud Volcanoes ; but that some of the clayey deposits associated with Volcanic rocks were formed in this way, is not unlikely. SECTION VIII. EARTHQUAKES. Though Earthquakes are not agents of so much geological importance as was at one time supposed, they are so intimately connected with volcanic action, that this chapter would be incomplete without some notice of them. How Earthquakes are caused. The following illustration will enable us to realize how earthquakes and the phenomena which accompany them are brought about. Suppose that we have a large steel armour-plate firmly supported in a horizontal position, and held so fast that it is impossible that it can move as a whole, and that a number of vases or similar objects are placed upon it. Suppose that a heavy blow is struck with a hammer in a vertical direction on the under side of the plate. The vases will be upset. How is it that a blow delivered on the under side will thus affect objects standing on the surface of the plate 1 It might be supposed that the plate was lifted bodily through a small space, and that it is in this way that the vases became jerked up and overturned. But the plate is too firmly fixed to allow of this happening, and it is possible to discover another reason why this cannot be the true explanation. If the overthrow of the vases was owing to the plate being lifted bodily upwards, they would all fall at the same instant. They do come down very nearly all together, but if the plate be large enough and the times of their falling be noted by instruments capable of recording very small inter- vals of time, it can be shown that they do not fall quite simultaneously. Those immediately over the spot struck fall first ; the others come down a little later, and those most distant from that spot are the last to fall. What upsets the vases is obviously not a motion of the plate as a whole, but something transmitted from the point struck through the plate which travels with very great rapidity but does take time to pass from one spot to another. 384 Geology. Let us now inquire how the effect of the blow is enabled to pass through the plate. We must first recollect that the plate is made up of a number of very minute particles called molecules and that these molecules are not touching one another but that there are spaces between them. These molecules are not absolutely at rest, but they move in such a way that for each of them there is a point called its mean position from which it never departs far. For our present purpose we may suppose that in the normal state of the plate each molecule is at rest in its mean position. Consider first the chain of molecules lying in a vertical straight line through the point struck by the hammer. The effect of the blow will be to force a number of these molecules nearer together than they are in their normal position. Now if the blow had been delivered on a mass of putty, very little more than a displacement of the molecules would be produced : some of them would be squeezed out around the base of the hammer, and the molecules thus displaced would remain in the positions into which they had been forced. But steel possesses a property known as elasticity in virtue of which any portion that has had its density or shape altered, tends to return to its original density or shape as soon as the force which produced the alteration ceases to act. When then the momentary effect of the blow has passed away, the molecules which were displaced by it begin to move back to their old mean positions. Each molecule on reaching that point is moving with a certain velocity : it will therefore be carried below it, and, if the elasticity were perfect, it would not stop till it had reached a point as far below its mean position as the point to which it had been driven by the blow was above it. It would then commence to move upwards, and would continually swing backwards and forwards between two points equidistant from its mean position. But no body has perfect elasticity ; each excursion which the particle makes will be shorter than the previous one, and it will at last come to rest in its original mean position. The body of molecules then set in motion by the blow will after a time settle down into their original positions ; the time during which they are in motion will be short. But the displaced molecules will act on a group of molecules just above them ; these will be set in motion, and after a short time will come to rest. The displacement of the second set will in its turn set a third set in motion ; these will disturb a fourth set, and so on. Thus the effect of the blow will be transmitted through the plate. Motion of this kind is called a Wave ; if at any instant we take two consecutive molecules each of which is at the same distance from its mean position and on the same side of it, the distance between these two molecules is the Length of the Wave ; the greatest distance traversed by a molecule is the Amplitude of the Wave. The intensity of the shock communicated by the wave to any object will vary, all other things being equal, as the square of the Amplitude. In the case considered the motion is produced by the alternate con- densation and expansion of successive portions of the substance, and the Wave is spoken of as a Wave of Elastic Compression. The effects we have described will not be confined to the particular Earthquakes. 385 line we have been dealing with ; very much the same results will be produced along every line that can be drawn through the plate from the spot where the blow was delivered ; a wave will travel along all such lines and it will move at the same rate along each. The general result may be thus represented. Suppose a number of spheres to be described with the point where the blow is struck as a common centre, and let the difference between the radii of two successive spheres be small. The shell between two adjoining hemispheres will be first com- pressed, will then expand, and will finally recover its original density : the shell that adjoins it will then go through the same series of changes; then the shell next beyond ; and so on. A wave in fact will spread in all directions from the point where the blow is struck, and all the molecules which are at a given time in the same state of disturbance will lie on the surface of a sphere whose centre is at that spot. The spheres will cut the upper surface of the plate in a series of circles whose common centre is the point vertically above that where the blow is struck, and all the points on any one of these circles will be reached by the wave at the same instant : its effect will be felt first at the common centre, then along the innermost circle, then along the one next outside, and so on. The wave however travels very fast, several thousand feet in a second, and the actual displacement of the particles of the steel which it produces is extremely small : neither the wave-motion nor the fact that it takes time to traverse the plate would be appreciable by the eye. But the wave makes itself apparent by its effect on the objects standing on the plate. If one of the vases be vertically above the point where the blow is struck, and if the aggregate effect of the upward motion of the molecules of the steel on which it stands, pass vertically up through its centre of gravity, it will be jerked into the air, but it may settle down again without being overturned. At any other point the motion of the molecules is inclined to the surface of the plate, and the result will be to produce a shock whose direction does not pass through the centre of gravity of the vase : this will not only throw the vase up but will turn it over as well. In the wave we have been considering, the molecules swing back- wards and forwards, that is they vibrate in the direction in which the wave travels ; it is distinguished as a Normal or Longitudinal Wave. Another wave may also be set up by the blow, in which the molecules will vibrate at right angles to the line along which the wave travels. This is styled a Transverse Wave. The Transverse Wave spreads out in spherical shells, but it travels more slowly than the Normal Wave. At every point then on the surface of the plate two waves may make themselves felt, the one following the other ; and the interval between the times of their reaching any spot on the surface of the plate will be longer the greater the distance of that spot from the point vertically over the blow. A further effect of the blow will be the production of sound. If the ear be laid close to the plate by one of the vases, a sound will be heard coming from the plate at the instant when the vase is upset. The wave which has travelled through the plate sets up a corresponding 2 B 386 Geology. wave in the air when it emerges, and this air-wave impinging on the ear produces a noise. But other sounds will reach the ear which come not out of the plate, but from the air above. Wherever the wave emerges from the plate it starts a sound-wave in the air : these waves travel much slower than the wave in the plate, only about 1200 feet per second, they therefore reach the ear later than the wave which has come direct through the plate. The general effect will be first a sharp jarring noise caused by a wave which has travelled nearly all the way through the plate, then a succession of rumblings caused by waves which have passed first through the plate and after- wards through varying distances of air. Again suppose a tank of water to be firmly bolted on to the plate. When the blow is struck, the water immediately above will be thrown up in a heap : this heap will subside, but a wave will spread out in concentric circles from it over the surface of the water. The rate at which this wave moves will depend on the depth of the water, but it will always travel much slower than the Wave of Elastic Compression in the plate. The motion of the particles in this wave, which is called a Wave of Translation, is totally different from that of the Wave of Elastic Compression ; each particle describes a circle or ellipse about its position of rest, and there is no condensation or expansion of the water. One more effect remains to be noticed. Instead of a tank of uniform depth let us have a tank which shallows gradually towards one end, the depth being decreased by a wedge-shaped mass of steel rigidly attached to the bottom. As the Wave of Elastic Compression passes through the steel beneath the shallow water, it will slightly raise the bottom ; this elevation of the bottom will lift up the layer of shallow water above and will give rise to a wave on the surface which is called a Forced Wave. The Wave of Translation, when it comes into the shallow water, will curl over and break with some amount of violence. The Forced Wave will also break on reaching the edge of the tank, but much less violently. The results just described would be difficult to distinguish on account of the small scale of the experiment, but exactly corresponding phenomena have been observed on a grander scale in the case of earth- quakes, and the different effects are here separated with less difficulty on account of their more marked development and the much larger area over which it is possible to extend our observations. Neglecting minor modifications we may say generally that when an earthquake happens the following phenomena, or some of them, occur, mostly in the order noted. First comes the great shock accompanied by hollow rumblings underground. If the earthquake has originated beneath the sea, a number of small waves break on the strand at the time when the shock is felt over the beach. Later on rumbling thunder comes through the air. Last of all in certain cases a great roller travels in from the open sea, and, if the beach shelves gradually, breaks on reaching the shallow Earthquakes. 387 water and rushes with fearful violence far inland sweeping everything before it in its headlong career. This is known as the Great Sea- wave. It is often more destructive than the earth-shock itself. The earth-shock is not felt at the same instant over the whole area affected : it reaches different spots at different times and is evidently produced by something transmitted through the ground. That that something is a wave or undulation is clearly proved by the effect which it produces the surface of the ground has been repeatedly observed to undulate when the shock is felt, and many of the displacements produced by earthquakes can be readily accounted for on the supposi- tion that the motion has been a swinging backwards and forwards, and scarcely seem to admit of explanation in any other way. As this wave passes through a solid mass it must be a wave of elastic com- pression. Again by noting the directions in which buildings are over- thrown it is possible to determine the direction in which the vibration took place. This was done by Mr. Mallet for the Neapolitan earthquake of 1857, and the directions of vibration at different places were found to converge approximately to the same spot. We are therefore justified in concluding that beneath this spot something corresponding to the hammer-blow in our experiment started a wave of elastic compression in the rocks of the earth's crust, that this wave spread outwards, and gave rise to the earthquake-shock wherever it reached the surface. What it is that starts the wave we do not exactly know. But earth- quakes always precede great volcanic eruptions, and in their case may be reasonably supposed to be produced by the jar occasioned when imprisoned steam rends asunder a rocky barrier that has held it in. It is therefore not unlikely that some similar action originates all earthquakes. This explanation is further supported by the fact that the belts on the earth's surface most subject to earthquakes coincide with the long trains of active volcanic vents. The wave travels with great velocity ; its mean rate in the Nea- politan earthquake was estimated by Mr. Mallet to be 788 feet per second. The same physicist has determined the rates of transit of waves produced artificially by explosions of gunpowder as follows. Feet per second. In loose Sand ...... 824 In jointed Granite ...... 1391 In solid Granite . . . . . .1773 In Slate and Quartz rock ..... 1220 The rate would be far greater if the rocks were homogeneous and unbroken by fissures and bedding planes. The reader must carefully bear in mind however that these are the rates at which the wave travels from spot to spot. The rate at which the particles swing backwards and forwards is far less. In the Nea- politan earthquake Mr. Mallet estimated this to be from 11 to 13 feet per second, and the amplitude to be from 2*5 to 4 '5 inches. It is likely that it is the Normal Wave that does the most mischief. Observation seems to show that a Transversal Wave or Waves do 388 Geology. exist, but that their effects are comparatively unimportant except for spots nearly vertically above the centre of disturbance. Wherever it emerges, the earth- wave starts a sound-wave in the air, Or perhaps transmits a wave through the body of any one who may happen to be on the spot, and this gives rise to the hollow rumblings that accompany the shock. Sound-waves may be started in the air at every point where the earth-wave emerges. If the earth-wave is passing through a rock in which its velocity is very large, it will travel faster than these sound- waves. Hence the sound-waves which started at distant points may reach the ear after the shock is felt, and the shock may be followed by continuous rolling thunder. If the centre of disturbance be beneath the sea, a Wave of Transla- tion is set going, which on a shelving beach produces the fearful sea- wave. The Wave of Translation travels less rapidly than the earth- wave. Hence the sea-wave arrives after the shock. In such a case too a Forced Wave is generated as the earth-wave passes beneath the shallow water, and small breakers are produced as the shock traverses the beach. A sound-wave may also travel through the sea, and hence a rolling thundery sound coming from the sea may precede, accompany, or follow the shock. The above is only a very broad general sketch of earthquake pheno- mena. In the actual case there are numerous circumstances which tend to complicate the results and produce effects which cannot be noticed here. Earthquakes as Geological Agents. The most important part which earthquakes play from a geological point of view is in helping on the work of denudation. Where an earthquake-wave traverses a country composed of inco- herent strata and channeled by deep steep-sided river valleys, great masses of the loose materials are detached by the shock from the steep faces of the valleys, -and large landslips are produced. If other con- ditions tending to produce landslips (see p. 581) are present, the shock comes in as an important auxiliary, and sets going bodies of rock which were already on the point of motion and required only a moderate impulse to start them. Fissures often of large size are formed along the lines where slipped masses have broken off, and the tumbled stuff itself becomes shattered and further fissured by its fall. These fissures serve as channels by which water finds its way to the strata below, softening them in many cases and rendering them unsafe foundations for the rocks above.* The slipped masses sometimes reach across a valley and dam back the river. Lakes are thus produced which have usually only a temporary existence. The dam sooner or later becomes cut through or undermined ; if it give way suddenly a devastating flood may result. In some cases the water is pounded back to such a height that it is able to flow away through some outlet above its former bed. The new channel is deepened, and in the end the river occupies it per- manently, deserting its former course, and the direction of the stream * Oldham, Quart. Journ. Geol. Soc. xxviii. (1872) 255. Earthquakes. 389 is altogether changed. The fissures produced by landslips are also in some cases enlarged by running water, and valleys altogether new are produced. It is possible that in hard rock fissures may be produced by the direct action of the shock, but no undoubted cases of such are known. The Great Sea- wave comes in with force enough to dislodge huge bodies of materials, and a part of these are doubtless swept back into the sea by its reflux. It would appear then that directly earthquakes are not agents of any geological moment, but that indirectly they furnish important help in carrying on the work of denudation. The exact study of earthquake phenomena is however of the utmost value in some branches of theoretical geology. The data afforded enable us to speculate as to what is going on in those deeply-seated portions of the earth's crust which are beyond the reach of direct observation. For further infor- mation on this subject the following works may be consulted : Mallet, Transactions of the Royal Irish Academy, xxi. (1848) 51 ; Reports of the British Association, 1850, 1851, 1852, 1858; First Principles of Observational Seismology; The Eruption of Vesuvius in 1872; The Neapolitan Earthquake of 1857; Phil. Trans, of the Royal Soc. 1861, 1862. Hopkins, Report of the British Association, 1847, p. 74. Numerous accounts of individual earthquakes will be found in the Quart. Journ. of the Geol. Soc., in the Neues Jahrbuch, the Zeits. der deuts. geol. Gesell., the Jahrbuch der k. k. geol. Reichs, and other geological periodicals. The published indexes in most cases allow of easy reference. CHAPTER VIII. PLUTONIC ROCKS. " Though there is a great characteristic difference between the Plutonic and Volcanic actions and their products, the two, when looked at largely, are seen so to inosculate that it is impossible not to refer them to an agency common to both." R. MALLET. Differences between Plutonic and Volcanic Rocks. An example will best make this point clear. I purchase of a dealer hand specimens of two Crystalline rocks from different places ; I deter- mine carefully the minerals of which they are composed, both by the unaided eye and by the help of the microscope ; I subject them to chemical analysis. The result of my investigations is that for all prac- tical purposes the two are made up of the same minerals, and that the relative proportions of the component minerals, and therefore the ultimate chemical composition, are the same for both. There was a time when the man who had done all this would have been held to have fulfilled the whole duty of a geologist ; the specimens would have been called by the same name, and the only difference between the labels affixed to the two would have been in the locality. But geolo- gists have long ago learned to be more inquisitive. I should certainly wish to make such an examination of my specimens as has been described, but I should want to do more ; I should not be satisfied till I had examined in the field the rocks from which they were taken. When I come to this, I find that one rock possesses all the characters which distinguish Volcanic rocks, and I place it in that class. In the other rock these characters are most conspicuously absent. Now the characters which decide that a rock belongs to the Volcanic class depend on the conditions under which these rocks solidified from a fused state. A rock in which these characters are wanting must have solidified under different conditions, and therefore from a geological stand- point ought to be separated from the Volcanic rocks and placed in another class ; and that too in spite of the fact that in mineral composi- tion it agrees very closely with certain Volcanic rocks. The rocks which we are going to class together under the head of Plutonic have a general agreement in mineral composition with rocks of the Volcanic group ; that is to say they are composed of the same minerals ; Felspars, Micas, Hornblende, Augite, and other com- Plutonic Rocks. 39 1 mon silicates are the chief components of both ; and further it is sometimes possible to find a Plutonic rock and a Volcanic rock which are made up of the same minerals mixed together in the same pro- portions. But Plutonic rocks differ from Volcanic rocks in these points. 1. They never show vesicular, slaggy, or glassy portions. 2. They have a tendency to be more coarsely and thoroughly crystalline. 3. In their crystals water-cavities are abundant and cavities con- taining liquid Carbon Dioxide are also met with ; glass- or stone- cavities are rare. 4. They are never bedded. The points of agreement between the two classes of rocks seem to indicate that both came from the same source; the differences will be accounted for if we suppose that the Plutonic rocks cooled and hardened under pressure and slowly, instead of having consolidated on or near the surface. Both we may conclude were made out of the same raw material ; but the process of manufacture was not the same in the two groups. Now under every volcano there must be a reser- voir of matter in a state of hydrothermal fusion, beneath, probably at some considerable depth beneath the surface. When the volcano becomes extinct the contents of this reservoir will cool and harden, but they will solidify under very different conditions from subaerial lava. The pressure of the overlying rock will prevent the boiling up of steam and the formation of vesicular structure : there will be no flow to produce slaggy structure ; cooling will be slow, and therefore no glass will be formed, and the circumstances will be favourable for the production of large crystals. The water will be retained by pressure, and portions of it will be caught up and enclosed in crystals as they grow. Cavities containing liquid Carbon Dioxide can have been formed only under considerable pressure. There will obviously be nothing to give rise to bedding. The hypothesis then that Plutonic rocks are the hardened contents of the deep-seated reservoirs of volcanoes, accounts fully for their distinctive characters, both negative and positive, and at the same time explains the resemblance which they bear in mineral composition to Volcanic rocks. The name has reference to their subterranean origin, and is derived from Pluto, the Latin god of the underground realm. Some Granites are the most typical examples of Plutonic rocks and the class is for this reason sometimes styled Granitic. But there are Granites which do not belong to the Plutonic class, so that this name is open to exception. Plutonic Rocks necessarily intrusive. The existence of a volcano implies that the contents of the reservoir which feeds it are in a state of expansion, and are struggling vigorously to tear a way through the rocks which form its walls. Many fissures will be rent open besides the one which serves as a chimney to the volcano, and every one will be injected with molten matter. The mass, when it hardens and is exposed to view, will be seen to send out dykes and veins into the surrounding rocks, and to behave generally in an intrusive manner. 392 Geology. It will be irruptive, but not eruptive. We may therefore add one more to the characters which distinguish Plutonic rocks. 5. They are all intrusive. It does not necessarily follow that there has been a volcano over every body of Plutonic rock. The rents which are torn open along its margin are attempts to establish a volcano ; but if none of them reach the surface the attempt was unsuccessful. Mode of Occurrence of Plutonic Rocks. Plutonic rocks always occur in masses, often many square miles in extent. These are occasionally approximately circular or elliptical in outline, but fre- quently their contour is most irregular. Fig. 123, which shows a ground-plan of a boss of this kind in the Plutonic Eock. ^ |=3=i Bedded Rocks. Section along the line A B. Fig. 123. PLAN AND SECTION OF A MASS OF PLUTONIC ROCK, Co. DONEGAL, IRELAND. Scale, 2 inches to a mile. north-west of Ireland and a section across it, will give an idea of the character of these masses. The section shows how the igneous rock cuts across the bedding of the stratified rocks which surround it, and its intrusive character is further made clear by the dykes which it gives off'. When the boundary is followed on the ground, innumerable little tongues and shoots are seen thrusting themselves into fissures too small to be shown in the drawing, and we readily picture to ourselves how this was once a seething mass of molten matter that tore and struggled against the bonds which held it in. Granite of Devon and Cornwall. The bosses of Granite met with in Devon and Cornwall furnish excellent instances of masses of Plutonic rock. The sketch-map and section in fig. 124 show the position and relation to the adjoining rocks of the Granite of Dartmoor. The surrounding country contains two very distinct groups of bedded sedimentary rocks. The upper, shown by the darker tint, consists of Plutonic Rocks. 393 Section along the line A B. 124. GEOLOGICAL SKETCH-MAP OF DARTMOOR. Sandstones and Shales, often with much carbonaceous matter, and here and there beds of impure Anthracite. Some earthy Limestones occur among the lower beds. The underlying group is made up of Clay-slates and hard Grits. Both groups are thrown into a number of complicated folds and undula- tions, but in spite of these have a general dip to the north, as shown in the section, so that their separ- ate beds come out to the surface in lines trending on the whole east and west. Now it will be noted that one boundary of the Granite runs nearly due north and south, or directly across the bedding. This is just what would happen if the Granite had been forced up from below, and had burst through the adjoining rocks. That this is the way in which it came here is shown by several points in the behaviour of the rock. The adjoining rock has often undergone just the same sort of baking and alteration as we have already seen occurs along the margins of intrusive dykes. The Granite also sends veins into the rock in contact with it, and has caught up and enclosed in itself portions of the beds which it pene- trates. All the evidence therefore leads us to the conviction that this is a body of rock which has been forcibly intruded in a molten state, and similar reasoning leads us to a like conclusion with regard to the other Granitic bosses which occur between Dartmoor and the Land's End. The probability is, that beneath the whole of this district there stretches a body of Granite ; that this was once in a state of fusion, and was then buried under a much greater thickness of bedded rocks than at present, and that every here and there bosses, projecting above the general surface of the mass, were thrust up into the overlying cover- ing. Denudation has stripped off enough of the capping to expose the summits of these bosses, but has not worked its way down to the spread of Granite beneath from which they all spring.* Granite Veins. In the cases just described among the facts tending to establish the intrusive behaviour of the Granite were the sending of veins into the adjoining strata, the occurrence of included blocks, and certain alteration of the adjoining rocks which goes by the name of contact-metamorphism. We will now give a few further illustrations of these occurrences. It was from the observation of veins proceeding from a Granitic mass * For a description of another area of intrusive Granite see Jukes' Manual of Geology, 3rd ed. pp. 241-245. 394 Geology. and penetrating the overlying rock that Hutton was led to assign an igneous origin to Granite.* In some cases perhaps the appearance of veins may be deceptive ; it is just possible that what look like veins may sometimes be portions of the adjacent rock, which yielded more readily to alteration than the main body of the rock itself and became converted into a substance indistinguishable from Granite, while the rest of the rock remained comparatively unaltered. But such an explanation is not admissible in those cases where Granite veins traverse rocks, such as Limestone, which no amount of alteration could convert into Granite. Of many such instances we may take the following as an illustration. The strata in the region between the Massawippe River and Canaan, in Canada, are in many places pierced by considerable masses of a beautiful Granite, which consists of white Quartz and Felspar, with a rather sparing amount of Mica uniformly mixed. Its intrusive nature is clearly shown by the Granite dykes, which proceed from it in various directions. One of the largest masses measures about six square miles ; it appears to displace the calcareous strata which it penetrates, as these are observed to dip from it in several places. At one spot, within a short distance of the edge of the granitic nucleus, a great number of Granite dykes are seen, cutting the basset edges of the Limestone beds, the whole having been worn down to a horizontal surface, a portion of which is represented in fig. 125. Some of the main dykes are from two to three feet in Fig. 125. DYKES OF GRANITE CUTTING THROUGH LIMESTONE. Scale, about 1 -500th. breadth, and divide into a multitude of irregular and reticulating branches, many of which are not more than the eighth of an inch wide. In the face of an escarpment, which rises from the Granite nucleus to this horizontal surface, a large dyke, of which all the others are probably ramifications, can be traced down to its source, t * Playfair, "Illustrations," Works, 1822, vol. i. 101, 308. t Report on the Geological Survey of Canada up to 1863, p. 434. Plutonic Rocks. 395 Good illustrations of Granite veins will be found in Plate V., and on pp. 168-187 of De la Beche's " Report on the Geology of Corn- wall, Devon, and West Somerset ; " and in Professor Ramsay's " Geology of the Island of Arran." Included Blocks in Granite. Intrusive masses of Plutonic rocks often enclose fragments apparently torn off from the rocks through which they have forced their way, and these blocks frequently show an external baked, or otherwise altered, coating. The reader will recollect that the same thing is often observed in the case of dykes and other intrusive igneous masses. The following is one out of many such cases. In the Pyrenees is a Granite containing many blocks of a dark-blue Limestone, identical with a rock found in the neighbour- hood. The outside crust of these fragments is converted into White Marble, the crystalline texture gradually disappears towards the interior, and the centre has the same colour as the rock from which they were derived. It does not necessarily follow that all included masses of foreign rock which are met with in Granite have come there in the way just described. We shall see in the next chapter that some Granites have arisen from the intense alteration of derivative rocks. Now if some parts of these rocks were more susceptible of alteration than others, it may well happen that portions of the less easily altered beds may remain comparatively unchanged in the middle of the crystalline mass which resulted from the reduction of the more readily altered strata. Cases of this sort have been noticed among the Granites of the south of Scotland by Dr. J. Geikie (Geological Magazine, iii. 533).* Instances have been described of included masses of enormous size in Granite, f One cannot help suspecting that in such cases we are dealing with an altered group of beds of variable composition ; that some have been converted into Granite, while others were better able to resist alteration ; and that the supposed included masses are really portions of the latter, rocks in fact that remained unaltered while the beds on either side of them were altered so as to put on a granitic form. Contact - metamorphism by Granite. We have already seen how rocks in contact with intrusive volcanic masses are frequently baked, hardened, and otherwise altered for a short distance from, the plane of junction : just the same effect has been produced in the neighbourhood of Plutonic masses. Thus at Grange Irish, in the Carlingford Mountains, in Ireland, a fine-grained Hornblendic Granite sends veins into beds of over- lying Limestone : the Limestone is converted into a bluish sugary Marble containing Garnets. Here too the Limestone has reacted on the Granite itself and wrought a singular change in its com- position. Professor Haughton gives the following results of his analysis of the * See also Geological Exploration of the 40th Parallel, i. 120. t Zirkel, Petrographie, i. 503. 396 Geology. 71-41 12-64 476 1-80 0-63 Quartz Potash Felspar . Hornblende 17 "16 per cent. 67-18 15-40 5-47 . 3-03 47 28 7 15 1 52 56 23 44 48 Anorthite or Felspar Hornblende Lime 85-84 1416 per cent. Granite ten yards from the point where it comes in contact with the Limestone. Equivalent to Silica . Alumina Protoxide of Iron Lime . Magnesia Potash Soda . The dykes which proceed from this rock and penetrate the Lime- stone are found to have the following composition. Equivalent to Silica . Alumina Protoxide of Iron Lime . Magnesia Comparing these two analyses we see that " the quantity of Horn- blende remains almost unaltered, and that the effect of the addition of Limestone to the melted Granite has been to convert the Quartz and Orthoclase into Anorthite. In this operation the alkalies of the Orthoclase have disappeared ; the Lime, being a more fixed base at high temperature, has altogether displaced the alkalies."* The change of Limestone into crystalline Marble by intruded masses of Hornblendic Granite in the Island of Skye has been described by Professor Geikie.t Under similar circumstances Clay- slates and Sandstones have been converted into Micaceous Schists, Gneiss, or similar foliated rocks. Contact-metamorphism is not in itself a proof that the Granite in whose neighbourhood it occurs was necessarily intrusive. Some Granites we shall learn shortly have been formed by the melting down of rocks in situ, without the process having gone far enough to give rise to intrusive behaviour. Alteration occurs round such Granite masses, but in their case there will be a gradual transition from Granite, through less highly altered rocks, to beds quite unchanged : where Granite has behaved intrusively there will be a more or less marked line of demarcation between it and the rock it invades. Lithological Varieties of Plutonic Rocks. A very large number of Granites, Syenites, Diorites, Gabbros, Diabases, and Mica- traps are Plutonic. So are most Elvanites, and some Quartz- trachytes. The Crystalline rocks that have been specified as Volcanic in the majority of instances, put on forms approaching those of the Plutonic class in the more deeply seated portions of Necks and Dykes. Connecting - links between Volcanic and Plutonic Rocks. If we look only at the extreme types of Volcanic and Plutonic rocks they are marked off from one another sharply enough, but a moment's reflection will convince us that there can be no hard and fast line between the two classes. A slaggy, vesicular, crypto- crystalline Lava is as distinct as can be from a compact and largely- * Quart. Journ. Geol. Soc. xii. 192-198. t Ibid. xiv. 18. Plutonic Rocks. 397 crystalline Granite. But the interior portion of a thick lava-flow consolidated under some degree of pressure, and therefore approaches in the conditions under which it was formed, and consequently in its structural characters, to rocks of the Plutonic class. This is still more the case with Dykes and Necks. Indeed passages from typical rocks of the one class into equally typical rocks of the other class must exist, if we could only see them. If we could lay open an extinct volcano from the surface down to the reservoir which fed it, we should see the lava as we traced it down the chimney passing gradually into the more compact material of the neck ; this again would become more compact and crystalline the deeper we followed it, and would in the end merge by imperceptible gradations into the Plutonic contents of the reservoir. We cannot point to any single section in which the whole passage is visible ; but denudation has laid bare for our inspection the interior of many extinct volcanoes, and by piecing together sections obtained at different spots around the same vent, we obtain a result as compre- hensive and convincing as if we had the whole before us at one view. In no case perhaps has this been so successfully done as by Professor Judd in the case of the extinct volcanoes of the Western Islands of Scotland and of Schemnitz in Hungary. Here denudation has so to speak dissected out the great volcanic piles and revealed their internal structure down to the entrances to the underground reservoirs which fed them, and a perfect passage from subaerial lavas with all their distinctive characters into Plutonic rocks of the most pronounced type can be traced.* Hydrothermal Origin of Plutonic Rocks. We have evi- dence quite as conclusive as in the case of Volcanic rocks that Plutonic rocks were formed not by the agency of heat alone but by Hydro- thermal Action. The presence of water-cavities in their crystals is alone proof of this. Again certain peculiarities in the lithological texture of Granite point to the same conclusion. Of the minerals that enter into its composition Quartz is the most infusible, and therefore, if Granite had been formed by the cooling of a mass liquefied by heat alone, the Quartz would have been the first to solidify. But the very reverse is the case. The Felspars and Micas are more or less com- pletely crystallized, the Quartz fills up the spaces between their crystals and has evidently been moulded upon them. If the con- stituents of Granite were before solidification in a state of hydrothermal fusion, the order in which the minerals crystallized might be very different from that of their fusibility, t An attempt has been made to prove this point by fusing Quartz under the oxyhydrogen blowpipe. The resulting product has a * Quart. Journ. Geol. Soc. xxx. (1874) 242 et seq.; xxxii. (1876) 292. t Durocher has however attempted to show by a most ingenious line of reasoning how Quartz might retain a considerable degree of fluidity or plasticity down to a temperature far below its freezing-point during the cooling of a molten mass having the same elementary composition as Granite (Comptes Rendus, 1845, xx. 1275). See also Vernon Harcourt, Report of British Association, 1860, p. 181 ; Sorby on Mount Sorrel Syenite, Geol. and Polytech. Soc. of West Riding of Yorkshire, May 28, 1863 ; Scheerer, Bull, de la Soc. Geol. de France, 2nd series, iv. 479, quoted by Scrope, Volcanoes, p. 283. 398 Geology. specific gravity of 2 '3 while the specific gravity of the Quartz of Granite is 2 '5. But the argument is of no value ; the substance obtained by fusion is a glass, the Quartz of Granite is crystalline, and the specific gravity of a substance is always higher in the crystalline than in the glassy state. If we could cool our fused Quartz slowly enough to allow it to crystallize, there is no reason to doubt that it would have the same specific gravity as the Quartz of Granite. All the fact proves is that the Quartz of Granite has not been formed by the rapid cooling of a fused mass, and this we know on other grounds. CHAPTER IX. METAMORPHIC ROCKS. " ' Was the world not made at once then ? ' said Felix. ' Hardly,' answered Jarno ; ' good bread needs baking.' " WILHELM JV1 BISTER'S TRAVELS. SECTION I. GENERAL VIEW AND INSTANCES OF METAMORPHISM. THERE yet remain a number of Crystalline rocks which cannot be placed either in the Volcanic or the Plutonic group. They are not volcanic, for they possess none of the structures characteristic of that class. They agree with Plutonic rocks in many respects : many of them are crystalline to a very high, degree ; water-cavities are plenti- ful in their crystals ; and they give other convincing proofs of having assumed their crystalline state under considerable pressure and slowly. But they can none of them be called Plutonic rocks : some cannot be placed in this class because they are distinctly bedded ; others because there is a more or less pronounced arrangement of their crystalline minerals in layers, a structure which is known as Foliation or Schistose structure ; others again are not foliated, and bear the closest resem- blance to Plutonic rocks in hand specimens, but show when they are studied on a large scale behaviour which makes it impossible to place them in the Plutonic class. Three such diverse kinds of rocks might seem to require three distinct classes to receive them : they are all placed under one head because there is reason to believe that they were all formed by the same process. Most likely they were originally all of them bedded derivative rocks, but they have been altered in such a way as to acquire a crystalline texture ; the three forms enumerated probably correspond to different degrees of alteration. This process of altera- tion is known as Metamorphism, and the rocks which have undergone it are styled Metamorphic Rocks. General Remarks on Metamorphism. The rocks we have hitherto considered, both of the igneous and derivative classes, have come down to us pretty much in the same state in which they were originally formed. Time indeed has not passed over their heads without leaving its mark upon them in various ways ; they have been hardened, new minerals have been introduced into them, and they 4OO Geology. have undergone other changes of a similar nature. But in all the cases that have so far come under our notice the utmost amount of alteration that has been effected does not amount to much ; the characters we rely upon as indications of origin may have been disguised to a small extent, but no rock we have yet met with has been so thoroughly transformed that we are no longer able to say without hesitation how it was produced and what was its original nature. The minor modifications however with which we have already become acquainted will suggest to us the possibility of there being rocks which have been altered to a much greater degree ; and obser- vation shows us many rocks whose peculiar character can be explained only on such a supposition. These are the rocks to which we specially apply the term Metamorphic, and to these we shall devote the present chapter. Strictly speaking, it would be hardly possible to find a rock which is not metamorphic to some degree, but the term is usually restricted to those rocks which have suffered transformations of so radical a nature that it is only by long and attentive study that we become convinced they are merely the altered forms of some of those rocks we have become already familiar with, and only by calling in the aid of the chemist and mineralogist that we can form any reasonable con- jectures as to the processes by which the alteration has been effected. It will be as well before we come to a formal description of the rocks usually classed as metamorphic, or indulge in any speculation as to the causes to which metamorphism is due, to lay before the student a description of one or two districts in which rocks of this class occur. In this way he will at the outset become acquainted with the nature of the evidence on which geologists base their belief in the metamor- phic character of the rocks in question, and will see that it is on broad geological grounds that they are led irresistibly to this conclusion. It will appear that when rocks of this class are studied in a large way in the field, they are found to possess on a great scale many of the distinguishing characters of derivative deposits. They consist of alternations of rocks of different character and composition, just as in sedimentary beds we meet with alternations of Shale, Sandstone, and Limestone. The several members are in many cases laid in regularly- bedded order one upon another, and range over the country according to the direction and amount of their dip. Among them we occasionally find beds still retaining their fossils, conformably placed with regard to the strata above and below them, and evidently forming part of the same series.* Some of the minor peculiarities of derivative deposits are moreover still to be detected in rocks of this class. Thus Dr. Sorby has recognised in Mica-schist exactly the same ripple-drift structure which we have already seen is so common in Sandstone, t Lastly, we occasionally meet with transitions of the most gradual character between these rocks and stratified fossiliferous deposits, the * Murchison, Siluria, pp. 163-169 ; Russia and the Oural Mountains, 402, 438, 465 ; Brochant, Annales des Mines, 1st series, vol. iv. ; Leonhard's Jahrbuch, 1840, p. 352. t Quart. Journ. Geol. Soc. of London, vol. xix. 401. General View of Metamorphism. 401 two shading imperceptibly into one another. From extensive obser- vations of this nature we arrive at the conclu- sion that the rocks now under consideration were originally sedimentary deposits, and that they have been subsequently altered so as to acquire a crystalline texture and certain struc- tural peculiarities. When we have thus seized on a clear view of the general nature and probable origin of the Metamorphic rocks, we shall have to notice sundry laboratory experiments which point to the same result as our field observations, and give us an insight into the way in which the transformation has been brought about. Metamorphic Rocks of County Donegal. The first instance that I shall lay before the reader is taken from a part of the large tract of Metamorphic rocks in the north- west of Ireland, and is illustrated by the section in fig. 126.* On the left we have sandy Limestones (1), and white Sandstones (2), with intrusive masses of Diorite (G). These beds appear to have undergone some, but not a very large, degree of metamorphism. Some of the Lime- stones are closely grained and semi-crystalline, and among the Sandstones we meet with beds of Quartzite here and there, but the main body of the rock shows feeble signs of alteration. Next comes a thick mass of Quartzite (3), forming the noble hill known as Errigal Mountain ; this is an intensely hard, very closely-grained rock, crystalline in parts, well jointed and splintery, very regularly and un- mistakably bedded. That it is a sedimentary Sandstone cannot be doubted, but the amount of alteration necessary to turn any Sandstone into a Quartzite of this nature must have been considerably greater than that which the under- lying beds (1) and (2) have suffered. Upon the Quartzite there lies a group of beds (4) consisting mainly of Mica-schist, with inter- bedded layers of Limestone, Gneiss, and a rock that cannot be distinguished from Granite. The progressive increase of altera- tion, which occurs in passing from (1) and (2) to (3). becomes still more strongly marked * On the Granites of Donegal, etc., British Association (1863) ; On the Granite Rocks of Donegal, R. H. Scott ; Journ. Royal Geol. Soc. of Ireland, i. 144 ; Geol. Mag. ii. 216, vii. 553. 2c < 02 4O2 Geology. here : the Mica-schist and Gneiss are foliated ; the Limestones are all highly crystalline, in some cases converted into Statuary Marble, and some of them contain plates of Mica ; some of the granitic beds are coarsely-grained crystalline aggregates. But in spite of the advanced stage of metamorphism through which they have passed, these rocks still retain a most characteristic and strongly-marked bedding. In this respect they cannot be distinguished from a group of interbedded Shales, Sandstones, and Limestones ; and it is only when we break into them and become aware of their intensely-crystal- line texture, that we realize the amount of alteration they must have gone through to reduce them from the condition of ordinary derivative sediment to their present state. As we go towards the right across the group last mentioned a gradual change becomes apparent ; the beds of Granitic Gneiss and Granite become thicker and more numerous, and the intervening bands of Mica-schist thinner and fewer, till we at last reach ground where the latter can no longer be detected and which is wholly occupied by Granite (5). We can here be no longer certain of the existence of bedding, but the rock is traversed by a number of divisional planes ranging parallel to the stratification of the undoubtedly bedded rocks on the left, and the layers into which the rock is divided by these planes differ from one another in grain, mineral composition, and other peculiarities, just in the same way as the successive beds of an ordinary stratified derivative deposit are observed to do. Very grave suspicions therefore arise in our mind that this apparently amorphous crystalline mass was once a bedded rock, and that the signs of stratification have been all but effaced by the intense degree of metamorphism to which it has been subjected. The section just described offers to our notice a group of rocks which, in bedding and other characteristics, presents the strongest analogy to derivative deposits, while it differs from these in possessing a more or less pronounced crystalline texture. On the first ground we are led to think that the rocks must have been originally derivative, while the impossibility of rocks as crystalline as these are having been formed by derivative methods alone, convinces us that the crystalliza- tion must have supervened after their deposition. And it is in favour of this view that, in proportion as the crystalline texture becomes more and more marked, the traces of bedding become less and less distinct. Metamorphic Rocks of the Western Territories of the United States. Numerous instances of a gradual passage from derivative rocks of the ordinary type into rocks possessing a more or less pronounced crystalline texture have been noticed in the Sierra Nevada and the ranges that adjoin it. The following case in the West Humboldt Mountains will serve for an instance. A group of Clay- slates, some siliceous, and other beds more purely siliceous in com- position, pass by most gradual transitions into rocks possessing all the characters of Elvanite. In the earlier stages of the metamor- phism white crystalline Felspars make their appearance massed together in parallel bands separated by belts of dark felsitic rock. In a more advanced stage distinct^ individualized crystals of Felspar General View of Metamorphism. 403 are scattered about in a felsitic ground mass which contains fully- developed crystals of Felspar or Quartz or both. The unaltered rocks and the various forms produced by the different stages of alteration have the same chemical composition. The close connection between the crystalline beds and the rocks from which they were derived is further shown by the fact that when the alteration extends into marly strata the felsitic matter of the altered rocks contains a good deal of reddish Calcite, this mineral not occurring in the rocks produced by the alteration of the sandy and clayey beds.* It is by a consideration of a large mass of evidence, similar to that furnished by the two cases just given, that geologists are led to a con- viction of the metamorphic origin of the rocks now under consideration. Effects of Metamorphism. The most obvious result of metamorphism has been to superinduce a crystalline structure in rocks originally derivative. Its effects upon the bedding vary in different cases ; sometimes the stratification can be still distinctly traced, some- times it has been obscured or replaced by the structure already alluded to as foliation, sometimes neither the original bedding nor foliation is present ; if a rock originally contained fossils, these are usually obscured and frequently altogether effaced by metamorphism. Subdivisions of Metamorphic Rocks. As was indicated in the remarks with which this chapter was opened, the Metamorphic rocks fall naturally into three groups. 1. Those which still retain their bedding. 2. Foliated or Schistose rocks. 3. Certain massive Crystalline rocks closely resembling in many respects rocks of the Plutonic class. About the origin of the first there can be no doubt : their bedded structure, the occasional presence in them of fossils, and the passage that can often be traced from them into the unaltered rock by whose metamorphism they were produced, put this beyond question. We get an inkling as to the way they were produced by observing that they closely resemble the bands of baked and altered rock, which have been already noticed as frequently adjoining intrusive dykes, and by sundry laboratory experiments by which some of them have been produced artificially. The origin of the Schistose rocks is not quite so evident at first sight. Considered by themselves and judged only by isolated hand specimens, there is little about them to suggest a relationship to the rocks of the first class ; but a study of them in the field does occasionally show a passage from them through the latter to unaltered derivative rocks, and leads us to believe that they represent a more advanced stage of metamorphism, in which the rock, without being actually fused, was so far softened and its coherence weakened that its constituent minerals were free to move among one another and group themselves in separate layers, and may be to become decomposed into their chemical elements, so that the latter were able to enter into new combinations. There is not much difference of opinion among geologists as to the * Geological Exploration of the 40th Parallel, i. 268, 271. 404 Geology, origin of the rocks which make up the two classes just described ; they are generally admitted to be altered derivative deposits. But by no means the same unanimity prevails with respect to many of the rocks which we propose to place in the third class. The reasons for believing them to be intensely-metamorphosed deposits will be given when we come to treat of the rocks themselves. SECTION II. DESCRIPTION OF THE PRINCIPAL VARIETIES OF THE METAMORPHIC ROCKS. We will now notice some of the chief varieties of the Metamorphic rocks. 1st Class. THOSE WHICH STILL RETAIN TRACES OF BEDDING AND OTHER, OBVIOUS PROOFS OF THEIR ORIGINALLY DERIVATIVE CON- DITION. (1 a) Siliceous Members. Quartz Rock or Quartette. An aggregate of Quartz grains bound together into a very hard compact rock with a splintery fracture. It is not very obviously crystalline, but the presence of Quartz crystals in cavities, and occasionally in the body of the rock itself, shows that a crystalline texture has been begun to be set up in it ; the matrix when examined by polarized light is seen to have crystalline texture ; and the grains when examined through a lens have often a semi-fused aspect. Intermediate varieties occur between the most crystalline form of Quartzite and the more closely-grained Sandstones, which lead us to believe that the former is an altered condition of the latter ; and this conclusion is confirmed by the following considerations. Blocks of Sandstone which have been for some time in use for the hearths of furnaces are converted into Quartzite, and have sometimes a prismatic structure given to them ; * and when igneous rocks have burst through Sandstones, a belt of the latter surrounding the intrusive mass is frequently found to be baked into a perfect Quartzite. A very good instance of the alteration produced by heat in siliceous rocks is furnished by the so-called Vitrified Sandstones (Verglaste Sandsteine) of West Germany and the Thuringerwald. Many cases are known in these districts where Sandstone has come in contact with Basalt and has acquired in consequence a columnar structure and experienced the following curious metamorphism. The spaces between the Quartz grains are filled up by a singly-refracting glassy substance, and the grains themselves are traversed by numerous cracks. The glass is full of minute crystallized bodies : some which seem to be hexagonal are very likely Nepheline ; other needle-shaped crystals may be referred to Hornblende ; besides these there are countless other diminutive crystalline needles variously aggregated, and sometimes forming an intricated interwoven mass of threads. The glass possesses true fluxion structure, which is shown by streams of Microliths, and empty, rounded, dark-edged vesicles occur in it. Everything indicates * De la Beche, Researches in Theoretical Geology, p. 109. Description of Metamorphic Rocks. 405 that the glass has been liquefied by heat, and it seems probable the high temperature of the adjoining Basalt fused the ferruginous and calcareous matrix of the Sandstone, that the fused mass hardened into a glass, and that the crystalline enclosures are devitrification products. Beyond cracking them the heat seems to have produced no effect on the Quartz grains. This view is supported by the fact that a product containing similar enclosures has been produced artificially by fusing a mixture of Clay and Quartz.* The metamorphisin in this case seems to have been brought about by dry heat : had there been water present and pressure enough to retain it, it would almost certainly have exerted a solvent action on the Quartz. But the action of heat is by no means the only way in which crystalline texture may be set up in Sandstones of clastic origin. In some Sandstones, which have certainly never been subjected to a high temperature, minute but perfectly-formed crystals of Quartz are found attached to the rounded Quartz grains of which the rock is made up ; and in some cases the spaces between the grains have been so completely filled up with crystalline Quartz that the rock has been converted into a true Quartzite. Here the Quartz must have been deposited in a crystalline state from solution in water or alkaline water ; it may have been derived from the decomposition of grains of Felspar in the rock ; by such decomposition we should obtain at the same time Silica in a soluble form and alkaline solutions eminently fitted to dissolve it.t The typical Quartzites are almost purely siliceous rocks, but varie- ties occur enclosing crystals of Felspar and other minerals, and one form contains Mica in sufficient abundance, and arranged with suffi- cient regularity in layers, to give it a schistose structure. Lydian Stone. Differs mainly from Quartzite in containing small admixtures of Alumina, Carbon, and Oxide of Iron ; the amount of the impurities varies very much in different specimens ; often ribboned or laminated. As the more typical Quartzites have arisen from the metamorphism of highly-siliceous Sandstones, Lydian Stone would seem to be the result of the alteration of the more impure argillaceous Sandstones and sandy Shales. A case has been already given where Shales have been converted by contact with a dyke into a sort of Lydian Stone (p. 376). Innumerable varieties of rock to which it is scarcely possible to assign definite names have arisen from the alteration of impure Sand- stones. The old-fashioned name Gramvacke, which is a somewhat vague, indefinite term for hard felspathic Sandstones, may be used to include them. Felspathic Sandstones, more or less altered into Grauwacke, make up a great part of the Silurian rocks of the southern uplands of Scotland. J * Zirkel, Mikros. Beschaff. 488. t Sorby, Quart. Journ. Geol. Soc. xxxvi. (1880), Anniversary Address, p. 62 ; J. A. Phillips, ibid, xxxvii. (1881) 13 ; Daubree, Geol. Experimental, i. 226. J. Geikie, Quart. Journ. Geol. Soc. of London, xxii. 513 ; A. Geikie (Me- moirs of the Geological Survey of Scotland), The Geology of East Lothian, The Geology of East Berwickshire, and Explanation of Sheets 3 and 15. 406 Geology. (16) Argillaceous Members. Clay -slate, Thonschiefer, Dachschiefer, Phyllite. This rock may well be placed in the Metamorphic group, though the main alteration that has been wrought in it is not so much the development of crystalline texture as hardening and the production of the peculiar structure already described under the name of Slaty Cleavage. There has been however in many cases a development of microscopic crystals, chiefly plates of micaceous, talcose, and chloritic minerals. In some Clay-slates too the spaces between the clastic grains are filled with a singly-refracting substance, which may be allied to Opal or may be a pectous silicate. Many of the varieties of Clay-slate correspond in mineral and chemical composition with the various forms of argillaceous Shale, and differ only from the latter in their more perfect induration and the possession of cleavage. Both these distinguishing characteristics are due, as we have already seen, to pressure. But Dr. Sorby has shown that many of the older Slates contain very little clayey matter and are mainly made up of very small plates of Mica. In sections cut perpendicular to the cleavage the edges of these plates look like very fine needles, which " depolarize " light ; but when sections parallel to the cleavage are examined, the rock is seen to be an aggregate of plates, lying most of them parallel to the cleavage planes, which have a very feeble " depolarizing" power. In their action on polarized light and in their optical properties generally these plates agree with Mica. The general character of the rock suggests that it was formed out of a micaceous mud derived from the decomposition of a finely-grained crystalline rock rich in Mica, and that the mud was sorted by gentle currents. The deposit has been subsequently hardened and cleaved by pressure.* The cleavage planes of some Clay-slates are flecked over with flakes of Mica, Talc, Chlorite, Chiastolite, and other mine- rals. In these varieties metamorphism has advanced a step further, and has given rise to an incipient foliation which allies them to the Schistose rocks. Porcellanite. The metamorphism of some Clays has given rise to a rock which, from its resemblance to earthenware or china, has received this name. When stained red it is known as Jaspery Porcellanite, or Porcelain Jasper. (1 c) Calcareous Members. Crystalline Limestone, Korniger Kalkstein. All tolerably pure Limestones seem to have a tendency to assume a crystalline texture. This is doubtless owing to the readiness with which Calcium Car- bonate passes into the crystalline state. Many parts of the Carboni- ferous Limestone of the centre of England for instance, which there is no reason to suppose has ever been subjected to any special meta- morphosing influence, have become very decidedly crystalline. The * Quart. Journ. Geol. Soc. xxxvi. (1880), Anniversary Address, p. 68. Description of Metamorphic Rocks. 407 stems of encrinites are entirely converted into Calcite, and sometimes the only trace of the original structure that remains is the little round hole in the middle of each plate of the stem. The dissolution and precipitation of Carbonate of Lime by percolating water has no doubt had much to do with producing this result, but very likely the slow working of molecular change may have aided. With a tendency of this nature, it is no wonder that Limestones, when metamorphic agencies are brought to bear on them, are readily converted into a semi-crystal- line or crystalline state. All sorts of varieties occur, some largely and coarsely crystalline ; others containing crystals of foreign minerals, such as Mica, Chlorite, Garnet, etc. ; and some of a beautifully even, closely- grained texture, resembling loaf-sugar and capable of taking a polish. The last, which are known as Saccharoidal Limestones, furnish the Statuary Marble of commerce. We have already noticed one case in which Limestone of the ordinary derivative type is observed to pass on a large scale into a Crystalline rock (p. 402) ; the same change is often noticed where Limestones are invaded by igneous rocks, as for instance in the case of the Chalk of the north-east of Ireland.* Crystalline Limestone has also been produced artificially. When Limestone is burned in the open air, the Carbonic Acid is driven off and Quicklime remains behind ; but Sir James Hall, by confining Limestone in a closed vessel so as to prevent the escape of the Car- bonic Acid, and heating it, succeeded in converting it into Statuary Marble, t The experiments were made with Chalk, common Lime- stone, Marble, Spar, and fish-shells. In all cases the Carbonate of Lime, which had been introduced in the state of the finest powder, was agglutinated into a firm mass, possessing a degree of hardness, compactness, and specific gravity nearly approaching to those qualities in a sound Limestone and some of the results, by their saline frac- ture, by their semi-transparency, and their susceptibility of a polish, deserved the name of Marble. One specimen, formed from pounded Spar, was so complete as to deceive the workman employed to polish it, who declared that, were the substance a little whiter, the quarry from which it was taken would be of great value if it lay within reach of a market. This last and some others soon crumbled and fell to dust ; but many others resisted the air arid retained their polish as well as any Marble. In some of the micaceous metamorphic Limestones the Mica occurs in sufficient abundance to give them a schistose structure. Crystalline Limestones very frequently show a characteristic structure under the microscope. The rock is seen to be a mass of closely-packed, irregularly-shaped grains. Each grain is made up of a number of fine parallel bands, and in many grains fine parallel cracks cut across the bands. The grains are Calcite ; the bands are produced by polysyn- thetic twinning parallel to a face of - J 72 (p. 1 36) ; the fine cracks are the edges of cleavage planes. The position of the twinning plane * Buckland and Conybeare, Transactions Geol. Soc. of London, iii. ; De la Beche, Geological Observer, p. 700. t Edinburgh Phil. Transactions, vol. vi. p. 71. 408 Geology. varies from grain to grain, so that each grain has its own independent system of bands. Inostranzeff examined a number of crystalline Limestones micro- scopically and found that in some all the grains were twinned. These were composed almost entirely of Calcium Carbonate. In others the cleavage planes were present, but none of the grains showed twin banding. These had very nearly the theoretical composition of Bitter- spar and were Dolomites. In others again twin-banding was present in some but absent from others of the grains, cleavage being well developed in all. These contained Magnesium Carbonate, and the number of the untwinned grains increased with the percentage of this salt : the Limestones were probably a mixture of Calcite and Bitter- spar or Dolomitic Limestones.* In the case of these specimens there is every reason to think that twinning is confined to Calcite, and is absent from Bitter-spar. Unfortunately this does not seem to be universally true.t Ophicalcite is a crystalline Limestone with nodules, specks, or veins of Serpentine. The Serpentine grains are often seen to be altered Olivine. Dolomite. We have seen that some Dolomites and Magnesian Limestones have probably been formed by chemical precipitation, others have undoubtedly been produced by the alteration of Lime- stone. Such a change is called dolomitization. Various views have been held as to the methods by which the metamorphism has been brought about. Arduino in 1779, Heim in 1806, and Von Buch in 1822, suggested that where augitic igneous rocks have burst through or been irrupted below Limestones, magnesian vapour had risen from the fused mass, insinuated itself through the body of the rock, and given rise to dolomitization.^ This view was stoutly upheld by the last great geologist ; and he pointed to the huge dolomitic masses of the Tyrol, below which great bodies of Melaphyre exist, as a case in point. Subsequent examination has shown that the explanation will not hold good in this case ; but there is no doubt that Limestones are sometimes dolomitized when they come in contact with igneous rocks containing magnesian silicates. Bischof quotes a case, men- tioned by Coquand, of a Limestone in contact with Basalt, which became more and more magnesian the nearer it approached the latter rock. || But he does not believe the change to be due to the action of vapour, and prefers to account for it by the percolation of water holding Carbonate of Magnesia in solution. The latter salt he be- lieves was obtained by the decomposition of the Silicate of Magnesia of the Basalt by carbonated water, which found its way down between the dyke and the adjoining rock. An experiment of Durocher's, on the other hand, shows that the vapour method is a possible one. He heated together fragments of Limestone and Magnesium Chloride in * Jahrbuch der k. k. Geol. Reichs. xxii. (1872) ; Min. Mittheil. p. 45. t See Meyer, Zeit. Deutsch. Geol. Gesell, xxx-i. (1879), 445. See Naumann, Geognosie, i. 763-765, for details and references. Fournet, Bull. Soc. Geol. de France, 2nd series, vi. 506-516 ; Naumann, Geognosie, i. 768. || Chemical Geology, iii. 179. Description of Metamorphic Rocks a close vessel, and succeeded in converting part of Dolomite.* Even supposing however that magnesian vapours have been in some cases the cause of dolomitization, we can hardly suppose that their effect would extend to any great distance from the source which gave them off, and we can scarcely apply this explanation to account for the transformation of great masses of Limestone into Dolomite. The explanation that dolomitization was produced by the percolation of water holding Carbonate of Magnesia in solution is not open to this objection, t Professor Harkness has explained in this way the occurrence of Magnesian Limestones among the Carboniferous Limestone of Co. Cork. J They occur under two forms ; in some cases they are inter- stratified with the ordinary Limestone, in others they form vertical ribs, cutting across the bedding like igneous dykes. Wherever the ribs occur the rock is well jointed, the walls of each rib being formed by joints ; but where beds prevail the rock has little or no jointing. In the first case the dolomitizing solution found its way most readily down the open vertical fissures, and spreading into the adjoining rock, altered a band bounded by the joints that gave it passage ; where there were no joints, the easiest path was parallel to the planes of bedding through the most permeable strata. Professor Harkness also points out that in some cases the alteration is greater in the upper than in the lower part of the rock, as if the producing cause had been some- thing introduced from above. He supposed the dolomitizing agent to have been the Sulphate of Magnesia and Chloride of Magnesium of sea-water. According to Favre, these substances can alter Limestone into Dolomite only under great pressure, and at a temperature of 200 C., conditions which we can hardly suppose to have been present. It seems more likely therefore that it was a solution of Carbonate of Magnesia that worked the change. Ribs of Magnesian Limestone, like those just described, are also met with in the Carboniferous Limestone of Yorkshire, where they are known as " Dun courses." Other cases of Dolomite occurring in ribs will be found in Naumann's Geognosie, i. 766, 767, with many references to the literature of the subject. Magnesian Limestones are very frequently full of cavities, and Elie de Beaumont has suggested that these are the result of the decrease in bulk which would accompany the transformation of Limestone into Dolomite, if it were effected by the process we are now considering. The manner of the change was thus. Water holding Carbonate of Magnesia in solution percolated through the rock ; the tendency of this salt to unite with Carbonate of Lime caused it to be precipitated, and the carbonated water took up in its place Carbonate of Lime. Out of every pair of molecules of Carbonate of Lime one was removed in solu- tion, and its place was taken by a molecule of Carbonate of Magnesia. * Comptes Rendus, xxiii. 64 (1851). t Bischof, Chemical Geology, iii. 164-167, 179 ; Dana and Jackson, Silli- man's Journ. xlv. (1843) 120, 141; Nauch, Poggendorf Ann. Ixxv. (1848) 149. $ Quart. Journ. Geol. Soc. xv. 100 ; see also Wyley, Journ. Dublin Geol. Soc. vi. 109. 4io Geology, Now Bulk of two molecules of Carbonate of Lime = Atomic weight _ 200 Specific gravity of Limestone '" 2*7 Bulk of one molecule of Carbonate of Lime + one molecule of Carbonate of Magnesia Atomic weight _ 184 Specific gravity of Dolomite ~~ 2 '85* Whence 184 Bulk of resulting Dolomite _ 2.85 _ Bulk of original Limestone ~ 200 27 So that the shrinking ought to be between 12 and 13 per cent. In some actual cases Elie de Beaumont estimated the cavities at 12 '9 per cent, of the whole rock. The same explanation will apply to the formation of what are known as Potato-stones. These are pebbles w r hich are known by the fossils they contain to have been once Limestone, but which have been con- verted into Magnesian Limestone. They are hollow inside and the walls of the cavity are coated with crystals of Bitter-spar. If the alteration was effected by a solution of Carbonate of Magnesia, the con- sequent shrinking would account for the internal hollow. A group of rocks belonging to the Permian formation, at Barrow- mouth near Whitehaven, seems to furnish an instance where dolomiti- zation has been brought about by a solution percolating from above downwards. The section is 3. Red Marl with a bed of Magnesian Limestone and lenticular masses of Gypsum. 2. Magnesian Limestone, pebbly in lower part. 1. Breccia, with dolomitic cement in upper part. The Breccia contains, besides other rocks, many pebbles of Carboni- ferous Limestone ; and in the upper part, where the cement has a dolomitic character, these are frequently converted into Potato-stones. The Limestone, according to Mr. Binney, is one mass of indistinct fossil shells, and is full of numerous small hollows lined with needle- shaped crystals of Calcite ; it contains, according to his analysis, about 77 per cent, of Carbonate of Lime and 11 '6 per cent, of Carbonate of Magnesia. The abundance of fossils in the rock makes it unlikely that it is one of those Limestones which were formed by precipitation from a solution of calcareous and magnesian salts ; it is most likely of organic origin, and was originally an ordinary Limestone. If this be true of the Limestone, it is highly probable that the Breccia contained at the time of its formation no magnesian matter. We may therefore look upon these two rocks as having been originally ordinary marine deposits. But the overlying Red Marl with Gypsum and Magnesian Limestone points to a change of conditions. These were probably formed in inland waters, perhaps by some such reaction as Sterry Hunt has suggested, whereby Gypsum and Carbonate of Magnesia were precipitated ; the latter was carried down in solution through the Description of Metamorphic Rocks. 411 underlying beds, slightly dolomitized the Limestone, giving rise to the drusy cavities, and penetrated some way into the Breccia, convert- ing the Limestone pebbles into Potato-stones. The fact that the lower part of the breccia does not appear to be altered seems to show that the water had been robbed of all its magnesian salt before reaching the bottom of the bed.* Dr. Sorby believes many of the Permian Magnesian Limestones, and also the magnesian portions of the Carboniferous and Devonian Limestones, to be altered Limestones. He says that in thin sections of these rocks fragments of the organic bodies of which they were com- posed may be sometimes detected, but that frequently the original mechanical structure has been entirely obliterated by the change. Such Dolomites would appear to be made up, in part if not wholly, of comminuted and decayed calcareous organisms, and to have been subsequently altered into Dolomite, possibly by the infiltration of Magnesian Salts of sea-water, when it had been so far concentrated that Rock-salt was deposited. In this case we may suppose that the water was for a time sufficiently free from dissolved matters to allow of the existence of animal life and the growth of organic Limestone, and that afterwards the area was flooded by a concentrated solution which transformed the Limestone into Dolomite by percolating through it. As a third means of explaining the origin of Dolomite, it has been suggested that water might remove from a Dolomitic Limestone the Carbonate of Lime, leaving the less soluble double carbonate behind. Some organisms do secrete and introduce into their hard parts Car- bonate of Magnesia as well as Carbonate of Lime, and it has been hinted that if an organic Limestone was formed of these remains and the two carbonates became united, the superfluous Carbonate of Lime might be removed and a Dolomite formed. The highest percentage of Carbonate of Magnesia known to exist in any hard organic structure is 7 '6 44 per cent., and therefore the application of this hypothesis to Limestone formed by animals at all approaching recent forms in this respect is quite inadmissible, on account of the enormous shrinking it would involve. Besides the Dolomitic rocks already mentioned, there are others interstratified with highly-metamorphosed rocks, such as Gneiss and Mica-schist, and containing crystals of Mica, Talc, and Quartz, which we can hardly look upon as anything but the products of alteration. By what means the change has been brought about is as yet uncertain. In dealing with those Magnesian Limestones which seem to have been formed by precipitation, we were obliged to admit that, though there was good reason to believe that they had a chemical origin, we were still quite in the dark as to the exact nature of the reactions by which they had been produced. Our position is very much the same with regard to the Dolomitic rocks we have just been treating of ; there are strong grounds for the opinion that they are altered Limestones, but how exactly the alteration was produced is still an open question. * For further particulars about these beds see Binney, Lit. and Phil. Soc. Manchester, 3rd series, ii. 374 ; Harkness, Quart. Journ. Geol. Soc. xx. 160. 4 1 2 Geology. Other forms of altered Limestone. The ready solubility of Calcium Carbonate has in many cases led to that salt being partially removed from Limestone and its place taken by other compounds. The following are instances. Dr. Sorby has shown that the Cleveland Ironstone was originally a Limestone, and that the Calcium Carbonate has been replaced by Ferrous Carbonate.* In a French Limestone where Phosphate of Lime occurs in pockets, the rock becomes dolomitic in the neighbourhood of the pockets. It is suggested that the deposi- tion of the Phosphate and the alteration of the surrounding rock may have been caused by the percolation of a solution of Magnesium Phosphate, t ^MgO.P0 5 + ?CaCO 3 = $CaO.PO 5 + $MgCO 3 . Gypsum. We must now say a word about those Gypsums which have probably been produced by the alteration of other rocks. There can be no doubt that Anhydrite, where it has been exposed to the action of the air or of percolating water, is converted into Gypsum. The process has been observed actually going on, and cases have been observed of masses of Sulphate of Lime which are com- posed of a coating of Gypsum wrapped over a nucleus of Anhydrite. Again, Limestone may be, and in many cases probably has been, converted into Gypsum. There are cases known, in the Alps and elsewhere, where a group of rocks contains at some spots great thick- nesses of Limestone, while at other spots the corresponding portions of the same group are composed of Gypsum ; the Limestone, in fact, is replaced by Gypsum. We can hardly suppose this to have been the original state of things ; it seems scarcely possible that marine organic deposits and chemical precipitates could have been formed thus closely side by side ; the more reasonable explanation is that the Gypsum is altered Limestone. The change may have been brought about in various ways. The action of sulphurous acid, which is given off from volcanic sources, would change any Limestone it came in contact with into Gypsum. It is doubtful whether vapour could make its way so thoroughly through the whole rock as to transform large masses ; but if converted into Sulphuric Acid it might be carried by water into the very heart of the rock. Sterry Hunt believes that this is the origin of certain masses of Gypsum in the Onodaga Salt Group of Canada ; springs are met with, in the district where the Gypsum occurs, containing free Sulphuric Acid, and he thinks that if this water came in contact with the Limestone of the group it would form a calcareous sulphate, nearly all of which, on account of its slight solubility, would be deposited in a crystalline form.^ He mentions that in one case Mr. * Geol. and Polytech. Soc. West Riding of Yorkshire, iii. (1856) 457 ; Quart. Journ. Geol. Soc. xxxv. (1879), Anniversary Address, p. 84. t Annales des Mines, xvi. (1879) 288. % On the Acid Springs and Gypsum Deposits of the Onodaga Salt Group, Silliman's Journ. 2nd series, viii. 175 (1849). In the Report on the Geology of Canada to 1863 (p. 352) it is stated that the Gypsums of the Onodaga group of Canada "seem to be contemporaneous with the Shales and Dolomites in which they are interstratified, and to have no connection with the springs of the present time." Sterry Hunt, however, maintains there are two Gypsums, one contem- Description of Metamorphic Rocks. 4 1 3 Murray observed a slender cylinder of Gypsum running up through several beds of Limestone, and extending into overlying Tertiary Clay. The conversion of Anhydrite or Limestone into Gypsum is attended by a considerable increase in bulk, and this circumstance has been applied to account for a puzzling occurrence often met with in the neighbourhood of large masses of Gypsum. Around such masses the overlying rocks are tilted up frequently at high angles and bent into an arch, while the rocks below lie perfectly flat. It has been suggested that, as the original rock swelled during its conversion into Gypsum, it bulged up into a boss and bent upwards the rocks that lay above.* The Canadian Gypsums just mentioned are stated by Sterry Hunt to occur in dome-shaped masses from 100 to 400 feet across, the overlying strata are tilted and wrap over the surface of the domes, while the beds underneath are undisturbed ; he also says that the ground rises in hillocks above the masses of Gypsum, and that houses are known in some cases to have been gradually raised by the elevation of the surface, arid masses of Gypsum have afterwards been found beneath them.t As another possible source of Gypsum, it has been pointed out that in volcanic districts Sulphuretted Hydrogen is given off, and by decomposing Silicates of Lime produces Gypsum and Sulphur. The theory of the metamorphic origin of Gypsum is however as yet in a very rudimentary state. For further hints the reader may refer to Bischof s Chemical Geology, chap. xix. ; Xaumann's Geog- nosie, i. 760 ; Zirkel's Petrol ogie, i. 268-273 ; Murchison, Quart. Journ. Geol. Soc. v. 172; Coquand, Bull, de la Soc. Geol. de la France, 2nd series, iv. 124 (1849); Roth, Allgemeine und Chemische Geologic, i. 89, 171, 204. (1 d) Carbonaceous Members. Graphite is the only rock coming under the head sufficiently common to be noticed here. It consists of Carbon with about 5 per cent, of impurities, such as Silica, Alumina, and Oxide of Iron. We have already had occasion to notice Graphite as a constituent or accessory mineral in several rocks. It also occurs in beds and lenticular masses among Schists, Crystalline Limestone, Gneiss, and other metamorphic strata. In general arrangement and microscopic struc- ture these layers of Graphite correspond frequently with Coal and some bituminous deposits, and there is every reason to believe that in many cases Graphite is a highly-metamorphosed Anthracite. Just as thick masses of pure Limestone may be looked upon as proofs poraneous with the rocks among which it occurs, the other now being formed in the manner explained in the text. Silliman's Journ. 2nd series, xxviii. 365 (1859). * Elie de Beaumont, Explication de la Carte Geologique de la France, ii. 69, 90. f LOG. cit. It is difficult to resist the notion that the latter part of the state- ment comes from some ingenious and kindly Yankee, good at a story and anxious to give Dr. Hunt a lift. 414 Geology. of the existence of animal life at the time of their formation, though no fossils may now be recognisable among them, the probability that Graphite was originally of vegetable origin is so great that its occur- rence makes it extremely likely that vegetable life existed during the period represented by the rocks in which it is met with.* (1 e) Altered Volcanic Ashes. The fragmental deposits associated with old lavas often retain so completely their clastic character, that their true origin is even now perfectly obvious. But they have occasionally undergone a consider- able amount of alteration, and it is then in some cases difficult to distinguish them from the lavas themselves. This is specially the case when they were originally composed of very fine dust : they are then transformed into flinty felsitic-looking rocks which bear the closest resemblance to true Felsites macroscopically. Even under the microscope they often show pseudo-fluxion structure, and are seen to have acquired more or less of true crystalline texture ; and they otherwise approximate to lavas. Weathering here sometimes brings out the clastic structure or reveals bedding, and thus indicates the true character of the rock. Sometimes such rocks can be traced in the field losing gradually their flinty character, becoming coarser and coarser, and at last passing into breccias or agglomerates. The micro- scope often shows clastic structure where it is quite unrecognisable by the unaided eye.t The fragments of which a fine ash is made up may be all broken crystals, and may therefore be all coloured when viewed between crossed Nicols, but when sufficiently enlarged they can often be seen to be fragments with ragged irregular edges, and can be distinguished from the more sharply-edged Microliths of a crypto- crystalline rock. 2nd Class. FOLIATED OK SCHISTOSE HOCKS. Nature of Foliation. Foliation, the structure which is the distinctive characteristic of the rocks we have next to consider, is defined by Professor Sedgwick^: to be "a separation of rock masses into crystalline layers of different mineral composition ; " and by Mr. D. Forbes is described as a parallel structure, which "makes its appearance in rock masses owing to the arrangement of certain crystal- lized minerals in more or less parallel lines, along which the crystals lie on their flat sides or lengthways, i.e. having their longer axes in the direction of, and not against, the grain of the rock." The first thing we notice about a foliated rock is a flaky structure or a tendency to split along planes rudely parallel to each other into leaves (folia) or laminae ; and further that this tendency is caused by the presence of * Dawson, Quart. Journ. Geol. Soc. xxvi. 112. t See the Geology of the Northern Part of the English Lake District (Memoirs of the Geological Survey), p. 24. J Paper already quoted on "Structure of Large Rock Masses." Popular Science Review (1870), p. 229. The last clause seems hardly wanted, because it is the arrangement of the crystals that gives the rock its grain. Description of Metamorphic Rocks. 4 1 5 roughly parallel layers running through the body of the rock composed in large measure of plates of a single mineral, such as Mica. But these characters alone do not constitute foliation ; a very similar struc- ture is found in some clastic rocks, such as micaceous Sandstone. The distinctive feature of foliated rocks consists in their mineral flakes being crystallized^ whereas in those derivative rocks which have a deceptive appearance of foliation the mineral plates are worn by attrition. It is the crystalline character of the constituent minerals, taken along with other peculiarities to be shortly noticed, which leads us to look on foliation as a superinduced structure. Classification of Foliated Rocks. Foliated rocks may be roughly divided into three groups, Schists, Gneisses, and Nodular Schists. An almost endless variety of rocks may be classed together as Schists from their possessing the following common characters. They consist very largely of Quartz, and the foliation is produced by rudely parallel layers of some other mineral such as Mica. A typical Schist contains no Felspar. Gneiss and its allies differ from the Schists in containing Felspar. The Nodular Schists are characterized by the presence of peculiar concretionary bodies arranged in rudely parallel planes. Schists. The following are the commonest varieties of the Schists. Mica-schist, Glimmer schiefer. Quartz and Mica separated more or less into alternate layers. The proportion of one mineral to the other varies almost indefinitely. The accessory minerals of Mica-schist are very numerous : Garnet is very common indeed, and Hornblende, Staurolite, Andalusite, and Cyanite are frequent. Argillaceous Mica-schist (Phyllite, Thonglimmerschiefer] is a rock intermediate between normal Mica-schist and Clay-slate, into both of which it passes by insensible gradations, according as the micaceous element becomes pronounced or gradually disappears. It might be termed either an imperfect Mica-schist or a foliated Clay-slate. Chiastolite-schist, Schiste made, Schiste maclifere. Argillaceous Mica-schist with crystals of Chiastolite disseminated through it. It is generally found where Clay-slate abuts on igneous rocks, and has been produced by the metamorphosing action of the latter on the former. There is a most instructive paper by Professor Fuchs describing the gradual growth by metamorphism of this rock in the Pyrenees, in Leonhard's Jahrbuch, 1872, p. 878, an abstract of which is given in p. 419. The presence of an allied mineral Staurolite, gives rise to Staurolite- scJtist ; the reader will do well to consult and compare with the paper just quoted one on the formation of this rock in the Geological Maga- zine, vol. x. 102. Talc-schist and Chlorite-schist. If the Mica in Mica-schist were mixed with or replaced by Talc or Chlorite we should have these rocks. Under their typical form they are usually poor in Silica ; sometimes they contain Felspar, and thereby pass into Protogine or Talcose Gneiss. Calcareous Mica-schist. Alternate layers of Mica-schist and Car- 416 Geology. bonate of Lime. It differs from the micaceous crystalline Limestones already mentioned in the presence of Quartz. Quartz-schist. When the quartzose element in Mica-schist becomes very large, we obtain this rock, which consists of compact, imperfectly- foliated white Quartz, foliated by thin parallel layers of Mica scales. If the Mica disappear, it passes into Quartzite. Felspathic Mica-schixt.Some Mica-schists contain Felspar, and form a transition between the normal type of that rock and Gneiss. The presence of accidental minerals gives rise to numerous varieties of the Schists just mentioned, which we cannot describe here. Tourmaline-schist or Schorl-schist. Mainly composed of Quartz and Schorl ; the foliation is produced by approximately parallel thin bands of Schorl. This rock, like Greisen, often contains Cassiterite, Topaz, Lithia-mica, Mispickel, arid other minerals. Just as Greisen is an altered Granite, Tourmaline-schist is a Clay-slate which has been metamorphosed in a similar manner. Where the Schorl is dissemi- nated irregularly so that there is no foliation, the rock is called Schorl or Tourmaline rock. Other Schists less common are Sericite-schist, Ottrelite-schist, Dipyr- schist, so called after minerals which specially characterize them. In some varieties of Schists the foliation is produced by metallic ores, such as micaceous Iron-glance, Zincblende, Iron or Copper Pyrites, Cobalt Ore, etc. Gneisses. Gneiss. A schistose aggregate of Felspar, Quartz, and Mica. The Felspar is in crystalline grains, the Quartz in grains or small lenticular discs, and through the mixture formed of these two minerals there run parallel layers or leaves of Mica, giving the rock a foliated structure. In some cases the Felspar and Quartz are to a certain extent separ- ated into distinct bands, but frequently these two minerals are intermixed and the Mica alone is collected into layers. Orthoclase is always present, but Triclinic Felspars occur as well. Microscopic cavities containing water and liquid Carbon Dioxide occur in the Quartz. Both macroscopically and microscopically Gneiss resembles Granite very closely in its mineral composition, the main difference between the two consisting in the presence of foliation. Garnet is a very common accessory mineral : Hornblende and Talc also occur, and they are sometimes present in such abundance as to give rise to the following varieties of the rock. Hornblendic Gneiss, in which the Mica is in part or wholly replaced by Hornblende; the Felspar is very largely Plagioclastic. When Hornblende is so abundant as to form the greater part of the rock, it is spoken of as Hornblende-schist: this form sometimes loses its schistose character, when it is known as Hornblende Rock. Protogine or Talcose Gneiss consists of Orthoclase, Plagioclase, Quartz, Mica, and a Talcose mineral ; it is sometimes spoken of as a variety of Granite, but the descriptions given of it seem to show that it has often more or less of a schistose structure. Graphitic Gneiss is a variety in which the Mica is wholly or in part replaced by Graphite. As the latter mineral is probably of vegetable Description of Metamorphic Rocks. 4 1 7 origin, this rock has most likely arisen from the metamorphism of sedimentary deposits containing carbonaceous matter.* Granitic Gneiss. When the foliation of Gneiss becomes indistinct the rock approximates lithologically to Granite. We not unfrequently meet with rocks of this vague character, and are hardly able to say whether they should be called Gneiss or Granite. Such intermediate forms are styled Granitic Gneiss. Granitic Gneiss sometimes loses its foliation entirely and passes into a rock which is lithologically Granite. These bedded Granites will be treated of among the third class of Metamorphic rocks. Granulite. The rocks described under this head vary a good deal in minor points but are all essentially composed of Quartz and Ortho- clase or perhaps Microcline with Garnets. In some varieties the Felspar constitutes a compact mass, while the Quartz occurs in flattened grains or very thin laminae running in parallel layers and producing a schistose structure. Garnets are disseminated irregularly. Schorl is a common accessory and Cyanite also occurs. Fluid- cavities have been observed in the Felspar of Granulite. The margins of many Granite masses are formed of a rock composed of the same minerals as Granulite, but without foliation. Serpentine. This rock may be conveniently noticed here. We have already seen that it has in many cases been produced by the alteration of Peridotite or some non-felspathic rock rich in Olivine. In these cases the alteration has been brought about by pseudomorphic changes rather than by what we usually understand by metamorphism. Serpentines formed in this way are, like the rocks from which they were derived, intrusive in many cases. But Serpentine also occurs distinctly bedded and intercalated among strata of crystalline Schists. Thus in the Laurentian gneissic rocks of Canada it is met with in bedded masses of great purity ; it occurs also associated with Lime- stone or Dolomite, sometimes in grains arranged in bands parallel to the stratification, and sometimes in veins traversing the rock.f In the same country a great thickness of bedded Serpentine occurs in the Quebec group of rocks ; J and Serpentine has also been found inter- bedded in thin layers with stratified Diorites at St. Stephen's. These bedded Serpentines are very frequently associated with Dolo- mite, and it is possible that they may be altered forms of that rock, but no satisfactory evidence of this transformation has yet been brought forward. Serpentine is also found with a laminated structure, and when the surfaces of the laminae are covered by plates of Mica or Chlorite it becomes a foliated rock. Such a rock passes into Talcose Schist, and is perhaps only a metamorphosed form of the latter. || Nodular and Spotted Schists. We may class together * See Geol. Mag. iv. 160, for an account of bituminous Gneiss and Mica-schist in Sweden ; and for a similar rock in the Vosges, Explication de la Carte Geo- logique de la France, i. 314. t See Report of Progress of the Geol. Survey of Canada (1863), pp. 471, 591. Ibid. p. 266. Ibid. (1870, 1871) p. 32. || See also Heddle, Transactions Royal Soc. Edinburgh, xxviii. chap, iv., and notice in Min. Mag. iii. 128. 2D 41 8 Geology. under this head a number of somewhat peculiar schistose rocks to which the Germans have given the names Fleckschiefer (Spotted or Blotchy Schists), Knotenschiefer (Knotty Schists), Fruchtschiefer (Grain-schists), Garbenschiefer (Sheaf -schists). The peculiarity in some is produced by dark spots or blotches ; in some small rounded concretions are present which stick up like knots on the planes of foliation ; in others the concretions are elongated and bear some resemblance to grains of corn ; in others again the concretions are collected into masses which resemble ears or sheaves of corn or are gathered into clusters.* In all cases spots and concretions show imperfectly-developed crystals of various minerals, Andalusite, Mica, and others. They may be called the embryos of crystals. These rocks are deserving of notice because in the field they can be traced passing on the one side into unaltered clastic rocks and on the other into more thoroughly crystalline foliated rocks. Both their position then between non-crystalline and highly-crystal- line beds, and their own imperfect degree of crystallization, show that they have been subjected to a degree of metamorphism small compared with that which the foliated Schists have undergone. They are the products in fact of an early stage of metamorphism, when the crystals were yet rudimentary. The foliated Schists are the products of a more advanced stage, when crystals had attained their full growth. How Foliation has been produced. We are led to the conclusion that in a very large number of instances foliation is an effect of metamorphism in more ways than one ; but the most con- vincing arguments are derived from a study of foliated rocks in the field followed up by microscopical and chemical examination. When we say that foliation is an effect of metamorphism we mean that the rock was not originally either crystalline or foliated ; that it was once a clastic derivative rock, and that it has been altered in such a way as to acquire crystalline texture and a foliated arrangement of its crystal- line minerals. The outdoor study of foliated rocks also leads us to the conclusion that the metamorphism which produced foliation was more intense than that which gave rise to altered rocks of the first class, such as Quartzite and Statuary Marble. One or two instances of observations leading us to the above belief will now be brought forward. Degrees Of Foliation. We may first notice that foliation varies very much in degree. Cleaved Clay-slate often shows an incipient foliation, the planes of cleavage being "coated over with Chlorite and semi-crystalline matter, which not only merely define the planes in ques- tion, but strike in parallel flakes through the whole mass of the rock."t Darwin mentions a case of what looks like arrested development of foliation in Terra del Fuego. " In several places I was particularly struck with the fact that the fine laminae of the Clay-slate, where cutting straight through the bands of stratification, and therefore indisputably true cleavage planes, differed slightly in their greyish and greenish tints of colour, in compactness, and in some laminae having a * v. Lasaulx, Neues Jahrbuch, 1872, p. 840. t Sedgwick, Transactions Geol. Soc. of London, iii. 471. Description of Metamorphic Rocks, 419 more jaspery appearance than others."* Had the process, of which we see in this instance the commencement, been carried on, the result would doubtless have been a foliated rock, with its flakes parallel to what are now planes of cleavage. From cases, such as those just mentioned, of rudimentary traces of this structure up to the most complete crystallization and parallel arrangement of the component minerals, all sorts of intermediate gradations exist, and may sometimes be observed melting impercep- tibly into one another in the same rock mass, as in the following case observed by Mr. Darwin at the Cape of Good Hope. A mass of Granite has there burst through Clay-slate. " At the distance of a quarter of a mile from the spot where the Granite appears on the beach (though probably the Granite is much nearer under ground) the Clay-slate becomes slightly more compact and crystalline. At a less distance some of the beds of Clay-slate are of a homogeneous texture, and obscurely striped with different zones of colour, whilst others are obscurely spotted (Fleckschiefer). Within a hundred yards of the first vein of Granite the Clay-slate consists of several varieties some compact with a tinge of purple, others glistening with numerous minute scales of Mica and imperfectly-crystallized Felspar, some obscurely granular, other porphyritic with small elongated spots of a soft white mineral (Fruchtschiefer). Close to the Granite the Clay- slate is changed into a dark-coloured laminated rock, having a granu- lar fracture, which is due to imperfect crystals of Felspar coated by minute brilliant scales of Mica." At the actual junction of the Granite and Clay-slate the latter was at one spot " converted into a fine-grained, perfectly-characterized Gneiss, composed of yellowish- brown granular Felspar, of abundant black brilliant Mica, and of few and thin laminae of Quartz." t The coloured striping, the micaceous laminae, and the elongated spots seem to point to an incipient foliation, which reaches its full development in the Gneiss immediately adjoin- ing the Granite, and this gradual growth of the structure clearly indicates a metamorphic origin. J Foliated Rocks of the Pyrenees. This instance, which has been carefully worked out by Professor Fuchs not only in the field but also by the help of the microscope and chemical analysis, is of the most convincing character. There are in the Pyrenees several detached masses of Granite, and around each of these there wraps a belt of altered rocks, the meta- morphism of which begins at the edge farthest from the Granite in traces which can be detected only by the most careful scrutiny, and in a general way increases as the Granite is approached. The following- are the principal steps in the gradual series of changes. * Geological Observations on South America, p. 155. t Geological Observations on Volcanic Islands, pp. 149, 150. % On the different stages of Foliation see also Kinahan, Royal Geol. Soc. of Ireland, June 1870. Die Alten Sediment-Formationen und ihre Metamorphose in den franzo- sischen Pyrenaeu, von Hern Professor C. W. C. Fuchs; Neues Jahrbuch, 1870, pp. 717, 851. See also Verharidlungen der k. k. Geologischen Reichsanstalt, 1869, p. 314. 42O Geology. Outside the zone where metamorphism first becomes apparent there is Clay-slate, which to the naked eye, and even under a pocket lens, seems perfectly homogeneous ; magnified 400 times certain dark points, indistinctly outlined, make their appearance, and under a power of 900 the rock resolves itself into an interlacing network of fine crystals of Quartz, Mica, Chlorite, and a little Magnetite. Next to this rock comes a Clay-slate in which the dark spots become more readily distinguishable (Fleckschiefer), but even here they are so small that they would not be noticed with the eye alone, were it not that their dull colour contrasts with the lustrous brilliancy of the rest of the rock ; when this rock is highly magnified the little spots are seen to be concretions, and Quartz and Mica are clearly distinguishable in the body of the rock. As we get further into the metamorphic zone the small concretions increase in size, and become more distinctly outlined, till the rock becomes a true Nodular Schist (Knotenschiefer or Frucht- schiefer), at the same time the Quartz and Mica are more distinctly developed, so that the rock passes into a form intermediate between Clay-slate and Mica-schist to which it is not possible to assign a definite name. The concretions continue to increase in number and size, till they at last assume the form of dark prismatic bodies, which are crystals of Andalusite. As we advance still further into the heart of the metamorphic region, the rock assumes more and more the form of a mixture of Quartz and Mica, and a curious change is noticed in the concretions and Andalusite crystals : their outline becomes indistinct, Mica makes its appearance in them, and at last they fade gradually away and are replaced by the latter mineral. The rock thus passes into a typical Mica-schist. As soon as the change has been effected Felspar begins to make its appearance, at first sparingly, so that it is no more than an accidental constituent, but it becomes step by step more plentiful, till the Mica -schist at last puts on the form of a foliated compound of Quartz, Mica, and Felspar, that is, passes into Gneiss ; and in many cases it is impossible to say where one rock ends and the other begins. The earliest appearances of Gneiss are finely grained and poor in Quartz, but as we approach the Granite the rock becomes coarser and the Quartz more plentiful. Lastly the Gneiss loses by degrees its schistose structure, and passes by the most imperceptible gradations into true Granite ; some of the intermediate varieties, which partake in a manner of the characters of both rocks, have been called Granitic Gneiss.* As a rule the degree of alteration increases steadily as we approach the Granite, but this is not universally the case. Occasionally beds actually in contact with Granite are less highly altered than others farther off. In such cases the amount of metamorphism has probably been determined by something in the original constitution of the rock, * Gradual changes of a similar character and occurring in a similar order have been observed in other instances. The case of the metamorphic rocks of Skiddaw and Carrock Fell is strikingly parallel up to a certain point. It differs only in the final term ; the Granite is there intrusive. See Ward, Quart. Journ. Geol. Soc. xxxii. 1. See also Jukes, Student's Manual of Geology, 3rd ed. p. 366. Description of MetamorpJiic Rocks. 421 which has caused some beds to be more easily metamorphosed than others. Instances like those just given could be multiplied without end,* and they leave no room for any conclusion except the one already mentioned, that the crystalline texture and foliation of the rocks we are considering have been produced by metamorphic change after their formation, and that a high degree of metamorphism was necessary to produce them. Other facts lead us to the same result. Thus in the Alps beds of sandy calcareous composition containing fossils are inter- stratified w r itli Mica-schist and Gneiss. In some cases Gneiss is the result of a higher degree of metamorphism than Mica-schist ; but the original composition has in other cases determined whether it was Mica-schist or Gneiss, or what form of foliated rock it was into which a rock passed. What determines the Planes of Foliation. In some cases it seems highly probable that foliation, even if it does not exactly follow, is closely related to planes of original sedimentary deposition. This appears to be the case with the great mass of schistose rocks that make up the northern highlands of Scotland, t The same has been observed to be the case in Anglesea,^ in Arran, in Ireland, || in Norway, H and elsewhere. Fig. 127, which is copied from Professor Ramsay's IV\ { 1\| Memoir, shows a case in point. He l\\\l\l says, "The beds consist of very hard Vu quartzose grits intermingled with schis- tose bands, and they seemed partly foliated and partly in lines without clear ,-,. " ,,,.,. T r , J , , , , T ., Fig. 12 1. FOLIATION i lohation. In the sandy beds No. 1, T0 BEDDING. marked with dots, I saw no separation of distinct layers of different mineral substances, such as would be called foliation in the sense in which it is used by Mr. Darwin, but only a sort of imperfect separation of material sometimes arranged in wavy lines ; while in the schistose beds complete foliation appears, with much twisting of quartzose and micaceous laminae, each bed however still retaining its identity. In No. 2 these contorted laminae follow the direction of the bed, and in Nos. 3, 4, and 5 they cross the beds, somewhat in the direction of the lines in the sandy beds, but, eliminating the contortions, at slightly different angles" (pp. 181, 182). * The student should read most carefully Dr. Sorby's observations on this head, Quart. Journ. Geol. Soc. xxxvi. (1880), Anniversary Address pp. 81-92 : and he may consult Allport, ibid, xxxiii. (1877) 407, 419 ; and Memoirs of the Geological Survey of Ireland, Explanation of Sheets 147 and 157, p. 8. t Murchison and Geikie, Quart. Journ. Geol. Soc. of London, xvii. 171. Ramsay, the Geology of North "Wales (Memoirs of the Geological Survey of England and Wales), pp. 177-188. Ramsay, Geology of Arran, pp. 88 and 89. || Kinalian, Royal Geol. Soc. of Ireland, January 10, 1866. IF D. Forbes, Popular Science Review (1870), p. 235. PAHALLEL 4 22 Geology. In considering this section we must first suppose the foliated laminae straightened out, because, as we shall shortly see, their crumpling was probably produced subsequently to the foliation. We must also bear in mind that the thicker lines in the figure are unquestionably the edges of bedding planes. In the case of No. 2 the foliation is parallel to these bedding planes, and therefore in all likelihood took place over the faces of laminae of regular and even deposition. In the other cases the folia- tion crosses the main lines of bedding, but it has a general parallelism to the subordinate layers into which the beds No. 1 are divided, and these latter have a singular resemblance to planes of current lamination. A suspicion therefore crosses the mind that here foliation has taken place, as suggested by D-r. Sorby, on planes of current bedding. The question whether foliation has ever taken place along bedding planes, though much may be said in favour of an affirmative answer, may perhaps be an open one ; but the observations of Mr. Darwin and others have demonstrated beyond doubt that in many cases the planes of foliation coincide with those of cleavage. Besides those instances of incipient foliation already given, in which this is the case, Mr. Darwin found in South America immense tracts of intensely metamor- phosed Schists, the folia of which had every one of the distinguishing characters of cleavage laminaB : their strike was uniform over large areas and parallel to the leading physical features of the country, and in all cases which he saw, where masses of cleaved and foliated rocks alternate together, the cleavage and foliation were parallel. He sums up the evidence thus : " Seeing then that foliated Schists indisputably are sometimes produced by the metamorphosis of homogeneous fissile rocks ; seeing that foliation and cleavage are so closely analogous in the several above-enumerated respects; seeing that some fissile and almost homogeneous rocks show incipient mineralogies! changes along the planes of cleavage, and that other rocks with a fissile structure alternate with and pass into varieties with a foliated structure ; I can- not doubt that in most * cases foliation and cleavage are parts of the same process : in cleavage there being only an incipient separation of the constituent minerals, in foliation a much more complete separation and crystallization." | It is perhaps scarcely the case that cleavage necessarily involves a separation of the constituent minerals, but it is certainly true that if foliation begin in a cleaved rock, the separation follows the planes of cleavage. The coincidence of the planes of cleavage and foliation is therefore probably accidental, but it is an accident that will be of very frequent occurrence. In fact this coin- cidence is so common that it is unnecessary to multiply examples here. Artificial Production of Cleavage-Foliation. Coincident cleavage and foliation have been produced artificially by Dr. Sorby and the late Mr. David Forbes. The first mixed a quantity of scales of Oxide of Iron with Pipe-clay so that they were distributed indis- criminately through the mass. After submitting the mixture to pressure, the scales were found to have arranged themselves in rudely parallel planes perpendicular to the pressure, along which the mass admitted * I would with deference suggest that for "in most cases" it would be safer to say, " in those cases where such an accumulation of evidence is met with." t Geological Observations in South America, pp. 162 and sequent. Description of Metamorphic Rocks. 423 of ready division into thin plates. Mr. Forbes exposed amorphous ISoapstone to a moderate heat not exceeding redness for some months, under a pressure of from seven to twelve pounds per square inch, and obtained an aggregate of finely-developed crystalline foliae of a brilliant white or greenish colour, identical with Talc ; in fact a Talcose Schist. Under similar circumstances Clay-slates were converted into rocks pos- sessing a beautiful parallel structure, closely resembling Gneiss.* Crumpled Laminae. In highly-foliated rocks the laminae are not plane, but wrinkled and contorted, sometimes gnarled and crumpled to an intense degree. An instance is shown in fig. 128, copied from one of the " Reports on the Geology of Canada." Fig. 128. CRUMPLED LAMINAE OF FOLIATION. a. Evenly-laminated hornblendic Gneiss. b. Crystalline Limestone with thin corrugated bands of Gneiss. Let //' hk'k Professor Ramsay has given the following explanation of this. abj cd, ef, gh, ik, in fig. 129, be the edges of planes of bedding, and the finer lines crossing these the edges of planes of cleavage, and suppose foliation to have been produced over the latter. Then suppose that subsequently the rock mass suffered compression in the direction of the arrows, so that the beds edef, gliik, be- come squeezed into the thinner beds cde'f, ghik'; the planes of cleavage-foliation crossing these beds, having now to pack them- selves into a narrower space, must become crumpled up into some such wavy lines as are shown in the figure, t Foliation is not confined to the metamorphic representatives of the * See also Daubree, Etudes sur le Metamorphisme, part. iii. chap. x. t The Geology of North Wales (Memoirs of the Geological Survey of Great Britain, vol. iii.), p. 188. For another explanation see Stapff, Neues Jahrbuch, 1882, vol. i. Abhand. p. 90. Fig. 129. 424 Geology. derivative rocks ; igneous rocks, both subaerial and deep-seated, and volcanic ashes and tuffs, where they have been subjected to the necessary conditions, exhibit this structure. Thus a ribboned or banded Trachyte would be eminently suited for its production, or an ash in which induration and cleavage have been brought about by pressure. Intrusive Schistose Rocks. The evidence already given would seem to show that in many cases crystalline Schists have undoubtedly been formed by the metamorphism of derivative rocks. Where a gradual passage from one into the other can be distinctly traced, we can scarcely come to any other conclusion ; and even in those cases where no such transition can be observed, we should be apt to infer from analogy that what \vas true of some schistose rocks was true of all, and that in every case they might safely be looked upon as metamorphosed sedimentary deposits. Whether such an inference would hold good is however more than doubtful. It is far from unlikely that some schistose rocks are truly intrusive, and have burst through the strata among which they occur in a fused or plastic state. Mr. Scrope has some very suggestive remarks (Volcanoes, pp. 140 and 300) on the analogy between foliation and the banded or ribboned structure of certain Trachytes. In the case of these rocks the unequal motion of different parts of a lava stream, as they dragged one over the other, has given rise to a laminated structure in the cooled rock and a platy arrangement of flattened crystals through the mass, 'and he thinks that just the same result may have been brought about in a body of fused Granite forced up under pressure, and so a foliated Gneiss may have been produced.* The subject has also been taken up by M. Daubree, and he has satisfactorily proved by experiment that when a plastic mass of suitable composition is forcibly driven through an orifice, the pressure exerted by the walls of the aperture will give rise to foliation. In his experiments, clay, with which sand or plates of Mica had been mixed, was placed in a strong vessel, and then squeezed out by powerful pressure through openings of various shapes in the lid. The prisms that emerged were distinctly cleaved, the planes of cleavage being perpendicular to the walls of the orifice, and at the same time a true foliation was set up by the sand or Mica ranging itself in layers over the cleavage planes. There is therefore nothing inconsistent in the notion that some foliated rocks have been intrusive ; but M. Daubree has gone further than this, and has shown that on such a supposition we can explain the origin of that peculiar arrangement in the planes of foliation which has been so often observed to obtain in the great schistose masses that form the cores of mountain-chains, and which the metamorphic theory accounts for only imperfectly. The arrangement in question, known as Fanlike Structure or Structure en Eventail, is as follows. In the centre the planes of foliation are vertical, but as we recede on either side from the axis of the chain they become more and more inclined, * See also Jiuld, Geol. Mag. [2] ii. (1875) 68. Description of Metamorphic Rocks. 425 dipping on each flank inwards towards the axis, so that in a transverse section their edges are disposed like the plaits of an opened fan. In order to imitate artificially the conditions which he supposed to have given rise to this structure, M. Daubree placed a square prism of clay between two iron plates of the same size as the faces of the prism, and subjected it to pressure. Sheets of clay were squeezed out from the four free faces of the prism, and these, as they extended themselves beyond the range of the pressure, gradually expanded in thickness till they as- sumed the forms of truncated wedges with their thicker ends outwards. The whole mass was found on examination to be foliated, and the planes of foliation presented a true fan-shaped arrangement ; in the portion between the plates they were parallel to the faces of the plates, in the truncated wedges outside the plates the planes of foliation opened out till they became parallel to the faces of the wedge. Exactly the same results would follow if the materials of a mass of Gneiss were extruded from below through a fissure in the earth's crust ; low down in the rent, where the pressure of the walls was approximately horizontal, the foliation planes would take a vertical position ; where the fissure in its upper part began to gape, these planes would bend so as to keep parallel to the diverging faces of the fissure, and would open out into the fan-shaped form which they actually assume.* Summary. In the present state of our knowledge it will perhaps not be safe to say more than this on the obscure subject of foliation. That rocks have been subjected to some form of metamorphic action which has set free their constituent minerals to move among them- selves, which has dissolved or melted these minerals, or in some way given them the power of assuming tabular crystalline forms ; which perhaps has decomposed them and allowed of the formation of new compounds out of their elements. That the minerals resulting from this action have separated themselves out from the body of the rock, and arranged themselves more or less in distinct parallel layers. That the process of alteration has not been carried far enough to efface those great planes of division, be they bedding or cleavage or any others, by which the rock was traversed when foliation began. That the segre- gation or separation took place along those planes which offered the Least resistance to tJie motion of the constituent particles, or to the passage of those agents which assisted in or produced the foliation ; so that if foliation took place in a bedded rock, before cleavage had been produced, the laminse would have a tendency to be parallel to the bedding ; but if cleavage had sealed up the bedding and opened out another set of divisional planes, it would be parallel to those that the foliage would range ; in the same way, if the rock contained nodules or concretions, the foliation would be turned out of its way by these and bend round them.t Lastly, we must not lose sight of the probability that some schistose * Daubree, Experiences sur la Schistosite des Eoches. Comptes Rendus, t. Ixxxii. 27 mars et 10 avril 1876 ; and Geologic Experimentale, i. 433. t See Professor Ramsay, Geology of North Wales, loc. cit. ; Kinahan, Dublin, Quart. Journ. of Science, vi. 185 ; Royal Geol. Soc. of Ireland, January 10, 1866, February 8, 1871. D. Sharpe, Phil. Transactions, 1852, p. 445 ; Quart. Journ. Geol. Soc. iii. (1847) 87. 4-6 Geology. rocks have been driven upwards in a fused or pasty state through rents, and that, as the mass rose, cleavage and foliation were simultaneously produced by the pressure exerted by the walls of the fissures. It is hardly possible that this can have been the cause which produced the foliation of those great bodies of schistose rocks which cover hundreds of square miles of country. Pressure could never be transmitted unimpaired through thicknesses of pasty rock so vast as these, and in their case we can in the present state of our know- ledge only look upon foliation as a structure which has been set up by widespread regional metamorphism. But in the case of the schistose nuclei of mountain-chains which exhibit the fanlike structure, the explanation comes in handily enough. We must however guard the reader from supposing that even in these cases it was the intrusion of molten or pasty matter that caused the upheaval of the mountain-chain. We shall see in Chapter XIII. that this was caused by thick masses of strata being squeezed by powerful horizontal pressure till long belts of rocks were forced to bulge up above the general surrounding level ; we shall also find that where the pressure has been most energetic the metamorphism is most intense ; and we shall come to look upon this compression as the cause which at once raised the mountain-chain, reduced the rocks in the centre to a pasty state, and squeezed them up through fissures, giving them as they rose foliation and fan-shaped structure. Early Theories about Crystalline Schists. In the early days of geological speculation, the crystalline Schists were not supposed to owe their peculiarities of structure to metamorphism, but were imagined to have been formed all of them at the same time, and pretty much as we see them now, during a very early period of the world's history, when conditions obtained very different from any that now exist. The constituents of these rocks were supposed to have been held in solution in an ocean of boiling water, and to have been precipi- tated as it cooled. A lurking fondness for this hypothesis seems still to linger in the minds of some eminent geologists. While they are driven by the overwhelming nature of the evidence to admit that there are many cases of schistose rocks which are metamorphosed sedimentary deposits, and that the process which gave rise to them has operated at many different times, they still think it possible, or even likely, that in some of those cases, where there is a very large thick- ness and extent of crystalline Schists, where no proofs of mechanical origin can be detected in them, and where indications of the existence of life at the time of their formation, either in the shape of fossils or limestones, are wanting, we may have rocks which owe their crystal- line structure, wholly or in part, to chemical precipitation from an ocean, the like of which can have existed only during the passage of the earth from some pre-existing state to its present condition. We shall see presently that in all likelihood the earth has passed through some such stage as this hypothesis requires, and it may be that during such a period rocks analogous to the crystalline Schists were produced ; though, as we have no experience to guide us, we can do no more than form vague conjectures as to what would happen Description of Metamorphic Rocks. 427 under such conditions ; but assuming that such was the case, there seems no necessity for the opinion that any of the rocks of this class now in existence date from so remote a period. When we have two masses of Gneiss differing from one another in no essential respect whatever, and when we know that one has been produced by metamorphism, it certainly seems more reasonable to believe that the other was produced in the same way, till some good ground can be shown to the contrary, than to go out of one's way to invent some other purely imaginary method by which it might have been formed, Reasoning from analogy is not necessarily conclusive, but it is safer than reasoning based on dreamy conjecture. The most telling point against such hypotheses however is the fact that the oldest rocks we know, the Laurentian formation of Canada, are crystalline Schists, which are conclusively proved to be meta- morphic by the traces they still present of mechanical origin and by the presence in them of what are almost certainly fossils. There may be rocks older than these, but the burden of proof rests with those who assert this to be the case, and proofs have not yet been forth- coming. 3rd Class. AMORPHOUS METAMORPHIC ROCKS. General Description. The rocks which we shall group to- gether under this head are quite indistinguishable, by the naked eye at least, from many members of the Plutonic* class as long as we study them by means of hand specimens alone. They are made up of the same minerals, they possess the same compact and generally largely crystalline texture as Plutonic rocks, and like them they are in no part vesicular, slaggy, or glassy. The resemblance extends even to minuter points ; fluid-cavities for instance abound in their crystals ; and it would seem in many cases that even the microscope fails to reveal characters by which we can distinguish between the two. In certain instances indeed microscopists believe that they have succeeded in detecting peculiarities which characterize these rocks and enable us to separate them from the irruptive rocks that constitute the Plutonic class ; but even those geologists who would be the first to hail such a discovery with delight are fain to admit that these points of difference are minute, recondite, and of a very vague nature ; and even if we admit their validity, it seems questionable whether they are of more than local application, t Lithologically then the rocks of this class are identical in many cases, perhaps in all, with certain of the Plutonic rocks ; for a long time no difference was recognised between the two ; and it is only of compar- atively late years that they have been removed from the Plutonic class. But when these rocks are studied on a large scale in the field, they * The reader must bear in mind that in the sense in which we have used the word all Plutonic rocks are irruptive. t Zirkel has pointed out differences between the Metamorphic and Intrusive Granites of the "Western Territories of North America, but he carefully guards himself against any appearance of extending his conclusions to other districts. Geological Exploration of 40th Parallel, vi. 58 and 59. See also the same work, i. 100. 428 Geology. are found to differ in one important respect from those rocks which we have classed as Plutonic. Masses of Plutonic rock are marked off by a hard and fast boundary from the rocks that adjoin them, and almost invariably behave in- trusively, thrusting out into the neighbouring beds dykes, tongues, and veins : they are all irruptive. Amorphous metamorphic rocks are never intrusive; they melt away by the most insensible gradations into the rocks that surround them. Modes of Occurrence. Amorphous metamorphic rocks occur under their most typical form in masses, and then pass gradually along their margins into some form of foliated rock ; this foliated rock in its turn shades away into less highly metamorphosed beds ; in these again the metamorphism grows step by step less and less till we reach unaltered clastic strata. There is a central crystalline mass encircled by belts of strata, each of which shows as we proceed outwards a smaller and smaller degree of alteration till we at last reach rocks perfectly unchanged ; and these belts are not separated by hard lines, but each melts imperceptibly into the belt on either side of it. Only one conclusion can be drawn from these facts. Just as foliated rocks are the products of a higher degree of metamorphism than such rocks as Quartzite and Statuary Marble, so the rocks we are now dealing with must have been formed by a degree of metamorphism more intense still. The metamorphism which gave rise to Gneiss for instance was not sufficient to destroy the bedding : the higher degree of metamorphism which produced rocks of the present class has effaced the bedding and formed an amorphous crystalline mass. The Granites of the Pyrenees, described on p. 419, are instances of masses of the present class, and this case is specially valuable because Professor Fuchs in investigating it did not rely solely on outdoor examination and the macroscopic characters of the rocks, but availed himself of the more refined tests which the microscope and chemical analysis have placed within the reach of the geologist. Many of these massive examples are Granites lithologically, and hence the name Granitoid is sometimes applied to the whole class. Again rocks of a kindred nature which may be conveniently treated of along with those just described occur in beds interstratified with foliated and other forms of metamorphic rocks. Their bedding prevents the term " amorphous " being strictly applicable to them, they form rather a connecting-link between Foliated and Amorphous Metamorphic rocks. Good examples are furnished by the bedded Granites of Donegal described on p. 401, and instances also occur among the metamorphic beds that surround the granitic masses of the Pyrenees (p. 419). Such bedded Granites differ from Granitic Gneiss only in the total absence of foliation ; passages occur from Gneiss into Granitic Gneiss and from that into bedded Granite ; and it seems likely that the metamorphism which produced Gneiss would if carried a step further give us Granitic Gneiss, and in the next stage Bedded Granite. Analogy leads to the belief that all the rocks of this division correspond to a very advanced stage of metamorphism. The reason why certain beds of a group have been more intensely altered than the generality of the rocks among Description of Metamorphic Rocks. 429 which they occur, is probably because in their original state their composition was such as rendered them specially susceptible of meta- morphic change. This is spoken of as Selective, Metamorphism. Relation to Plutonic Rocks. It has been necessary to lay special stress on the strongly-marked line of demarcation which in many cases separates the non-irruptive Amorphous Metamorphic rocks from the irruptive Plutonic rocks ; and unless a word of caution were given, the reader might be led by the words we have used to believe that there was always a great gap between the two classes. We will therefore without delay warn him not to come to this conclusion, for he will before long see that in all probability the very reverse is the case. Here as everywhere classification is merely a device to help the shortness of our memory and the feebleness of our intellect ; nature laughs at our attempts to parcel out the products of her handiwork into compartments ; we go on smoothly for a while and then she brings forward something that will not go into any of our neatly-arranged pigeon-holes, and the sharp boundaries between our classes which we began by fondly believing in are ruthlessly broken through. The present case is no exception to this rule. We will now give some additional instances of the two forms under which Amorphous Metamorphic rocks occur. Claystone of Chili. Nearly fifty years ago the notion that there were rocks, Plutonic as far as lithological characters go, which were really only the products of intense metamorphism, presented itself to Dr. Darwin. He thus describes a deposit to which he gives the name of Porphyritic Claystone Conglomerate, which occurs in Patagonia and covers large areas in Chili. " The formation, which I call Porphyritic Conglomerate, is the most important and most developed in Chili. From a great number of sec- tions I find it to be a true coarse Conglomerate or Breccia, which passes by every step, in slow gradations, into a fine Claystone Porphyry \ the pebbles and cement becoming porphyritic, till at last all is blended in one compact rock. The Porphyries are exceedingly abundant in this chain, and I feel sure that at least four-fifths of them have been thus produced from sedimentary beds in situ. There are also Porphyries, which have been injected from below among the strata, and others ejected, which have flowed in streams ; and I could show specimens of this rock, produced by these three methods, which cannot be distin- guished."* Halleflinta. A very instructive rock in connection with the present subject is Hdlleflinta or Felsitic Schist."^ It is in character intermediate between Felstorie and Gneiss, is foliated or unevenly laminated, and sometimes contains an admixture of Chlorite and occasionally some Mica. It occurs in Sweden interbedded with Gneiss and Granulite, into which it passes insensibly. It is therefore meta- * Letter from Mr. Darwin to Professor Henslow, December 1835. Printed for private distribution among the members of the Cambridge Philosophical Society. For details see Geological Observations on South America, pp. 148, 149, 169. t Neues Jahrbuch fur Mineral. 1874, p. 140. 43 Geology. morpliic, and in all probability is the result of the alteration of beds more early metamorphosed than the surrounding strata ; so that while they have been altered only so far as to become Gneiss, it has had a narrow escape of being fused altogether and converted into something indistinguishable from an irruptive Felstone. Rocks of Carrick in Ayrshire. In this district there occurs in the midst of a country composed mainly of hard Felspathic Sand- stones with occasional beds of Limestone, a tract of rocks so exactly like intrusive igneous rocks in look, composition, and general character, that they were for long referred without hesitation to that class. A careful examination however has shown that they have been produced by the metamorphism of the surrounding strata. Dr. James Geikie has traced a gradual passage from unaltered Sandstones, through forms in which an amygdaloidal texture begins to be developed, up to porphyritic and closely-grained Felstones, which, judged by look and mineral character alone, could not be distinguished from intrusive members of the Felspathic class of Crystalline rocks ; and he has even detected here and there, in the very heart of these apparently intrusive tracts, areas of unaltered rock gradually shading off into the Crystal- line rocks which surround them. In other cases the same process, acting probably on beds richer in Magnesia, has given rise to Dioritic rocks which approach quite as closely intrusive rocks of basic composi- tion. * Priestlaw. Granite, we have already seen, is in many cases an intrusive rock of Plutonic origin, but Granites also occur which are the result of extreme metamorphism, and no better instance of such a rock occurring in the shape of a mass can be found than that of Priest- law in Berwickshire.! This is a triangular mass, about one square mile in extent, surrounded on all sides by Felspathic Sandstones and Shales. The rock of which it is composed varies in texture and grain, but is for the most part a well-marked Granite, made up of Felspar, Quartz, and Mica, with occasional Hornblende. Professor A. Geikie has described with great care a section, which starts about a mile from the hill, and shows a series of gradual changes in the surround- ing beds, which terminate in Granite. He notes first that the strata become exceedingly fine grained and compact, ringing with a metallic sound when struck ; the Sandstones however still retain a granular texture, and the Shales, even when almost converted into Jasper, still show their fissile structure along a weathered face. The next change noticed is the appearance of a number of veins, beds, or dykes of rock indistinguishable from Felstone. These are not, as might be supposed, intrusive masses, they shade off so imperceptibly into the Sandstone adjoining them, that they are evidently metamorphosed portions of the latter. One of these crystalline beds has a base of pink crystalline Felspar, with scattered specks of black Mica and Hornblende; in another case, a bed of * Quart. Journ. Geol. Soc. xxii. 513. t Flayfair, "Illustrations," Works (1822), i. 328; The Geology of East Lothian (Memoirs of the Geological Survey of Scotland), p. 15. For another similar instance see the Geology of North Berwickshire, p. 29. Description of Metamorphic Rocks. 431 exceedingly hard Sandstone contains granules of dark vitreous Quartz, and is so extremely altered that it might readily pass for a Felstone, were it not that its bedded structure is still distinct. The number of these crystalline masses and the intensity of the alteration continue to increase, till we at last reach rocks of which it is hard to say whether they are to be called Sandstones or Felstones ; then follows a rock with much of the character of a Sandstone, but which soon passes into an undoubted salmon-coloured Felstone ; this becomes again more finely crystalline, until it once more resembles Sandstone. It is here composed of Felspar, Quartz, and Hornblende, with Mica. To this compound immediately succeeds by a rapid increase in the size of the crystals the true Granite of Priestlaw. This section shows beyond doubt that the Granite occupies a centre from w T hich metamorphic action extended itself around among the adjoining rocks. The question arises, Is the Granite itself only the final step in the series of changes by which the surrounding beds have been rendered more and more crystalline ; or is it a mass, intruded in a molten state, from which heat has spread outwards, and brought about these changes ? There is much against the latter vie\v. The alteration produced by intrusive igneous masses seldom extends so far as in the present case into the adjoining rocks. But what tells most strongly against the second explanation is the almost insensible grada- tions by which the passage is effected from slightly-altered Sandstones and Shales into Granite. An intrusive mass would produce alteration in the surrounding rocks ; but there would still be a line of demarcation between the irruptive and the altered rocks unless the latter had been converted along its inner margin into a substance exactly identical with the former. This is not a thing very likely to happen, but it is only in this way that the gradual melting away of the more highly crystalline into the less altered rocks could be made as perfect as it is in the present instance. A case like this, then, shows that Granite can be produced by the metamorphism of rocks in situ, for some at least of the Priestlaw Granite must have originated in this way. Thus much we must admit ; and if we admit this, there is no ground for refusing to believe that the whole could have been formed in the same mariner. South-west of Scotland. Dr. J. Geikie has described some Granites in the south-west of Scotland which seem to have had a similar origin to that of Priestlaw. The mass of the country is com- posed of Felspathic Sandstones interbanded with occasional beds and broad belts of Shales and Mudstones. The former approach Granite in composition more nearly than the latter, and we might therefore expect that if a group of alternations of such beds became converted into Granite, the transformation would be carried to a larger extent in the Sandstones than in the Shales. This is found to be the case. There are certain patches of Granite to the north-east of Loch Doon, which make their appearance in broad bands of vertical Felspathic Sandstones flanked on either side by hard flinty Shales. The Shales are finely crystalline along their line of junction with the Granite, but 432 Geology. the metamorphism quickly ceases as we recede from the Granite along the trend of a belt of Shale ; on the other hand, when we proceed along a band of the Sandstone we find alteration extending to a much greater distance. Where the hard slaty Shales impinge on the Granite, we have no difficulty in laying our h'nger upon the line which separates one rock from the other ; but at the point where the Granite and the Sandstones come together, the union of the two rocks is so intimate that we hav r e usually no line of demarcation, but, on the contrary, a gradual passage.* The same author, in describing a similar tract of Granite and Minette in the same neighbourhood, which he believes owe their present form to metamorphic action, says : " The metamorphism of these rocks has been deduced from a variety of considerations. The chemical com- position of the unaltered strata and the Crystalline rocks is similar, and distinct passages can be traced from granular and slightly-altered Felspathic Sandstones through masses of various textures (the main constituent being Felspar, with Quartz and Mica more or less abun- dant in places) into Crystalline rocks, such as Minette and Granite. When the relation of these crystalline masses to the surrounding unaltered Sandstones is considered, the metamorphic character of the former becomes still more apparent. The Sandstones are not broken through and violently displaced, nor is there any appearance of con- fusion, as the centres of greater metamorphism are approached. On the contrary, the dip and strike of the strata continue unchanged and perfectly distinct until the rocks begin to assume a 'baked' and semi- crystalline texture, and the bedding gradually becomes obscure, and at last vanishes altogether. But after the metamorphic area is traversed, and the unaltered strata on the farther side are reached, the Felspathic Sandstones again appear with exactly the same dip and strike, giving no evidence of disruption by great masses of igneous rock. It seems reasonable to conclude therefore that the Felspathic Sandstones were once continuous across the area now occupied by Crystalline rocks, and that these Crystalline rocks have not been erupted from below, but are in truth only the Felspathic Sandstones under a different form. The Sandstones in this area have been simply metamorphosed into Crystalline rocks : they have changed their texture while retaining the same general composition, "t Bedded Granites of Brittany. We have already had instances of a rock which is lithologically Granite occurring in beds (pp. 401, 420), a similar case is met with in Brittany. J In that pro- vince two very distinct forms of Granite occur. One forms the flanks of the hill-ranges. It is finely grained, and contains interstratified beds of Mica-schist, Common Gneiss, Granitic and Talcose Gneiss, into which * Geol. Mag. iii. 529. t Memoirs of the Geological Survey of Scotland, Explanation of Sheet 22, par. 10. For a case in the Vosges in which a passage can be traced from Granite through Gneiss into Mica-schist see Explication de la Carte Geologique de la France, i. 307 and 327 ; and for an instance of a passage from Granite into Clay- slate, Explanation of Sheet 70 of the Geological Survey Map of Ireland, 11. Explication de la Carte Geologique de la France, i. 192; Geol. Mag. [1] x. 102. Description of Metamorphic Rocks. 433 it passes so insensibly that it is impossible to define a boundary between the two. On the borders of the Granitic areas the Granite is flanked by Gneiss into which it passes imperceptibly, and the latter shades off through Mica-schist into broad tracts of Clay-slate. As we approach the Granite, crystals of Staurolite begin to make their appearance, and become more plentiful and more perfectly formed the nearer we get to that rock. All these indications seem capable of explanation only on the supposition that the Granite is a truly bedded rock, which has assumed its present form through metamorphic action. Whether the result of this action has ended in the production of Granite or some other form of altered rock, would depend partly on its intensity and partly on the composition of the rock submitted to its influence. Where we get beds of Granite interstratined with other rocks, the former were probably strata whose composition rendered them more susceptible of metamorphism than the latter. Large tracts of Granite may have arisen from the alteration of a great thickness of such rocks, or from an intensity of metamorphism sufficient to convert into Granite rocks of different mineral composition. That the extent of metamorphism does actually depend in some measure on the composition of the rock operated on is found to be the case in many instances, of which the following may be taken as an example. In speaking of some metamorphosed beds in the Southern Uplands of Scotland, Professor Geikie states that the character and extent of the metamorphism have been largely determined by the original composition of the rock. Its Quartz grains have suffered little or no change ; it is the dark argillaceous base or matrix that has undergone metamorphism. Hence when a coarse quartzose grit occurs, it has suffered little alteration; but where on the other hand the rock has been formed out of a fine sandy Silt or muddy Sand, the metamorphism reaches its maximum.* The other Granite of Brittany is more coarsely grained than the one just described and porphyritic. It does not show intercalations of other rocks ; it penetrates and sends veins into the first, and contains fragments of Gneiss. These facts show that this Granite has been forcibly thrust in among the group of beds of which the first Granite is a member. This Granite forms the peaks and summits of the hill- ranges. Metamorphic Granites of the Western Territories of America. A large number of the Granites of this region are believed by Mr. Clarence King on the strength of their behaviour in the field to be the products of intense metamorphism, and this view is supported by 'the high authority of Professor Zirkel, who has examined specimens under the microscope. We have here again a case which does not depend on outdoor observations and macroscopic characters alone, but is supported by an observer who has brought to bear on the question all the refinements of modern geology. In some cases Mr. King believes that the first step in the transfor- mation of Gneiss into Granite has been effected by mechanical means. Where Gneiss has been intensely contorted, the layers of Mica, which * Memoirs of the Geol. Survey of Scotland. Explanation of Sheet 3, par. 25. 2 E 434 Geology. give it its foliation, are so shattered and smashed up that the schistose character of the rock is all but lost, and we get an approach to Gneiss without foliation, that is to Granite. He says, " The original sheets of Mica may still be traced through the rock, but they are twisted and broken and crowded into such positions that at first the rock looks like a coarse Granite." * Porphyroids of Nevada. Under this name Mr. Clarence King describes rocks which in hand specimens cannot be distinguished by the naked eye from those Porphyritic Felstones and Quartz-felsites that behave irruptively, and which therefore we should call Plutonic : some of them even approximate very closely to those rocks in their microscopic characters. They can be traced however in the field passing gradually into bedded Clay-slates and Quartzites. Judged by hand specimens alone they would certainly be pronounced to be irruptive rocks and would be placed in the Plutonic class, but outdoor observation clearly shows that they are merely intensely altered clastic rocks. Chemical analysis confirms this view. The composition of the Porphyroids agrees fairly well with the mean of the analyses of the Clay-slate and Quartzite, and there is one special point in which the resemblance is noticeable, all three are comparatively rich in Potash and poor in Soda.t Caution needed with Rocks of this Class. The instances we have given show conclusively that many rocks which are lithologi- cally identical with irruptive members of the Plutonic class, in the sense in which we have used the word, are not intrusive but are the products of the intense metamorphism of clastic strata. Great care however is needed in dealing with cases of this sort, and it is always desirable to have recourse to microscopic examination before coming to an opinion. A conglomerate in the neighbourhood of Llanberis furnishes an instance. There is no doubt that the rock is clastic, and that it has been formed very largely by the denudation of an adjoining Quartz- felsite ; it contains pebbles of that rock, and its matrix is in many cases little else but mashed-up Felsite. It has undergone a certain degree of metamorphism which has given its matrix a felsitic flinty look, and the crystals of Quartz are so little worn or damaged that they cannot be distinguished from the crystals of the Felsite itself. Even under the microscope it is scarcely possible with low powers to distinguish between the Felsite and the matrix of the Conglomerate. The Felsite has good fluxion-structure ; the pressure which the Con- glomerate has undergone has spread out the grains of its matrix in rudely parallel wavy bands which bend round the crystals in a manner that imitates most closely true fluxion-structure. We may call it pseudofluxion-structure if we like, because we happen to know that it occurs in a conglomerate and that it therefore could not have been formed in the same way as the fluxion-structure of lava, but there is nothing in the structure itself to lead to this conclusion. If our slice indeed be so cut as to steer clear of pebbles, the matrix of the con- glomerate cannot be distinguished from the Felsite either macroscopi- * Geological Exploration of the 40th Parallel, i. 67, 100, 119-121 ; ii. 379, 536. t Ibid. ii. 721. Causes of Metamorphism. 43 5 cally or under low microscopic powers. So close is the resemblance between the two that if we contented ourselves with these tests we might well be led to the conclusion that only a little extra metamor- phism was needed to convert the matrix of the conglomerate, perhaps even the conglomerate itself, into Felsite. We are far from asserting that such a transformation is not within the bounds of possibility, but examination under high powers shows that in the actual instance metamorphism has stopped far short of this. When magnified sufficiently and viewed under crossed Nicols, the matrix of the conglomerate is seen to be wholly composed of fragments, relatively large, very irregular in size, and very irregularly distributed : each fragment is a broken bit of a crystal, and the whole field is a mosaic of colours. In the Felsite there is an isotropic base which remains dark during an entire revolution, thickly studded with microliths very much smaller than the grains of the matrix of the conglomerate, with more regularly-defined outlines, and all more nearly of a size. The rock has all the characters of a glassy lava largely devitrified. In this case the pebbles prevent any possible risk of error ; but if the rock had been a grit instead of a conglomerate, it would not have been possible to distinguish that grit, where it was altered, from the Felsite by the naked eye or even under the microscope with a low power. Examination under higher powers however would even here have shown us the difference between the two. A rock which from its description seems to be exactly identical with the Llanberis conglomerate, to have been formed in the same way, and altered to the same extent, occurs in the Vosges. It is a conglomerate whose matrix has all the appearance of a Quartz-felsite, and hence it has been called Poudingue a pate euritique. At certain points however it loses its crystalline texture and passes into Grau- wacke. It contains pebbles of Quartz-felsite, and has therefore been formed out of a rock of that character ; its matrix is doubtless com- minuted Felsite, and has been altered to such an extent as to become itself felsitic in appearance.* SECTION III. CAUSES OF METAMORPHISM. The principal phenomena and products of metamorphism having been now described, it remains to inquire whether we can offer any reasonable explanation of the means by which it has been brought about. Local Metamorphism by Intrusive Igneous Rocks. It will no doubt have already occurred to the reader, in connection with this part of the subject, that we often find, along the margin of intrusive masses of igneous rock, that the beds through which they have forced their way are altered into substances identical with some members of the metamorphic class. Sandstones are baked into Quartz- ites, Limestones put on a crystalline texture, and Shales are converted into Lydian Stone or Porcellanite. The alteration does not extend * Explication de la Carte Geologique de la France, i. 343. 436 Geology. usually to any great distance, and this sort of metamorphism is there- fore spoken of as local (metamorphismc accidentel, de juxtaposition, de la roche encaissante) to distinguish it from the widespread meta- morphism (metamorphisme normal, general, regional] which has ex- tended its influence over large areas. Heat one Agent. It is clear that in such cases the heat of the molten intrusive rock has had an important share in effecting the change. We are also led to look upon heat as one of the agencies that aided in producing metamorphism, on account of the striking analogies that Metamorphic rocks offer to rocks which we know were produced by igneous fusion, in the character of their crystalline minerals, specially in the presence in them of anhydrous Silicates. Considerations of this nature lead us to believe that in a very large number of cases heat has been one and an important agent in produc- ing metamorphism. And all the successful attempts to imitate experimentally metamorphic processes have required the intervention of heat. Heat alone not enough. But there are a host of reasons why heat alone would not be sufficient. It is not always the case that in- jections of igneous rocks produce local metamorphism, as would be the case if nothing but heat were wanted for the task. Again the low conducting power of rocks makes it difficult to understand how heat could have found its way through the vast areas and enormous thick- nesses of rock that show throughout a uniform degree of intense meta- morphism. Then we often find crystalline minerals, which must have been generated subsequently to the formation of the rock in which they occur, in rocks the main body of which is very little altered at all. If these were produced by heat, how is it that the heat has had so little effect on the rock surrounding them ? On grounds like these we conclude that though metamorphism cannot be produced without heat, something else is wanted. Heated Vapours. A study of volcanic phenomena suggests as possible aids heated vapours, such as Sulphurous, Carbonic, Hydro- chloric, and Hydrofluoric acids. These we know are given off from lavas, and give rise by chemical reaction to sundry crystalline products. Water. Agents like these have doubtless assisted in the work of metamorphism ; but there is one substance so universally present in all rocks, and so thoroughly capable of effecting, when in a heated state and under pressure, so many of the changes which observation shows to have taken place, that we must look upon it as the grand helpmate of heat in bringing about metamorphic changes. That sub- stance is water. We have so few opportunities of becoming acquainted with the portion of the earth below the surface, that it does not readily occur to us how widely water must be diffused throughout the ground beneath our feet, and to what great depths it penetrates. But a little reflection will soon bring home to our minds the conviction that water must exist in as large quantity below as above the surface. In all underground explorations, as the miner knows to his cost, it is met with ; and wherever a path is open, it tends steadily downwards. Friction and the narrowing of suitable channels cause a decrease in the Causes of Metamorphism. 437 amount in many cases as we descend, arid sometimes the fortunate presence of a natural means of escape gives rise to what is practically a dry mine, but in all attempts to penetrate below the surface water is the one enemy which we always have to cope with to a greater or less degree. That it descends to depths greater than any we have been able to reach is rendered highly probable by the existence of thermal springs, which in many cases can obtain their heat only by rising from a very considerable depth. The enormous quantity of water given off during volcanic eruptions also shows that it exists in large quantities at depths below the surface. Even when friction and other impedi- ments counteract the effect of gravity in diffusing water, capillary attraction comes in to entice it through the minute pores and interstices of the rocks, and causes them to be saturated with it. Water again exists in a state of chemical combination in many minerals, and is set free when they are decomposed, as in all probability has happened during the process of metamorphism. So much for the diffusion of water. We must next take into account the increase of temperature as we descend into the earth. This subject belongs to a subsequent chapter ; it will be enough to state here that as we go below the surface the temperature rises, and that the average rate of increase may be taken at 1 Fahrenheit for every 60 feet : at this rate we should, at about a depth of 2 miles, arrive at the temperature at which water boils at the level of the sea. Under the conditions which are present at great depths, it is impos- sible to say for certain in what state water would exist, but it is certain that it would be a far more powerful agent than at ordinary temperatures and pressures, for acting as a solvent, for promoting chemical decomposition, and for softening and diminishing the coher- ency of the constituent minerals of rocks. We shall have to notice subsequently that the crumpling and crushing which the rocks have undergone is another possible source of heat, which may well have aided the work of metamorphism at less depths than that mentioned. Water too would not pass down pure ; in its passage it would take up the various minerals soluble in it with which it came in contact, and would do this to a larger and larger degree as it became more and more highly heated.* Seeing then that water is everywhere present, that it is so well able to promote and effect alteration in rocks beneath the surface, and that we know of no other substance that can compare with it in these respects, we are justified in concluding that it must play a leading part in any changes that are wrought among the rocks of the earth's crust, in fact that much of the metamorphism we are acquainted with has been brought about by Hydrothermal action. Pressure and Depth. Sundry considerations lead to the con- clusion that metamorphism must have gone on at considerable depths below the surface. A thick coating of overlying rock would be necessary to check the escape of heat, and to prevent the water and other metamorphosing * In connection with this see Sterry Hunt, Quart. Journ. Geol. Soc. xv. 488. 43 8 Geology. agents, or some of the constituents of the rock, being driven off in a state of vapour ; for instance, in the conversion of Limestone into crystalline Marble, we must have pressure to prevent the escape of the Carbonic Acid. It is quite clear too that the Crystalline rocks could not have been produced by heat alone at atmospheric pressure, because if they are fused and allowed to solidify, they harden either into a glass or a stony mass totally different from the original rock ; and if they originated from the joint action of heat and water, we shall still require the assistance of depth to raise the latter to the requisite temperature and pressure to prevent it escaping as steam. Further \ve have the fact that Metamorphic rocks usually show intense folding and contortion, and it will be explained by-and-by that, as far as we know, this could have been produced only under the pressure due to a thick mass of rock atop. Lastly the successful attempts to imitate artificially metamorphic processes have all called in the assistance of pressure.* In connection with this part of our subject it is instructive to notice that the twinning which is so charac- teristic of the calcite grains of Statuary Marble has been produced artificially in Calcite by pressure alone, t We must not suppose however that metamorphism will necessarily be produced if a rock is sunk deep enough into the earth. There are cases where we can show that rocks have been piled one on the top of the other to a thickness of 10,000 or 12,000 feet, and yet the bottom beds show no signs of what is usually called metamorphism ; on the other hand, we can point to rocks which have not had at the outside half the above thickness of cover on when they were metamorphosed, and are yet converted into crystalline Schists. When we reflect how complex the process of metamorphism probably is, and how many causes heat, water, pressure, and may be others we do not know of are necessary for its production, the seeming inconsistencies of these cases vanish. We shall also see by-and-by that the heat required was pos- sibly not derived directly from the heated interior, but was a result of crumpling and crushing ; the energy necessary to produce the compli- cated puckering of the Metamorphic rocks must have been enormous, and if any of this took the form of heat, it would probably furnish an amount amply sufficient for the work of alteration. If this be so, the amount of metamorphism ought to increase with the contortion, and not necessarily with the depth to which the rocks have been sunk. We will now glance at some of the attempts to imitate experiment- ally the process of metamorphism. Experiments of Daubr^e. Among the most instructive are those of Professor Daubree. He enclosed the substances to be operated on along with some water in a glass tube, which was introduced into a strong iron cylinder to prevent its being burst by the expansive force * Professor Geikie has tried to determine the depth at which the metamor- phisra of the Scotch Highlands was produced ; and has shown that, though it was considerable, it was probably not so great as has been sometimes supposed (Transactions Edinburgh Geol. Soc. ii. 287). t See also the experiments of Walth. Spring, Bull, de 1'Acad. Boy. des Sciences de Belgique, [2] xlix. (1880) 323, and a notice in Neues Jahrbuch, 1882, vol. i. Ref. p. 42. Causes of Metamorphism. 439 of the steam ; subjected the whole to various temperatures ; and after- wards allowed it to cool slowly. Among the results obtained were the following. Pure water was placed in the tube and the whole exposed to a dull red heat. The glass, which was a Silicate of Lime and Alkali, was partly decomposed. Numerous well-shaped crystals of Quartz were formed out of some of the Silica. Another part of the Silica combined with Lime and formed a mineral identical in every respect with the natural Silicate of Lime known as Wollastonite. Using a tube of glass which contained Oxide of Iron, he obtained Diopside. When Obsidian was operated on in the tube, a substance was obtained which when powdered and examined under the microscope had all the char- acter of Sanidine or glassy Orthoclase. In another experiment the mineral waters of Plombieres, which are rich in Silicates of Potash and Soda, were substituted for pure water. The -walls of the tube were found to be coated with Silica in the form of Quartz crystals and Chalcedony, which appeared to have been derived from the decomposition of the alkaline Silicates of the water. When a mass of pure Kaolin was treated with these waters, it was converted into a solid substance, confusedly crystallized in small prisms, which proved to be a double Silicate of Alumina and an Alkali with all the characters of Felspar ; mixed with this was a little crystallized Quartz. A mineral was also produced which there was every reason to believe was Mica. A point of great importance brought out by these experiments was the small quantity of water necessary for the transfor- mations ; in some cases this did not amount to a third part by weight of the substance transformed. It is also specially worth notice that the temperature employed was far below the fusing-point of the anhydrous silicates obtained. * In similar experiments which have been described subsequently the glass was partly transformed into a substance which approached a Zeolite known as Pectolite in composition. When the altered glass was sliced and examined under the microscope, it was found to be crowded with microliths often arranged in radiated clusters and to contain Sphserulites and small crystals of Augite.f Researches of Sterry Hunt. Dr. Sterry Hunt believes that the following changes might be brought about in a Sandy Clay com- posed of comminuted Potash- and Soda-felspars, and containing Cal- careous, Magnesian, and Ferric Salts. If the mass be permeable, atmospheric water will decompose the Soda- more easily than the Potash-felspars, and will remove the Soda, Lime, and Magnesia. If organic matter be present it will reduce the Ferric to Ferrous Salts, and these will go away as Carbonate or in combination with an organic * For details of these experiments see Annales des Mines, 5th series, xii. 289 ; Etudes et Experiences Sy nth etiq lies sur le Metamorphisme, Memoires, Academic des Sciences, xvii. (1860) ; Bulletin Soc. Geol. de France, 2nd series, xv. 97, xvi. 588. See also Durocher, Etudes sur le Metamorphisme, Bulletin Soc. Geol. de France, 2nd series, iii. 547 ; Metamorphisme dans les Pyrenees, Annales des Mines, 3rd series, vi. 78 ; Delesse, Etudes sur le Metamorphisme des Roches, and Annales des Mines, 5th series, xii. and xiii. ; Vernon Harcourt, Report of British Association (1860), p. 175 ; Mitscherlich sur la Production artificielle des Mineraux crystallises, Annales de Chimie, xxiv. 258 (1824). t Geol. Experimentale, i. 154-177. 44 Geology. acid. There will remain the Silica, Alumina, and Potash, that is to say the constituents of the minerals of the Acid Crystalline rocks. But if the original mass be clayey and impermeable, it will retain the components of Lime- felspars and of the minerals of the Basic Crystal- line rocks. If now these residues be acted upon by water holding in solution alkaline Carbonates and Silicates in the presence of heat and under pressure, the various siliceous minerals of an Acid or Basic Crys- talline rock, as the case may be, will be formed. His opinion is that a temperature of 212 Fahrenheit would suffice for the production of Silicates of Lime, Magnesia, and Iron, and that at 480 the Felspathic and Micaceous Silicates generally could be formed.* Observations of Dr. Sorby. Dr. Sorby, who was among the first to call attention to the geological bearing of the existence of water- and stone- or glass-cavities in the minerals of the Crystalline rocks, has pointed out the presence of water-cavities in the Quartz and Garnet of Mica-schist and Gneiss. He shows how Felspathic Clays might be converted into crystalline Quartz and Mica, so as to constitute Mica-schist, by the removal of part of the alkaline bases, and argues, from the presence of water-cavities, that the alteration was not the effect of dry heat and partial fusion, but was due to highly- heated water disseminated through the rock. The fact that the process took place at a high temperature is inferred from the presence of bubbles in the cavities ; and by determining the degree to which the crystals must be heated in order to make the liquid expand so as to fill the cavity, he has endeavoured to fix what that temperature actu- ally was. This last step of the problem cannot be solved unless we know also the pressure under which the operation took place ; but by making reasonable assumptions on this head most instructive results are arrived at.t Variations in amount of Metamorphism. In considering the action of the various metamorphosing agents we must recollect that they would act unequally on different rocks. Some rocks would conduct heat more readily and be more pervious to water and vapours than others ; the final result would also depend on the original compo- sition of the rock operated on ; and thus we can easily understand how it is that in a mass of inetamorphic rocks we find some beds much more altered than others immediately in contact with them. Thus Professor Geikie tells us that among the altered rocks of the Southern Uplands of Scotland there appears to be always a close connection between the nature and extent of the metainorphism and the chemical constitution of the rocks in which it is manifested. It is always most developed in those strata into whose composition Felspar enters as a main ingredient, while on the other hand in the more quartzose rocks little or comparatively little change has taken place. } ]S T ew elements besides might well be introduced by water into the rocks * Quart. Journ. Geol. Soc. xv. 488 ; Report on Geological Survey of Canada (1856), p. 479 ; Silliman's Journal, 2nd series, xxx. 135, xxxvi. 214, and July 1864 ; Chem. and Geol. Essays, pp. 12-21. t See the references to this subject in the note to p. 310. Memoirs of the Geological Survey of Scotland. Explanation of Sheet 15, par. 35. Causes of Metamorphism. 441 through which it finds its way, and we must therefore not be sur- prised if the chemical composition of a metamorphic rock differs from that of the mechanical deposit from which it was derived. Subsidiary Metamorphosing Agencies. The causes men- tioned seem, as far as our knowledge goes, to have been the main agents in the production of metamorphism ; but in particular in- stances other subsidiary influences no doubt gave their help. Thus, for instance, Forchhammer believes that the presence of sea-weeds has conduced largely to bring about the present condition of the Alum- slate of Scandinavia.* Summary. Our present knowledge does not enable us, and it is doubtful if we ever shall be able, to unravel fully the intricacies of the subtle process of metamorphism ; but reasoning, such as that which has been laid before the reader, enables us to form what is probably a very just notion of the general way in which the result has been brought about. Heat is in a large number of cases an essential requisite, but the difficulty of accounting for the transmission of heat by mere conduc- tion through such vast masses of rock as have been thoroughly meta- morphosed, obliges us to look for some vehicle which would propagate 'it by convection. Such a vehicle we find in water. The universal presence of this substance, its incessant state of circulation, and its high specific heat, fit it admirably for the task of acting as a diffuser of heat ; and at the same time by its power of softening and acting chemically upon rocks it aids materially in promoting rearrangement of the constituent minerals, decomposition and the formation of new compounds, and the introduction of fresh elements. Other sub- stances, in the state either of liquid or vapour, may have had a share in the process. Lastly, we can obtain the requisite widespread heat, and the means of preventing the escape of water and other volatile substances which aided it, only under the pressure of a considerable thickness of overlying rock ; and on this and other grounds we con- clude that metamorphism went on deep under ground, and that we see metamorphic rocks at the surface now only because they have been uplifted, and the covering of rock under which they were once buried has been removed by denudation, t Three Stages of Metamorphism. So far we have recognised three subdivisions in the metamorphic rocks, and the differences which distinguish these three classes seem in the main to depend on the dif- ferent degrees of metamorphism which their members have undergone. A certain amount of metamorphic change gives us rocks like Quartzite and Statuary Marble : a further degree of alteration gives rise to Foli- ated Schists : metamorphism still more intense produces rocks litho- logically identical with those of the Plutonic class, and which differ from Plutonic rocks only in never being intrusive. The reader surely asks, Does metamorphism stop here] Would it not be possible to carry on the process till an irruptive Plutonic product was obtained 1 We will discuss this point in the next chapter. * Report of British Association, 1844, p. 77. t The student may further consult Gunibel, Ostbayeriscb.es Grenzegebirge, p. 838, and Kalkowsky, Zeit. d. Deutsch. Geol. Gesell, xxxviii. (1876) 747. 44 2 Geology. Metamorphism no Proof of Antiquity. It is a fact which has been long noticed that rnetamorphic rocks are more plentiful among the older than among the younger members of the rocks of the earth's crust. So generally true is this that at one time the fact of a rock being a Crystalline Schist was looked upon as conclusive proof of great antiquity. Such an inference is however by no means sound and good. We can point to rocks of this class which have certainly been produced during geological periods comparatively recent ; and there can be no doubt that metamorphism has always been going on and is now in progress. But if we bear in mind the position of the metamorphic workshop, we shall see that the prevalence of metamorphic products among the older rocks is only what is to be expected. The meta- morphic rocks of recent date are for the most part hidden from sight, because denudation has not yet had time to strip off the covering of rock beneath which the alteration was effected. It is only in districts like the Alps, where great upheaval and extensive denudation has gone on in comparatively recent times, that we can hope to get a sight of rocks that have been metamorphosed during the later portions of the earth's lifetime. We must also bear in mind that the older a rock is, the greater chance will it have had of having been sub- jected to metamorphic influence, and this will tend to make meta- morphic products more abundant in the older than in the newer rocks. On the other hand, while we are bound to admit that metamorphism has always been going on, we shall see in Chapter XIV. that there is reason to believe that at very remote periods its action must have been more vigorous than at present. Whether however any of the metamorphic rocks formed during these periods still survive, is another and a very open question. CHAPTER X. GENERAL VIEW OF THE CRYSTALLINE ROCKS. " In eternal restless change Self- fed and self-consumed. " MILTON. Natural Classification of Crystalline Rocks. We have now looked at the Crystalline rocks from two points of view. In Chapter VI. we regarded them solely from a lithological standpoint ; their mineral composition and textural peculiarities were the only matters to which we paid any attention, and they were named and arranged in classes with reference to these characters alone. In the three following chapters these rocks have been treated from the petrological standpoint; the larger peculiarities of structure which they present when they are studied on a large scale in the field have been brought before the reader, and we have by this means been able to arrive at explanations more or less satisfactory as to the ways in which they were formed. As we proceeded, the truth which had forced itself upon our notice in the case of the Derivative rocks was found to hold good in the case of the Crystalline rocks also. Evidence of the unsatisfactory nature of a purely lithological classification presented itself from time to time. Among the Derivative rocks all these rocks which are mainly composed of Carbonate of Lime obtained the name of Lime- stones, and were, as long as mineral composition was the only point we had in view, all placed together in the same class. But further investigation showed that some Limestones had been formed by organic agency and others by chemical precipitation. The incon- sistency of grouping together rocks which had been formed in ways so totally distinct, became evident, and we found it necessary to rearrange the Limestones and the Derivative rocks generally accord- ing to the manner of their formation. Just in the same way a litho- logical classification of the Crystalline rocks is found to be unsuitable and even inconsistent. A number of rocks for instance, all of which had the same mineral composition and general texture, came to be called Granites. But field observations indicated clearly that there are Granites and Granites, and that some had been formed in one way and some in another, and it then became no longer possible to include them all in the same class. And rocks which lithologically are widely apart are seen, when their method of formation is taken into account, 444 Geology. to be closely akin. The majority of Dolerites and Trachytes for instance are lavas, and are therefore much more nearly related geologi- cally than Trachyte and Granite, in spite of the fact that Trachyte and Granite are both rocks of acid composition, and Dolerite is a basic rock. It is an unfortunate circumstance that in the case of the Crystalline rocks our nomenclature is so purely lithological, but there is no help for it. The names were given before geology had advanced beyond the lithological stage, and they are far too firmly rooted to be easily discarded. Nevertheless a rational classification, based on the method of forma- tion, must be had, and such a classification we will now attempt to frame (see p. 445). . Are all Crystalline Rocks Metamorphic Products ? A classification such as that sketched out on p. 445 has weak points enough, but it is a decided advance on a classification based on litho- logical characters alone, and its very failings prompt lines of thought which may lead us up to clearer and wider views on the subject of the formation of the Crystalline rocks than we have yet been able to reach. We know that there is no hard and fast line between the Volcanic and the Plutonic rocks ; the raw material was the same for both, the difference in the product depends solely on a difference in the way in which it was worked up. The thought naturally suggests itself that the Plutonic arid Metamorphic classes may in like manner pass into one another. Let us inquire whether there are any grounds for such a supposition, and to give greater definiteness to our views let us take the case of an individual rock Granite, examples of which occur in both these classes. It is unfortunate that those who have speculated on the origin of Granite have in many cases looked only at a part of the facts on which the solution of the problem depends. A case is found in which Granite shows every sign of having been forcibly intruded in a fused state into the beds among which it occurs, and the observer thereupon jumps to the conclusion that all Granites are irruptive. Another ob- server detects Granite under circumstances which raise the strongest suspicion that it has been formed in situ by intense metamorphism, and forthwith refuses to believe that it has ever behaved irruptively. We have endeavoured to avoid these one-sided ways of reasoning by laying before the reader instances of both these methods of occurrence ; and now the question arises, Have these two forms been produced by different causes, or are they only the results of different stages of the same operation ? We have already seen good grounds for believing that Foliated Schists have been formed out of Clastic Derivative rocks by a somewhat advanced stage of metamorphism : Granite occasionally occurs among Foliated Schists in beds. Here there can be no reason for refusing to believe that Granite has been formed in the same way as the rocks among which it occurs, viz. by metamorphic action. The metamorphism in these cases may have been more advanced than that which produced the Schists, but it was not intense enough to efface General View of Crystalline Rocks. 445 o g ^^> ^ li ^ ?. i 1 X ^0 ~V3 s PL, ^ PH O 9 5 CO ^ c M ^ -^ -2 C3 s | o> ^ -M o > 'S o b O "CO 1 ^ rt ^ V -3^ r^ W o p^ I> 3 iJ CO c S o 01 -3 0> ;_, c 'IS > 1 aJ o 3 > s 5 r o 1 5? 3J ^ '_ ~, al * eS cc & i^ co >, S 2 o ^ V2 'ri M 32 2 b '3 CJ > ? K> o P4 ,_^ 8 ^_ 'oj N o s o o 4d s ?C 'o fe 03 OJ ctf S C3 M CU ^s O G>rf ^ ^4 S5 "S C " _o 4J 'z, Q )1^ iC e r a ad oo h M O J t3 5 . 00 O o ^ o "SQ Q ^ -^ -3 ^ o M S S 1 E 1 | C/} C3 'o J > 1 o 2 ~ > .-. 3 __ i c S 5 S -3 | i i | $ O> > i of j g .s ^ 3 | i-I S _2 r2 ^i ^ 00 eS "o ^Q oe S C o gf el ^ f ^ C/3 o M o 71 ^ p*^ XJ * K*"3 ~t~* B -^ t'j ^3 -n u ^ S ^c 1 00 c3 t a 'a c "~r 'S *^C .2 g __ j-j E- 9 1 J "oo cc o S O $ o rt a _0 3 oo CJ a5 1 J rd $ i q5 x 3^ c O a OJ -3 _o 05 ^ P* ^ i a '1 J ^ O 1 S3 '> cS O 1 S >> q OB 1 t oi cT o CG |l 1 < ^ 1 o i "3 ^ 5^5.2 S^^ |^ cavities c Genera ^s 1 i s *"*' - z s OTT-^ -d 5 ^ ll'S go I .S y S I P5 P5 446 Geology. the bedding. Again Granite occurs in amorphous masses which melt away on all sides into the surrounding strata but show no signs of having burst violently through them, rather convey the idea that they fill up spaces once occupied by rocks similar to those which now sur- round them ; and there is a most perfect passage from a central gran- itic mass into foliated rocks, from foliated into less altered rocks, and from these into unaltered clastic strata. Here the evidence is of the strongest character in favour of the granitic core being, like the un- doubtedly metam orphic rocks into which it shades off, itself meta- morphic in its origin. The metamorphism was probably more intense than in the case of bedded Granite, for the stratification is completely destroyed. Again cases are known of masses of Granite which melt away into the surrounding rocks in the manner just described along a portion of their margin, but elsewhere behave intrusively and send out veins and dykes into these rocks. No difference in composition and general character exists between the non-intrusive and the intrusive portions, they both form part of the same mass, and it would seem that both must have been formed by the same process. ]S T ow the process which gave rise to the non-intrusive Granite must have consisted in a soften- ing and decomposition of the particles of the rock to an extent that allowed of a molecular rearrangement of their constituents. Let this same process be carried further, and it is perfectly conceivable that the rock might be reduced to a high degree of plasticity or fluidity : that increasing heat and the pressure of the overlying rocks might drive it against the surrounding beds with sufficient force to strain and rend them ; and that it would then be injected into the fissures so formed. There is nothing a priori unreasonable in the supposition that intrusive Granite has been produced by metamorphism more intense than that which gave rise to the non-intrusive bodies of that rock, and the occurrence of the two forms in different parts of the same mass scarcely leaves room for any other conclusion. We want but one step more. Let the energetic metamorphism which suffices to produce intrusive Granite be no longer limited to certain spots but act throughout the whole mass, and all the Granite becomes intrusive. Here then we have good reasons for the belief that one kind of Irruptive Plutonic rock can be produced by very intense metamorphism ; and there is no ground for limiting the statement to Granite. The mineral composition of the metamorphic product will depend on the composition of the rock from which it was derived, and while certain clastic rocks yield, when sufficiently metamorphosed, Granites, others will give rise to Plutonic rocks of different composition. Instances are not wanting. Professor A. Geikie tells us of an intrusive sheet of a rock, which can be classed as Diorite, among the altered rocks of the southern uplands of Scotland, but which, in spite of its Plutonic character, is found in the space of a few 'yards to pass into a rock which does not differ in general aspect from beds which are un- doubtedly metamorphosed Felspathic Sandstones.* And he mentions * Memoirs of the Geol. Survey of Scotland. Explanation of Sheet 3, par. 26. General Vieiv of Crystalline Rocks. 447 that while some of the stratified rocks, probably originally more quartzose, have been changed into Granite, others, which were pro- bably more felspathic and argillaceous, have been altered into various Porphyries and Diorites.* If the reasoning just advanced be admitted, the line between the Plutonic and Metamorphic classes breaks down ; Plutonic rocks are produced by a metamorphism more intense than that which gave rise to the rocks we have styled Amorphous Metamorphic. And since Volcanic are merely the subaerial form of Plutonic rocks, we arrive at the conclusion that in the end all Crystalline rocks owe their existence to the metamorphism of clastic strata. Instances of Irruptive behaviour by Metamorphic Rocks. The fact that a crystalline rock which is in the main meta- morphic behaves intrusively at certain points is not very easy to substantiate, and hence the cases of this nature which can be put in evidence are not numerous, and are not all of them as well established as could be wished. The following are a few instances. The Granite north of Dundalk described on p. 11 of the Explana- tion of Sheet 70 of the Geological Survey Map of Ireland (p. 432, note) is apparently metamorphic, but at some spots it sends veins and dykes into the adjoining beds. The late Professor Jamieson has described Granite in Banffshire which he gives good grounds for thinking has been formed by the raetamorphism of rocks similar to those which abut against it, but this Granite is very frequently intrusive. t In neither of these cases are the details quite full enough to render it certain that the Granite has been produced by the metamorphism of rocks similar to those which now adjoin it. A far more satisfactory instance, which has been described by Professor Judd,J is found in the district of Schemnitz in Hungary. We have there the ruins of a great volcano. Paroxysmal eruptions have blown away a large portion of the cone, and denudation has further eaten down into the crateral hollow thus produced to such an extent that sections are laid bare through the very bowels of the volcano. In this way the great Plutonic masses, which are the hardened contents of the old volcanic reservoirs, have become exposed to view. These intrusive bodies have wrought local metamorphism among the bedded rocks through which they were forced and have converted them into Schists, Quartzite, Crystalline Limestone, Gneiss, and Aplite. There can be no question about these Crystalline rocks being true products of metamorphism, for they can be traced passing by the most insensible gradations into the unaltered bedded rocks which adjoin them. For the most part they retain more or less of bedded structure and show no disposition to behave intrusively, but at some spots metamorphism has been carried further in the case of the Aplite, and that rock breaks through and sends veins across the Gneiss which overlies it. The evidence is conclusive that the metamorphism which could turn a clastic rock into * Memoirs of the Geol. Survey of Scotland. Explanation of Sheet 15, par. 35. t Quart. Journ. Geol. Soc. xxvii. (1871) 105. Ibid, xxxii. (1876) 320. 448 Geology. Aplite 'was competent when carried a stage further to convert the Aplite into an irruptive rock. We have noticed the occurrence of two kinds of Granite in Brittany (p. 432), one intrusive, and the other apparently truly interbedded with Metamorphic rocks. The first forms the peaks and ridges, and the second is found along the flanks of the hill-ranges. This restric- tion of each kind to a separate region seems full of meaning, for before the rocks were brought into their present position, the beds out of which the first Granite was formed were the lowest of the group, while those which gave rise to the second Granite were higher up in the series. * The metamorphism of the first Granite therefore went on at a greater depth than that of the second and was proportionably more intense, so that the rock came not only to be rendered more or less fluid but to be driven forcibly into the strata around it ; the second Granite, having been formed at a smaller depth when the metamorphism was less complete, still retains traces of its original bedding. Counter-views on the Origin of Crystalline Rocks. The views which have been just propounded were for a long time stoutly opposed by many eminent geologists, and they are far from being universally accepted even now. They were put forward before the use of the microscope had become common among geologists, and some of the instances which were advanced in support of them broke down when subjected to the test of microscopic examination. But among the cases we have brought forward there are some where all the refinements of the geology of the present day have been employed without in any way shaking the evidence. It may be well to give a sketch of the counter- opinions which have been held as to the origin of the Plutonic and Volcanic rocks. There is good reason to believe that the whole earth was once in a state of fusion, and, according to one school, it consists now of an external solid crust, within which a mass or detached masses of the original fluid material still remain in a molten condition. Taking this view of the constitution of our planet, some geologists will have it that all lavas are portions of this molten interior mass, which have been forced up to the surface, and that those rocks known as Plutonic, which differ from subaerial lavas in being more compact and crystal- line, are portions of the same molten interior mass which have cooled and hardened under pressure. This school then draws all igneous products from an internal permanently molten reservoir. The two views as to the origin of molten igneous products are not necessarily antagonistic. They may be both true. Some lavas and traps may have come from an interior permanently fluid reservoir, and some may have been produced by the melting down of portions of the solid crust. On the other hand the existence of any permanently molten masses near enough to the surface to allow of their contents being forced out into the air, seems on many grounds extremely improbable. Objections to Metamorphic Theory. Those geologists * The reader will realize why we are justified in making this statement when he has gone through Chapter XII. General View of Crystalline Rocks. who dispute the metamorphic origin of Granite seem to rest opposition mainly on two lines of argument. They say that the com- position of the Crystalline rocks all the world over is so uniform that it is not likely that they could have been derived from beds so variable as they assert the Derivative rocks to be ; and they maintain that there is no Derivative rock which agrees in chemical composition with Granite. The first of these statements is certainly not strictly true ; and if it were, it would not prove the point it is intended to establish. Crystal- line rocks are far from observing the constancy in mineral and chemical composition that is assigned to them. And if the assertion is only intended to be taken in a wide general sense, it will apply equally well to those Derivative rocks from which the metamorphic hypothesis sup- poses Granite to be derived. Certain Sandstones looked at broadly are quite as uniform in composition as any Granite ; and therefore there is no ground for surprise if, when these Sandstones are metamorphosed, the products also are very much alike. The first objection therefore falls to the ground.* The second statement, that there are no Derivative rocks of the same composition as Granite, is decidedly open to question. Instances to the contrary might be multiplied without limit, but one example must suffice here. The two analyses given below are taken almost at random from Zirkel : ( 1 ) shows the composition of a Clay-slate from Prague ; (2) that of a Granite in the Carpathians. (i) Silica 67-50 Alumina Lime Magnesia Potash Soda Oxide of Iron and Manganese 15-89 2-24 3-67 1-23 2-11 5-85 (2) 69-31 16-40 3-06 0-83 2 87 3-29 1-79 There can be no difficulty, as far as chemical composition goes, in believing that the constituents of the first rock might be so rearranged as to give rise to something very like the second. But even if this objection were well founded, it would not be fatal to the metamorphic theory ; for since water plays so important a part in metamorphism, we can readily conceive that any ingredients necessary for the trans- formation of the rock into its new shape may have been introduced in solution either during or after the process of metamorphism, and that any superfluous constituents may have been carried away in the same manner. Summary and Conclusions. We have now for the space of ten chapters been plodding through dry descriptions of various kinds of rocks, and explanations, some in the highest degree probable, others involving more or less of speculation, of the ways in which these rocks were originally formed and have been subsequently modified. A stone is no longer to us a stone and nothing more ; every stone carries with it a story, and the experience we have gained enables us to decipher for each individual stone, with more or less of certainty, the characters * See Allport, Geol. Mag. ix. 188. 2 P 45 o Geology. in which that story is written, and translate it into our own tongue. But so far we have done little more than relate so many detached incidents in the history of the formation of the earth's crust, little more than collect a bundle of historical anecdotes, such as is put into the hands of children to awaken in their minds an interest for historical reading and lead them up to the study of history itself. The questions arise, Is our geological knowledge as yet sufficient to enable us to do any thing further than gather a budget of geological tales 1 Do the iso- lated facts we have been reviewing naturally lend themselves to a con- nected narrative ? Can we ascend from them to broad general views, and frame out of them something deserving the name of a history of a portion at least of the lifetime of the earth on which we live 1 If it seems likely that we can, we shall do well boldly to make the attempt. For the history of every science shows that, if generalizations are made in a truly cautious and philosophical spirit, and when necessary, looked upon as merely provisional working hypotheses, the gain that follows from them is immense ; nay more, that if they are not made when the right time for making them has arrived, the loss that results is still greater. Not only do well-grounded hypotheses serve as a string on which to hang our facts, where they can swing in full view and be readily got at when wanted, but they also point out the direction we must go in if we wish to add to our collection. And in the opinion of many eminent thinkers the time is ripe for some degree of generaliza- tion in geology. It matters not that many pages of the geological record are so blurred and blotted that we can only grope our way stumblingly through them ; that many can be read in several different ways, so that the interpretations of them are almost as numerous as the interpreters ; that many are altogether blank, and many torn out and gone for ever. We take heart when we find very many written in characters which cannot be misunderstood, and find too that the pages we can read are numerous enough to justify us in attempting conjectural restorations or emendations of those which are lost or corrupt. The reader has no doubt by this time pretty shrewd notions as to the nature of the theoretical conclusions which will follow the above introductory remarks, and it only remains to put them into a formal shape. We will take Granite as an instance, but merely for the sake of rendering the explanation more definite ; all that is said about Granite will apply equally well to the rest of the Plutonic rocks. We have found that Granite occurs under three forms. Under the first it still retains traces of bedding or is interstratified with undoubt- edly bedded rocks ; here there can be little doubt that it is an intensely- metamorphosed rock. Under the second form Granite occurs in amorphous masses, which melt away insensibly on all sides into unaltered strata, show no signs of having burst violently through the adjoining beds, but look as if they filled up spaces once occupied by rocks similar to those that surround them. Such appearances are best explained by supposing that portions of the rock-mass, in the heart of which these bosses occur, have been altered into Granite, the meta- morphism having been more intense than that which produced the first General View of Crystalline Rocks. 45 1 form of the rock because the bedding is effaced, but yet not energetic enough to cause the Granite to behave irruptively. Under its third form Granite gives proof of having been forcibly intruded into the rocks among which it occurs, and its irruptive behaviour may reason- ably be attributed to an increased degree of energy in the metamorphic process which gave rise to it. There is reason then to believe that these three forms of Granite have not been produced by different causes, but are the results of three successive stages of the same process ; and now that we seem to have seen our way to three links in a chain of operations, we are led on to inquire whether any others of the geological processes we have become acquainted with may not belong to the same series, and to try and assign to them their proper places in it. Now the metamorphism which produced bedded Granite differs in all probability from that which gave rise to Gneiss and less highly- altered products, only in intensity. On the one side then we have a passage from the bedded form of Granite through Gneiss and other Metamorphic rocks into the Derivative rocks out of which the latter were produced. Here then we seem to have an unbroken chain link- ing on Derivative rocks to one form of Granite. Looking in the other direction, the bedded Granites pass through the second form of that rock into Granites which have been shifted from the spot where they were metamorphosed and driven violently into rents and fissures. If such openings fail to reach the surface, the injected masses harden under pressure arid give rise to Plutonic pro- ducts. The passage for instance that has frequently been observed between intrusive Granite through Elvanite into compact Felstone (Petrosilex) shows one case of a Plutonic rock which is nothing but Granite modified by the circumstances under which it cooled. But if Granite should be injected into a vent opening above ground, we can scarcely doubt that the portion of it which hardened under ordinary atmos- pheric pressure would take the form of some such rock as Quartz-trachyte. On this side then Granite is connected with one of the commonest forms both of the deep-seated and subae'rial Crystalline rocks. The complete chain of operations then would seem to be as follows. First, Derivative rocks are formed by the wear and tear of crystal- line strata. Certain of these Derivative rocks, coming within the range of metamorphic action, pass through various stages of metamor- phism into Gneiss, and thence into the three successive forms of Granite. By the final step an intrusive product is obtained, which if it harden under pressure takes the form of a Plutonic rock of Acid composition, such as Granite, Elvanite, or compact Felstone ; but if it be ejected on to the surface, hardens into an acid lava, such as Trachyte. And so in the end we come back to the Crystalline rocks with which we started.* * See the paper of Judd's, already referred to, on the Secondary Rocks of Scotland, Quart. Journ. Geol. Soc. xxx. 233-237, 289-295 ; A. Geikie, The Geology of East Berwickshire (Memoirs of the Geol. Survey of Scotland), p. 33 ; Ramsay, Address to Geol. Section of British Association, 1866. According to Sterry Hunt, Keferstein in his Nuturgeschichte des Erdkorpers (i. 109) suggested as far back as 1834 that Crystalline rocks have arisen from the metamorphism of sedimentary deposits. Chem. and Geol. Essays, p. 16. 45 2 Geology. Granite has been selected as a particular instance for illustrating this great cycle of changes, but the line of reasoning applies equally well to all the members of the Plutonic class of rocks. Derivative rocks of suitable composition are capable, when subjected to the same process, of giving rise to Basic and other varieties both of deep-seated and subaerial Crystalline rocks. If the hypothesis just explained be true, we might expect to find that periods of great metamorphism would be also periods of great vol- canic activity. Professor Geikie has pointed out one instance in which this has certainly been the case ; * and he has suggested a very pro- bable reason for this connection, which we shall have to consider when we come to inquire into the cause of volcanic energy. An attempt has been made to present to the eye a diagrammatic representation of the round of changes from Derivative to the different forms of Metamorphic and Igneous rocks in fig. 130. On the right widespread regional metamorphism is going on over a large area, the action increasing in intensity from left to right. On approaching this tract the bedded rocks gradually put on metamorphic forms and shade off into Gneiss and bedded Granite ; as we get more into the heart of the metamorphic region, the latter passes into molten amorphous Granite. The beginning of the change and the final pas- sage into Granite take place at a greater distance from the metamor- phic centre in some beds than in others, those first affected being more susceptible of metamorphic influence than the strata above and below them. In this way we get along the margin of the metamorphic region beds of Granite interstratified with crystalline Schists. Still further to the right, where the metamorphic energy has reached its maximum, portions of the fused Granite have been injected into the overlying rocks. Some of the rents do not reach the surface, and the matter that hardens in them gives rise to intrusive Plutonic products ; two other rents do reach the surface, and the portions of the fused mass that are forced up through them flow out and harden into lavas, some of which are poured out on land, and others, streaming over the sea-bed, become interstratified with the derivative deposits that are being laid down beneath the water. To the left is a centre of local metamorphism, around which portions of the Derivative rocks are converted into Granite ; here, owing to the smaller intensity of the metamorphism, none of the molten matter has behaved intrusively ; the irregularities in outline of the Granite masses are not caused by intrusion of that rock, but are due to the fact that some beds are more readily metamorphosed and converted into Granite than others, and accordingly the process of change has spread further into them than into the strata less open to alteration. A belt of altered rock fringes this mass and also the Granite dykes. to the right ; but it will be noted that this is much narrower than the broad band of altered rock which abuts against the tract of regional metamorphism. Further Classification of the Crystalline Rocks. The nomenclature adopted in the table on p. 445 is open to objection on several points. The Volcanic and Plutonic rocks are both closely con- * Transactions Edinburgh Geol. Soc. ii. 287. General View of Crystalline Rocks. 453 nected with volcanic action, and it seems hardly desirable to limit the term " volcanic " to a portion of these rocks. Again Plutonic and Metamorphic rocks are both of them the products of deep-seated action, and there is an inconsistency in applying to a part of them only a name which implies that they were formed underground. We may in a measure avoid these blots thus. The rocks we have called 454 Geology. Volcanic are all Eruptive, and none of the other Crystalline rocks are ; we may therefore substitute the term Eruptive for Volcanic. For a similar reason Irruptive may take the place of Plutonic. The term Metamorphic may include all the Crystalline rocks which are not and never have been intrusive. Again there are three things we want to know about a rock if our knowledge of it is to be complete. First and foremost, not its mineral composition which is the point to which petrographers have hitherto paid most attention, but the way in which it was formed. Secondly, the petrological manner of its occurrence. Thirdly, its mineral composition and lithological character. Now in the case of the Crystalline rocks we may subdivide them into three classes as far as the first head goes. First those which we have agreed to call Metamorphic, in which the metamorphism has not gone far enough to produce intrusive behaviour. These again fall into three subdivisions, the Non-foliated, the Foliated, and the Amorphous or Granitoid. The second class includes those in which metamorphism has advanced so far as to render them intrusive, but which have not reached the surface. These are to be styled Irruptive. The members of the third class have burst their way up to the sur- face and are distinguished as Eruptive. Our next care will be to specify the petrological form under which each rock presents itself ; and the principal petrological forms under which Crystalline rocks occur are Beds, i.e. true strata, Masses, Dykes, Veins, Necks, Intrusive Sheets, and Contemporaneous Sheets or Flows. Lastly, the mineral composition may be denoted by an adjective, such as Granitic, Dioritic, Basaltic, and so on. Our scheme will then stand thus. Mode of Formation. Mode of Occurrence. Principal Mineral Varieties. A. Metamor- phic ' Non-foliated Foliated . . Amorphous or Granitoid . ( . More or less | distinctly -J bedded | I Masses Quartzite, Porcellanite, Crystalline Limestone. Schists, Gneiss, Gran- itic Gneiss, Bedded Granite. Granitic, Syenitic, Fel- \ sitic, Dioritic, etc. B. Irruptive Masses Dykes Veins Sheets \ Granitic, Syenitic, Dior- f itic, Felsitic, Trachy- f tic, Basaltic, etc. ; ; some Schistose. C. Eruptive Flows Necks Dykes / Felsitic, Trachytic, ( Basaltic, Doleritic, etc. General View of Crystalline Rocks. 45 5 An application to one or two actual instances will perhaps make the above scheme more intelligible. The stratified Granites of Donegal (p. 401) will according to it be described as Granitic Metamorphic Beds. The Granite of Priestlaw (p. 430) will be a Granitoid Mass. The Granites of Devon and Cornwall (p. 392) will be Granitic Irruptive Masses. In Arthur's Seat (p. 364) the Intrusive Sheets will be Doleritic Irruptive Sheets ; the Interbedded Lavas, Doleritic Eruptive Flows, or simply Doleritic Flows ; the rock of the summit a Basaltic Eruptive Neck, or simply a Basaltic Neck. Necks and Flows being necessarily Eruptive, the adjective in their case may be dropped. CHAPTER XL HOW THE ROCKS CAME INTO THE POSITIONS IN WHICH WE NOW FIND THEM. ' ' They are raised for ever and ever, And sink again into sleep. " TENNYSON. SECTION L NATURE OF THE DISPLACEMENTS WHICH ROCKS HAVE UNDERGONE. WE have now learned the ways in which the different kinds of rocks composing the earth's crust were formed, and the modifications of structure which some of them have undergone since their formation ; our next step will be to inquire into the changes of position they have suffered, and how these changes were brought about. Displacements which Submarine Beds have suffered. A large number of the rocks of the earth's crust were originally formed in approximately horizontal beds at the bottom of the sea, but this is not the position we now find many of them in. They have frequently undergone two very important displacements. First, they have been raised high and dry into the air, sometimes even up to the summits of lofty mountains ; secondly, the beds into which they are divided are no longer horizontal, but inclined to the horizon at all angles from the gentlest slope up to becoming absolutely vertical, oftentimes bent into broad folds or puckered up into the sharpest arifl most complicated curves, in some cases even turned over so that the stratum originally at the bottom is now uppermost. It will conduce to clearness of ideas if for the present we consider these two displacements, the upward rise of the beds and the displace- ment of them from their originally horizontal position, as separate facts ; but we shall see in the end that it is very possible they are both due to the same cause. SECTION II. VERTICAL ELEVATION. Two possible Explanations of Elevation. The presence of beds, which were formed beneath the sea, at different heights above its present level, may be accounted for in two ways. Either the sea Vertical Elevation. 457 has shrunk and had its level lowered, or tracts which were once beneath its waters have been raised into dry land, other tracts being depressed to form receptacles for the water thus displaced. Arguments against a lowering of the Sea-level. There are many insuperable difficulties in the way of accepting the first explanation. According to it the ocean must have stood in some cases upwards of 10,000 feet above its present level, and as a rise in the sea-level must be universal, the whole of the globe must have been submerged to this depth. Here we are at once met by the difficulty of getting rid of so enormous a bulk of water, a difficulty which is very much increased by the fact that geology shows that the relative level of the sea and land has oscillated upwards and down- wards over and over again ; so that we have not a general decrease in the waters of the ocean to account for, but countless repetitions of alternate swelling and shrinking. Further, to take a particular instance, marine strata belonging to what is known as the Nummulitic formation, are found in the Alps and other mountains at a greater height than that just named ; if the sea ever reached up to the level where they now occur, nearly the whole earth must have been under water, and nothing would have been left in the shape of dry land but a few islands formed by the peaks of the highest mountains. But in France and England there are Estuarine and Lacustrine deposits containing the remains of land animals, of the same age as the Nummulitic beds, and this shows that at the time when the latter were being formed extensive tracts of land existed at no great distance from the spot where they attain so great an eleva- tion. It is evidently impossible that the sea-level should have retained its present position in England and France, and at the same time stood so much higher in Switzerland. Besides, the very existence of Deri- vative rocks requires land, from the waste of which the materials necessary for their formation may be derived. But where shall we find land enough if the whole globe w r as submerged to these great depths ? Again, the hypothesis of the lowering of the sea explains only one- half of the facts we have to account for ; no alteration in the depth of the ocean will tilt beds originally horizontal, or fold and contort them. The Land has gone up, not the Sea gone down. For these and other similar reasons we cannot allow of the- possibility of such oscillations in the sea-level as are required by the first explana- tion, and are driven to attribute the occurrence of marine beds in inland and lofty situations to an elevation of the sea-bed, by which tracts once below its waters have been upraised and turned into dry land.* Denudation gives Proof of Elevation. The phenomena of denudation point to the same conclusion. The wear and tear of the land, which is everywhere going on slowly but without ceasing, if it had been allowed free play without any counteracting influence, must long ago have swept away everything exposed to its action, and have reduced the land to a dead flat but little raised above the sea-level. * The whole question is lucidly treated by Playfair, Works, i. 432. 45 8 Geology. This has not happened, and there must have been therefore some antagonistic force at work to counteract the levelling tendency of denudation. Just what we want would be supplied by forces of eleva- tion, which from time to time raised sea-bottoms into dry land and so formed new continents to take the place of those which had been worn down by denudation. Instances of observed Oscillation of Land. Considera- tions such as these would be quite enough to convince us that changes in the relative level of the land and sea have occurred, and have been produced by movements of the solid crust and not by an alteration in the bulk of the ocean, even if no cases of such movement had actually come under observation. Our position will however be all the stronger if we can point to actual instances where movements of the land have been observed, and this we fortunately can do. The well-known case of the Temple of Serapis, at Puzzuoli near Naples, shows that within the historic period the spot where it stands was once beneath the sea ; was afterwards upraised and became the site of a temple older than the one whose remains are now standing ; was possibly again submerged and again upraised before the building of the present ruin ; was again let down till the sea rose at least some 20 feet above the pavement of the temple ; was again raised into dry land, and is now slowly sinking again.* Then again we have the case of the Scandinavian peninsula, where there is good reason to believe that within the memory of man the northern part of the country has been rising, perhaps at the rate of 2 or 3 feet in a century ; that the movement lessens as we go southwards till about Stockholm the land is stationary ; and that still further south motion is going on in the opposite direction and the land is slowly sinking, t It will be seen how a case like this, and it is not an isolated one, effectually disposes of any attempt to explain the phenomena we are considering by a lowering of the sea-level. We have distinct proofs of oscillations of level in our own country at no very distant period. Every here and there round the northern part of the island we find, at a height of from 20 to 30 feet above the present mean-tide level, a flat terrace stretching inland for a distance varying from a few yards to several miles, and bounded on the land- ward side by a line of bluffs which bears a strong resemblance to a sea- cliff. The subsoil of this terrace consists of Sand, Silt, and Shingle, occasionally enclosing shells and other marine remains and in some cases human implements and canoes. This terrace is evidently an old sea-beach, and it shows that the land at one time stood some 20 or 30 feet lower than it does now, and remained in that position long enough to give the sea time to cut a notch in the solid rock as a record of its former level, and to strew the floor of that notch with shore deposits. In the elevation that followed the land was raised to a greater height than at present, for we constantly find stretching out below the sea the remains of buried forests, the trees of which not only grew on dry * For details see Lyell's Principles, vol. ii. chap. xxx. ; Quart. Journ. Geol. Soc. iii. 186. t Lyell's Principles, vol. ii. chap. xxxi. Displacement of Rocks. 459 land, but could have attained their size and luxuriance only in situa- tions sufficiently far inland to be removed from the blighting influence of sea-breezes. We have therefore evidence of a time when the land was lower than now, of a subsequent upheaval which raised it above its present level, and in fact probably connected it with the continent of Europe, and afterwards of a depression which produced its present insular condition.* Submergence produced by a Polar Icecap. There is one possible means by which a change might have been produced in the position of the sea-level without any movement on the part of the land, that ought not to be passed over. We have evidence that there have been times when the climate of the polar and temperate regions became far more severe than at present, and it seems likely that these cold periods shifted from one hemisphere to the other perhaps several times over during a long lapse of years. It has been supposed that in consequence of this enormous accumu- lations of ice gathered, now round the northern and now round the southern polar regions, which reached down far into the temperate zones. Should such caps of ice ever form around the poles, their attraction would tend to draw the water of the ocean towards the pole around which they were placed, and so raise the ocean-level in the corresponding hemisphere. There is however considerable doubt whether this cause has ever really been in action. That there has been one, perhaps more than one, period of intense cold is beyond question ; but there is no evidence to show that there ever was a continuous cap of ice spreading away in every direction from the pole. There is distinct proof that during this period every region, whose configuration made it a good gathering- ground, became a great snow-field and a centre from which ice-sheets and glaciers were shed off, but all the known facts are dead against the idea of the northern regions having been ever swathed in one general covering of ice.t SECTION III. DISPLACEMENT OF THE ROCKS FROM THEIR ORIGINALLY HORIZONTAL POSITION. The instances we have given furnish proof of up and down move- ments by which rocks formed beneath the sea are raised above its level. But this is not all that has happened to them. In the case of stratified deposits we know that their beds must have been at the time of their formation approximately in a horizontal position ; there would * For other cases of oscillation of level see Geol. Mag. vol. viii. pp. 300, 430 ; Nature, i. 381. "t On this subject see Adheniar, Revolutions de la Mer (Leipzig, 1843) ; Croll, The Reader, September 2, December 2, December 9. 1865, January 13, 1866 ; Phil. Mag. 4th series, xxxi. 301 (Ap. 1866) ; Heath, Phil. Mag. 4th series, xxxi. 201, 323 ; Pratt, Phil. Mag. 4th series, xxxi. 172, 532 ; Figure of the Earth, 4th edition, p. 236 ; 0. Fisher, The Reader, February 10, 1866 ; and The Reader, January 20, February 24, March 17, 1866 ; and for a summary, Croll's "Climate and Time," chap, xxiii. 460 Geology. be exceptions when sediment was thrown down by currents in sloping layers, but these are unimportant in a general view, and speaking broadly, we may say that beds of sedimentary rocks were originally horizontal. But this is not the position in which in many cases we now find them, and hence we learn that rocks have been affected by other movements besides that of mere vertical elevation. A moment's reflection on the way in which sedimentary rocks were formed would be quite enough to convince us that they could not have been deposited in the inclined positions in which we often see them ; but one or two other facts leading to the same conclusion may be just mentioned. The surface of such beds often bear ripple-marks, rain-pittings, and the tracks of animals, which could not possibly have been impressed on them in their present highly-inclined position. We occasionally find embedded in rock the trunks of trees still rooted in the soil in which they grew, and inclined at the same angle to the vertical as the beds are to the horizon. We cannot suppose these trees grew in such an unusual position ; but if we suppose them to have sprung up when the beds were horizontal and to have shared in a subsequent tilting, their position will be satisfactorily explained. Again, if we examine a deposit of Shingle we find a tendency among the pebbles to arrange themselves with their flat surfaces and longer axes horizontal ; but wherever we find inclined beds of old Shingle or Conglomerate, the flat surfaces of the pebbles are parallel to the bedding, showing that since the former were deposited horizontally, the same must have been the case with the latter. We will now go on to consider the displacements rocks have under- gone from their originally horizontal lie, and define the terms used in describing them. Dip. Where strata have been tilted from a horizontal position, their inclination to the horizon is called the Dip. Thus in fig. 131, if Fig. 131. ABC be the surface of an inclined stratum, OBC a horizontal plane, AO a vertical straight line, and AD a straight line in the plane ABC perpendicular to BO, then the inclination of the plane ABC to the horizontal plane BOG is measured by the angle ADO. ADO is there- fore the dip of the bed ABC. Strike. The line BC, or the intersection of the inclined bed with a horizontal plane, is called its /Strike or Level line, and is described by its compass-bearing. Perhaps the simplest illustration of dip and strike may be given by Displacement of Rocks. 46 1 holding a board or slate in an inclined position in a trough of water. The intersection of the surface of the water with the slate is necessarily horizontal and gives the line of strike ; if a drop of water be placed on the slate it will run down the steepest line on it, and this is the line of dip. In practice a quarry partly filled with water is the best possible place for determining dip and strike. To put the definitions as shortly as possible, we may say that the line of dip is the line of greatest inclination that can be drawn on the surface of a bed ; the line of strike is the line of no inclination. Amount and Direction of Dip. In order to determine com- pletely the dip of a bed we must know two things, its amount and its direction. The line DO in fig. 131 is the direction of the dip. Its bearing may be taken by a compass and may be expressed in several ways. We may divide each quadrant between the four cardinal points into 90 and say that the direction is so many degrees east of north, south of west, and so on as the case may be. Thus if we write the direction S. 30 W., we mean that the line OD points in a direction 30 west of south. Or we may reckon our degrees all round the circle from up to 360. In this case, if the graduations start at the north point and are reckoned round towards the east, the direction of the dip of the bed mentioned would be described as 210, for the south point is 180, and 180 + 30 = 2 10. A third way, often used by the German geologists, is to divide the whole circle into twenty-four equal parts and call each an hour. Each division then contains 15. Due south is noon, and the bed in question would be said to be dip- ping to two hours after noon. English miners use a similar phraseology : they would say that the bed is dipping to two o'clock sun. If the dip were N. 30 E., the bed would rise to two o'clock sun. The amount of the dip may be expressed by the number of degrees in the angle ADO. Thus if this angle contains 18, we say that the dip is 18. In this case DO is very nearly three times AO, and we may also express the amount of dip by saying that it is 1 in 3. What comes to the same thing is 4 inches to a foot, 1 foot to a yard, 587 yards to a mile. When the number of degrees in the angle of dip is known, we may express the dip as one in so many and vice versa in several ways. If 6 be the number of degrees in an angle of 1 in n, then n = cotan 6. (1) Or we may carefully lay down the angle on paper with a protractor, from any point in one bounding line let fall a perpendicular to the other, measure the perpendicular and base of the right-angled triangle so formed, and then base perpendicular For angles not greater than 20 the following formula gives n approxi- mately. 60 ,. *-!.-/- ( 3 ) 462 G In the following table the values of calculated from (1) are given correct to two decimal places up to a dip of I.V. For angles larger than 45 it is convenient to say that the bed rises or falls so much in 1, instead of saying that it rises or falls 1 in so many; the corre- sponding value of n is found opposite the angle in the right hand column. Thus if the dip be Cn , the bed f ; ,l| s ] -7.", |Wt in a foot, or 17-:? feet in 10 feet, or 173 feet in 100 feet, or 1-73 x 1760 = 2044'S yards in a mile. In the columns on the left the approximate values of n obtained from (.'5) and the percentage of error belonging to each are given, correct to one decimal place, up l> L'I! : here tin- error exceeds 10 per cent., and for larger angles the error is so large, that the formula can be no longer used. Error o/o M it N 6 Krror 0/0 en e M i) 66' 60 18 26' 8 47 60" 1 57-29 89 8-8 ^'2 19 2 '90 71" 1 55' 80 !)-J 8 20 2-75 7ir 4 : 8 30" 2 28-64 '.'7 2-9 21' 2-61 69' 2 52' 20 K)-J 27 2,2 2-48 68 4 ; S 2()" :; 19-08 87 28 2-86 Q7 3 49' 15 'Jl 2 '25 .;.; 4 ; 9 15 4 l 1-80 86 a 25 2-14 66 4 4,7 12 2<; 2'05 <;i fro 12 f,' 11-48 85 26 88' 2 f, 1 2" 10 27' J 96 5-1 10 <> ;> ;,i 84" 28 ' '88 62 i; 20' 9 2tf SO (>1 " 5-2 8-6 7 8-14 88 80 fl '73 00 7 7' 8 :; i 66 59 5-4 7-5 7-12 82 82 60 8 8' 7 :;:, 6i i ; 66 9 6-31 81 84 48 56 '.' -IT 6 43 .,., 5 '-8 6 10 5-67 88 54 6 5 '4 6 11 :,-l t 7'.' 87 83 58 ... ... 1 i 1 :' 28 52 o" l-j 7" 7S'' 89 28 51 ;; 4-6 ]:r 88 77 10 19 50 6-9 43 M 01 7'i n ]'> 49 i r -' 42" 11 4H" 7-2 4 I.V 8-78 1.-, 07 17 7-5 8-8 10* 8-49 74' 11 04 40 s 8-0 1 7 ' 8 '27 7'6 10 8'3 18 r>' Measurement Of Dip. If wfj liave, tho surface of a bed laid bare, we can determine, by an instrument for measuring angles called a Clinometer, the direction of a lovol line on the bod, and then, by measuring the inclination along a line at right angles to the level line, we get the amount of the dip. Or the dip may bo measured on ;\. vertical face of rock, such a.s a cliff or the wall of a quarry ; but in such a case, in order to determine its full amount, it is necessary that the face Displacement of Rocks. 463 should be, like AOD (fig. 131), perpendicular to the strike; if for instance a measurement was made on faces such as ABO, A CO, the observed angles would be less than the full amount of the dip. In practice it often happens that we cannot find a vertical face run- ning along the full dip, but we can generally get measurements of the apparent dip along two faces making a large angle with one another, from which the amount and direction of the full dip may be determined either by the formula given in the note below,* or by the following graphical method. Let OA, OB be the bearings of the two faces on which dips are measured, and let the dip along OA be 1 in m, along OB 1 in n. Take in OA Oa = m, and in OB Ob = n: join ab, and draw OC per- pendicular to ab. Then OC is the direction of the full dip, and the amount of the full dip is 1 in OC. If from we draw any line OE meeting ab in D, then the dip along a vertical face running along OE is 1 in OD. Hence if we know the full dip and want to know the dip along any vertical face we proceed thus. Let the full dip be 1 in d, arid let OC be its direction, OB the direction in which the dip is required. Make OC = d, draw Cb perpendicular to OC cutting OB in b. Then the dip along OB is 1 in Ob. Or we may use the formula given in the note below. t A table calculated from this formula will be found in the Memoir on the South Staffordshire Coalfield (Memoirs of the Geological Survey of England and Wales), p. 215. Both these methods and calculation from the formulae may occa- sionally be usefully employed, but in many cases a little practice will enable us to estimate the full dip from two observations quite near enough for all practical purposes. It must be borne in mind that the surfaces of beds are far 'from being true planes, so that the angle of dip can seldom be determined within a degree at any given spot ; and even in the case of very evenly-bedded rocks where this might be pos- * Let d and d' be the two observed dips ABO, ACO (fig. 131), D the full BOC-2a, BOD-a-b, COD-a + b, then sin (d d) Tan b = -. ^ =-' cot a sin (d + d) Tan D= 2 cos a cos d cos d' cos b t Let be the full dip, 6' the dip along OB, COB = (fig. 132). Tan ff = Tan 6 cos 0. Then 464 Geology. sible, the dip is never constant, but varies in amount and direction slightly one way or another from spot to spot. In measuring dip then it is desirable to make as many observations as possible, and, if these do not differ much from one another, to take their mean as represent- ing the average dip. Problems connected with Dip. By means of the methods just given a number of useful practical results may be worked out, of which the following is an instance. In fig. 132 0, A, and B are three pits sunk to the same bed of Coal, which is supposed to be dipping regularly at the same amount and in the same direction throughout the surrounding country. We have given OA = 2400 yards,. OB = 880 yards and Depth of Coal in Yards. Height of Ground in Feet above Sea-level. At . . . 130 150 At A . . . 200 180 At B . . . 158 114 AtE. . . 141 To find how deep the Coal will be at the point E, OE being 1953 yards Apparent fall of Coal from to A = 200 - 130 = 70 yards. Subtract rise of ground =30 feet =10 ,, True fall of Coal from to A = 60 2400 Dip along OA = -^- = 1 in 40. Apparent fall of Coal from to B= 158 - 130 = 28 yards. Add fall of ground = 36 feet =J^2 True fall of Coal from to B =~40 Dip along = l in 22. In OA take 0a = 40and in OB 06 = 22, join ab. Let OE cut ab in D ; then it will be found that 0Z> = 21. Therefore the dip along O^is 1 in 21. Fall of Coal from to E Depth at ^Y = 93 yards. = 130 Depth at E, if ground were flat, would = 223 Subtract fall of ground from to E=9 feet = 3 Depth of Coal at E = 220 Measurement of Thickness of Strata. If we saw across a plank in different directions, the breadth of the edge will depend on the direction in which the cut is made, but when we speak of the Displacement of Rocks. 465 thickness of the plank we mean the distance between its bounding surfaces measured in a direction perpendicular to those surfaces. So the thickness of a group of strata, the beds of which are parallel to one another, is the distance from the top of the top bed to the bottom of the bottom bed measured perpendicularly to the planes of bedding. This is easily ascertained when the amount and direction of the dip, and the difference in level between the points where the top and bottom beds come to the surface are known. Fig. 133. Let a vertical plane running along the full dip cut the bottom and top beds of the group of strata whose thickness we wish to find in AB and CD ; and let the same plane cut the surface of the ground in AC.* From C draw CE perpendicular to AB. Then GE is the thickness required. CE may generally be found nearly enough for all practical purposes by protraction, or it may be calculated from the following formula. Draw AF horizontal, CF perpendicular to AF '; and let the full dip, FAB, = 6 ; the slope of the ground, FAG, =<. Then CE= thickness = CA sin or = cos 9 (1) (2) If we measure the distance on the ground, we use (1). If we take the distances off a map, the distance between A and C will be repre- sented by AF. For gentle slopes the difference is scarcely appreciable, but it is important on steep hillsides. It may happen however that we cannot succeed in getting the position of A and C along a line running in the direction of the full dip. In this case we may still construct our figure in the same way as in fig. 133, but the line EC will be greater than the true thickness. If be the full clip, a the angle between AF and the direction of the full dip, t the true thickness, t' = EC the apparent thickness, then 1-sin 2 6>sin 2 a (3) * This figure represents what we should sea in the side of a deep ditch dug across the country. It is called a Geological Section. 2ft 466 Geology. We may therefore first find t' as in the former case, and then obtain the true thickness by means of this formula. For a given dip the error increases with the obliquity, and for a given obliquity it increases with the dip. But if the dip be small the error is scarcely appreciable however oblique the section may be ; and if the section be not very oblique, the error is correspondingly small whatever be the dip. If however both dip and obliquity be large, the error becomes very serious. For instance if the dip be 60, the error when a = 60 is 51 per cent., and for sections very oblique will amount to nearly 100 per cent. The following table will help us towards deciding when it is necessary to apply a correction. 6 or a Maximum Error per cent. in determining the Thickness. 5 .... 0-4 10 .... 2 15 .... 4 20 . . . .6 25 ... 10 30 .... 15 35 . . . .22 40 .... 31 45 .... 41 50 .... 56 55 .... 74 60 .... 100 The table serves a double purpose. If the angle in the first column represent dip, then the number in the second column shows what will be the largest error that can be made by an oblique section with that dip. This will be the error in a section running along the strike. Secondly, if the angle in the first column be the obliquity of the section, the number in the second column shows what will be the largest error that can be made in a section with that obliquity. This will be the error made when the beds are vertical. For instance if the dip be 5, the error for the most oblique section does not exceed '4 per cent. ; for sections not very oblique it will be much smaller : the best measurements that we can obtain in geological work are usually so rough that such an error is quite inappreciable. Again if the dip is 20, the error cannot exceed 6 per cent, however oblique the section may be. Secondly, if the obliquity of the section be 5, the error cannot exceed '4 per cent, however high the dip may be ; if the obliquity be 10, the error cannot exceed 2 per cent. With low dips the corre- sponding errors will be much less. But if both dip and obliquity be both high the errors become serious. It follows then that with low dips we may use any sections we like, however oblique they may be, to determine thicknesses without any appreciable error ; and with high dips we shall not commit any serious mistake if the section be not very oblique. But if both dip and obli- Displacement of Rocks. 467 quity be large, we must correct thicknesses determined from oblique sections by means of formula (3). In fig. 133 draw FG perpendicular to AJ3, and let CF produced cut AB in B. Then FG is the thickness of a group of strata the breadth of whose outcrop measured along the full dip is AF, and FB is the depth to which the bed AB descends in a horizontal distance AF measured along the full dip. If we call the values of FG and FB, when AF '= 1000, t and d, the following table gives the values of t and d correct to one decimal place for different values of the dip (6). 9 t d 6 t d 1 17-5 17-5 46 719-3 1,035-5 2 34-9 34-9 47 731-4 ,072-3 3 52-3 52-4 48 7431 ,110-6 4 69-8 69-9 49 754-7 ,150-3 6 87-1 87-5 50 766-0 ,191-7 6 104-5 1051 51 777-1 ,234-8 7 121-9 122-8 52 788-0 ,279-9 8 139-2 140-5 53 798-6 ,327-0 9 156-4 158-4 54 809-0 ,376-3 10 173-6 176-3 55 819-2 ,428-1 11 190-8 194-4 56 829-0 ,482-5 12 207-9 212-6 57 838-7 ,539-8 13 225-0 230-9 58 848-0 ,600-3 14 241-9 249-3 59 857-2 ,604-2 15 258-8 267-9 60 866-0 ,732-0 16 275-6 2867 61 874-6 ,804-0 17 292-4 305-7 62 882 -9 ,880-7 18 309-0 324-9 63 891-0 ,962-6 19 325-6 344-3 64 898-8 2,050-3 20 342-0 364-0 65 906-3 2,144-5 21 358-4 383-9 66 913-5 2,246-0 22 374-6 404-0 67 920-5 2,355-8 23 3907 424-5 68 927-2 2,475-0 24 406-7 445-2 69 933-6 2,605-0 25 422-6 466-3 70 939-7 2,747-4 26 438-4 4877 71 945-5 2,904-2 27 454-0 509-5 72 951-1 3,077-6 28 469-5 5317 73 956-3 3,270-8 29 484-8 554-3 74 961-3 3,487-4 30 500-0 577-4 75 965-9 3,732-0 31 515-0 600-9 76 970-3 4,0107 32 529-9 624-9 77 974-4 4,331-4 33 544-6 649-4 78 9781 4,704-6 34 559-2 674-5 79 981-6 5,144-5 35 573-6 700-2 80 984 -8 5,671-2 36 587-8 726-5 81 987-7 6,313-7 37 601-8 753-6 82 990-3 7,115-3 38 615-7 781-3 83 992-5 8,144-3 39 629-3 809-8 84 994-5 9,514-3 40 642-8 839-1 85 996-2 11,430-0 41 656-1 869-3 86 997-6 14,300-6 42 669-1 900-4 87 998-6 19,081-1 43 682-0 932-5 88 999-4 28,636-2 44 694-7 965-7 89 999-8 57,289-9 45 707*1 1000-0 90 1000-0 468 Geology. Outcrop. The line along which any bed cuts the surface is called its Outcrop, Outgoing, or Bassett. A bed is also said to "come to day " or to " come to grass " when it reaches the surface. There are three cases in which the outcrop of a bed coincides with its strike. 1. If the surface of the ground be horizontal. 2. If, without the whole surface being horizontal, that portion of it along which the bed crops out is horizontal. 3. If the bed be vertical. In all other cases the outcrop winds about with the inequalities of the ground, and the different ways in which the outcrops of moder- ately-inclined beds bend round hills and run up and down valleys are somewhat puzzling to understand. A small set of models constructed after a pattern devised by the late Mr. Sopwith may be bought and will be found useful in explain- ing this matter. But the student will derive much more profit from constructing models for himself. Let him take some modelling clay and mould it into a surface broken by hills and valleys according to his fancy. Then let him thrust a plate of sheet tin into the clay, holding it sometimes gently and sometimes steeply inclined to the surface, sometimes sloping up, sometimes sloping down, and sometimes sloping across the valleys. The cuts made on the surface of the clay will represent the outcrops of beds dipping in the direction in which the plate is held, and the course of the outcrop in each case will be seen at a glance. Another very excellent exercise is to track out the course of outcrops over a map on which the height of the ground is shown by contour lines, by the following graphic method. Contour lines are lines drawn on a map through all those points which are at the same height above a given datum. We will now give an illustration of this method. Fig. 134 is a contoured map, the fine dotted lines being contour lines. All the points on the line or lines against which 100 is written are 100 feet above the sea-level; every point on the line numbered 125 is 125 feet above the sea-level ; and so on for all the contour lines, which go on ascending by 25 feet. It will be seen that along the strong clotted line WT all the contour lines run up into sharp-pointed Vs. This is obviously the line of a valley, for if we walked along any line EFG crossing WT, the contours show that we go down-hill from E to F, and up-hill from F to G. We have no name in English for the line WT which runs through the points of all these V's, but the Germans call it a Thalweg (Valley path). Again the strong dotted line WS which runs from the head of the valley across the highest ground cutting the contour lines at right angles, evidently divides the water which runs into the valley WT from that which runs into the next valley. It is called a Water- parting. Now suppose at a point P on the 200 contour we find a bed of Coal cropping out and dipping in the direction PQ at 1 in n. In PQ take PA 1 = A^ 2 = A 2 A 3 = 25 n feet. From P draw a line PF perpendicular to PQ to cut the 200 contour on the opposite sides of the valley in Displacement of Rocks. 469 Fig. 134. OuxcRor OF BED DIPPING UP VALLEY. 470 Geology. P'P". From A l draw A 1 B 1 B\ parallel to PP' cutting the 175 con- tour in B l and B\: from A 2 draw A 2 B 2 B' 2 parallel to PP' cutting the 150 contour in J3 2 and E\\ and so on. Then B^ B\, B^ J5' 2 , etc., will be points on the outcrop of the bed. For at P the bed is 200 feet above the sea. In the distance PA l PA it falls - 1 feet, or 25 feet. Therefore at A l the bed is 175 feet above the sea. A^ is the direction of the strike or level line. Therefore the bed is 175 feet above the sea all along A^B^. And the surface of the ground is 175 feet above the sea at B^ Therefore the bed comes out at B^ In the same way we can show that it comes out at B<>, B 3 , B' 3 , B'.^ B\, P. We shall find the outcrop by drawing a line through these points. It is the thick black line. Just in the same way if we take in QP produced, PC l = C 1 C. 2 = etc. = 25 n, and draw C J D l D' l perpendicular to QP to cut the 225 contour in D! and D\, and C 2 D 2 D' 2 to cut the 250 contour in D 2 and D' 2 , A, A, D' 2 , D\ are points on the outcrop. The outcrop evidently bends back towards the water-parting and makes a V where it crosses the valley. The dip is up the valley and the V also points up the valley. It is evident that the steeper the dip, the shorter are the lines PA lt A ^2, etc., and consequently the shorter the V. When the bed is vertical A^ A 2 , etc., all coincide with P, and the outcrop runs in a straight line parallel to the strike across hill and valley. Between P' and P" the 200 contour is nearly straight and parallel to the strike. The outcrop will evidently coincide with the strike between these points. This will explain the second case in which out- crop and strike coincide ; the ground is not flat, but the contour line from P to P" is parallel to the strike. Next take the case where the bed is horizontal. Here the outcrop obviously runs parallel to the contours: in fig. 134 the outcrop of a horizontal bed cropping at P will coincide with the 200 contour. Next take the case where the bed dips down the valley but at a less angle than the slope of the Thahveg. Fig. 135 shows that here also the outcrop forms a V pointing up the valley. The construction is exactly the same as in fig. 134 except that PA^ PA 2 , etc., are taken off to the rise, and there is only one point requir- ing explanation. Let QP produced cut the successive contours in a^ 2 , %, etc. It is clear that PA l must be greater than Pa^; for the ground rises 25 feet in Pa l and the bed rises the same distance in PA^ and the ground rises faster than the bed. Lastly we have the case shown in fig. 136, where the dip is down the valley and at a greater angle than the slope of the Thalweg. Here PA 1 is less than Pa^ PA 2 less than Pa. 2 , etc., and the outcrop forms a V pointing down the valley. Displacement of Rocks. 471 In this figure a second valley is shown at the top on the left side. By prolonging the construction lines P'P^ A^B^ etc., to cut the corre- A* 200 Fig. 135. OUTCROP OF BED DIPPING DOWN VALLEY AT LESS ANGLE THAN SLOPE OF THALWEG. spending contours in this valley, the outcrop has been traced across the water-parting between the two valleys and followed down into the second valley. In figs. 134 and 135, and in the small valley on the left in fig. 136, the direction of the dip is nearly parallel to the Thalweg. The two branches of the V here lie symmetrically with regard to the Thalweg. 4/2 Geology. In the larger valley in fig. 136 the direction of the dip is very oblique to the Thalweg. On the side of the valley where the direction of the dip makes a small angle with the contours, the outcrop runs down gradu- Fig. 136. OUTCROP OF BED DIPPING DOWN VALLEY AT GREATER ANGLE THAN SLOPE OF THALWEG. ally in a long line ; on the opposite side where the direction of the dip cuts across the contours, the outcrop also runs up sharply across the contours. Breadth of Outcrop. On looking at a geological map in which the different subdivisions of the rocks are traced out in detail, w r e notice that the belts of colour representing these different subdivisions Displacement of Rocks. 473 vary very much in breadth from place to place. These variations in breadth are not owing to changes in the thickness of the subdivisions, they may and do occur in cases where the thickness undergoes no change. Now a geological map shows the rock that is at the surface at each spot upon it, so that these belts of colour are outcrops. Their breadth at any spot depends partly on the dip of the beds and partly on the shape of the surface of the ground. Fig. 137 will explain why this is so. The figure is a geological section across a country in which both dip EF G a b c d ef g h Fig. 137. SECTION SHOWING OUTCROPS OF UNEQUAL BREADTH. and shape of surface vary from point to point. Among the rocks there are four beds, shown by dark bands, each of which has exactly the same thickness. But the breadths of the outcrops of these bands are all different. The outcrop AB is much broader than CD : the ground in both cases is nearly flat, but the dip at CD is much steeper than at AB. Again EF is narrower than CD, though the dip at EF is less than at CD : but the ground at EF is sloping very steeply in the opposite direction to the dip and this narrows the outcrop. Lastly at GH we have a bed dipping at a gentle angle and the ground sloping at nearly the same angle in the same direction ; the result is the broadest outcrop of all. The differences in breadth are perceptible on the surface of the ground ; they become still more conspicuous on a map, for there we see not the actual breadths of the outcrops but their projections, ab, cd, ef, gh, obtained by letting fall perpendiculars from their extremities on a horizontal plane. The difference between the actual outcrop and its projection is largest in the case of the steep hill- side at EF. Undulations and Contortions. It very rarely happens in nature that the dip of the beds is constant for any long distance ; it frequently varies both in amount and direction from point to point. When the changes are small and gentle, a series of easy rolls or undulations is produced. In other cases the foldings are excessively sharp and sudden, and the beds are then said to be con- torted. 474 Geology. Undulations and contortipns may be present on a small scale with- out interfering with the general dip of the beds; thus in fig. 138 the beds on the left have been thrown into a series of broad gentle folds, and towards the right have been puckered up into sharp curves, but Displacement of Rocks. 47 S preserve, in spite of these lesser irregularities, a general dip from the right towards the left. A case of violent contortion on a small scale is given in fig. 139, which is a natural section of Shale and thin Sandstones in North Staffordshire. Fig. 140* shows another case, where beds of solid Limestone have been bent to the form of an inverted W. It is in mountain-chains * Roughly reduced from a photograph issued by the Geological and Polytechnic Society of the West Riding of Yorkshire. 4/6 Geology. that such foldings and crumplings occur on the grandest scale, the beds sweeping up and down in curves of enormous radius, and bending in and out in countless and most abrupt plications. The following terms are used in connection with the larger undula- tions. Anticlinal and Synclinal ; Dome and Basin. When the beds have been bent into the form of long ridge-like arches, these are called Anticlinal* or Saddles and the hollows between them Synclinals or Troughs. In both Anticlinals and Synclinals the line in each bed, along which the change in the direction of the dip takes place, is called the Anti- Fig. 141. MAP OF AN ANTICLINAL. clinal or Synclinal Axis of that bed ; and the planes containing all the axes of an anticlinal ridge or synclinal trough are called Axis planes. If the beds dip away in all directions from a centre, they are said to have a quaquaversal dip or to be domed ; and if they dip everywhere towards a centre, they have a centroclinal dip or form a basin. An anticlinal runs on as long as its axes are horizontal or only gently inclined ; it is brought to an end when they begin to bend down sharply. A complete anticlinal consists of a long ridge ter- minated at each end by the half of a dome ; in fact, an anticlinal is nothing but an elongated dome. A synclinal, in the same way, Displacement of Rocks. 477 is a long trough with half a basin at each end, or an elongated basin. Anticlinals and synclinals are however often abruptly truncated by the dislocations known as faults. Anticlinal. A sketch-map and sections of an anticlinal ridge are given in figs. 141, 142, and 143, the arrows showing the direction of A 5 43 45 B Fig. 142. SECTION ALONG THE LINE AB IN FIG. 141. the dip. In the southern part of the map the beds are thrown off both to the east and west from a central line or axis, as shown in the section, fig. 142; their outcrops wind about with the inequalities of the ground, but keep on the whole a northerly and southerly strike. On the east, owing to the smaller dip and the flatness of the surface, the outcrop of No. 3 is much broader than on the west side. Towards the north however the regular dip to either side becomes gradually exchanged for a dome-shaped bedding, the strata fall away in all directions, and the anticlinal is terminated by a half-dome, around which the easterly and westerly outcrops, bend till they join one another. The second section shows this change in dip ; its southern part runs along the anticlinal axis, and the beds are therefore flat, but Fig. 143. SECTION ALONG THE LINE CD IN FIG. 141. towards the north they bend down, and the successive members come on one over the other, just as along the flanks of the arched area on the south. When, owing to the axes of an anticlinal becoming inclined, the beds sink in one after the other along its crest, they have been said in homely but very expressive language to " nose in," the pointed nose- like shape of their outcrops naturally suggesting the phrase. An anticlinal whose crest has been sliced off by denudation, as in fig. 142, is called by the Germans an "air-saddle" (Luftsattel). Dome. As an illustration of dome-shaped bedding or quaqua- versal dip, I have chosen Simon's Seat, a conspicuous hill in Wharfe- dale, the structure of which has been kindly explained to me by my friend Mr. J. R. Dakyns. Figs. 144, 145, and 146 show a sketch- 478 Geology. plan of the hill and two sections across it. The rocks of which it is composed are 5. Gritstone. (Top.) 4. Shale. 3. Gritstone. 2. Shale. 1. Limestone. (Bottom.) The beds, as shown by the arrows on the plan and by the sections, dip away in all directions from a centre, around which their outcrops run x D c P. Fig. 144. GEOLOGICAL SKETCH-MAP OF SIMON'S SEAT. Scale, 1 inch to a mile. in concentric curves. Thus in the middle we have a patch of No. 2 ; this is enclosed by a belt of No. 3, which is in its turn surrounded by a ring of No. 4. The outermost band, formed by the outcrop of No. 5, is not complete, a portion having been removed by denudation. 23 Fig. 145. SECTION ALONG THE LINE AB i^ FIG. 144. The Limestone does not actually come to the surface in the centre of the dome, but it is brought up by a change of dip on the north-west side of the hill and is there seen to underlie the Shale No. 2, and it must therefore form the great mass of the interior of the hill. Displacement of Rocks. 479 Very beautiful illustrations of domes are sometimes seen on the sea- coast, when the waves have planed away the summit and laid bare a horizontal section perpendicular to the axis of the dome. A good instance occurs near Berwick-on-Tweed. The rocks consist of alterna- Fig. 146. SECTION ALONG THE LINE CD IN FIG. 144. tions of hard Limestones or Sandstones and soft Shales, and at one spot they have been thrown into a dome almost perfectly circular in outline. The top of this has been sliced across in the way just mentioned, and the concentric rings formed by the outcrops of succes- sive beds are most distinctly exhibited. The structure comes out with singular clearness, because the outcrop of each hard bed stands up above those of the softer measures on each side ; and thus there is produced a set of concentric circular low reefs, separated by grooves in which water remains after the fall of the tide has laid dry all around. Synclinal and Basin. A case of a synclinal trough is shown in figs. 147, 148, and 149, which are a map and sections of a part of a long synclinal in North Staffordshire, known as the Goyt Trough. The general synclinal lie over the greater part of the map is well shown by the beds 1, 2, and 3, whose outcrops, in spite of windings due to inequalities in the ground, strike persistently north and south, and which dip inwards on both sides towards a central axis, as shown in the second of the sections. The trough however is subdivided by lesser undulations into several minor basins. The first section runs across one of these. Here the general easterly dip of the western half of the trough is exchanged about the centre for a dip to the west, and a smaller interior trough produced. The beds then roll over and resume their easterly dip up to a fault, beyond which they put on the westerly dip which prevails along the eastern half of the trough. Another well-marked basin, in the centre of which a detached patch of the bed 6 nestles, is seen towards the southern end of the map. The beds here, as shown by the arrows, dip on every side inwards towards a centre, and the outcrops run in concentric rings round the central area of the highest member. This trough is terminated on the south by a half-basin, the simple synclinal lie being exchanged for inclinations to the north-east, north, and north-west, and the outcrops on either side wind round till they meet. Parallelism of Anticlinals. It frequently happens that anti- clinal ridges show a tendency to run rudely parallel to one another over large areas. 480 Geology. Classes of Anticlinals. Anticlinals may be distinguished ac- cording to their transverse section into three classes, examples of which are seen in fig. 150, which is a general section after Professor H. D. Rogers, across the Appalachian Mountains. Displacement of Rocks. 481 In the first class the beds on opposite sides are equally inclined to the horizon, and the axis plane is therefore vertical. This symmetrical form is common among gentle undulations of considerable width, such as are seen in the left of the section. In the second class the beds are more steeply inclined on one side than on the other, so that the axis plane is no longer vertical ; foldings of this kind occur in the middle of the section, and it will be noted that in all of them the steeper side of the arch faces the west. In the third class the rocks are doubled under on the steeper side of the fold, so that on that side the beds which before the folding were uppermost, now plunge down beneath those which originally lay below them, and what is called "Inversion" is produced. The axis plane here is inclined to the horizon, but at a smaller angle than the beds on the steep side. This form prevails towards the right hand of the section. These three forms are seen in the instance before us to pass into 2n 482 Geology. one another, and the theory of their formation will be touched on shortly. Inversion. Instances of inversion of the beds, such as that which occurs on the steep side of the anticlinals last mentioned, are not un- common, specially in intensely-contorted mountain regions. A simple case is shown in fig. 151, which is a section in the neighbourhood of Pembroke. There are three groups of rocks. 3. Carboniferous Limestone. (Top.) 2. Lower Limestone Shale. 1. Old Ked Sandstone. (Bottom.) There is ample evidence in the neighbourhood that, where they have not been disturbed from their original position, the three groups lie one on the other in the order indicated above ; and they are found in their normal position on both sides of the synclinal trough on the south. On the north side of the anticlinal arch which follows however Fig. 151. INVERTED BEDS NEAR MILFORD HAVEN. Scale, l inches to a mile. they have been so completely folded over, that the Old Red Sandstone is at the top and the Limestone Shales and Limestone dip under it, so that an observer who had seen only this end of the section would be led by it to believe that the Old Red was the uppermost and the Limestone the lowest of the three groups, whereas exactly the reverse is the case. * A case of more violent inversion is shown in fig. 152, which is a section in the eastern part of the Jura.t The rocks when undisturbed, as at the northern end of the section, occur in the following order. 7. Fresh-water Marl. (Top.) 6. Nagelfluh. 5. White Jura Beds. 4. Brown Jura Beds. 3. Lias. 2. Keuper. 1. Muschelkalk. (Bottom.) * For another case of Inversion see Memoirs of the Geological Survey of Great Britain, vol. ii. part.i. pp. 153, 154. t Copied from Beitrage zur Geol. Karte der Schweiz, vol. iv. For other cases of startling inversion see Der Glarnisch, ein problem Alpinen Gebirgsbanes, Dr A. Baultzer. Zurich, 1873. See also the latter author, Neues Jahrbuch, 1876, p. 118, 1878, pp. 26 and 449, and Zeit. d. Deutsch. Geol. Gesell. xxx. 267; and Heim, Mechanismus der Gebirgsbildung, Tab. vii. Profil xiii. and xvii. Displacement of Rocks. 483 On the south a very sharp fold occurs, by which the beds have been thrown over, till the lowest member, the Muschelkalk, has come to lie at the top, and the other subdivisions appear beneath it in an order exactly the reverse of the above table. The reader will realize the enormous amount of displacement and denudation neces- sary to bring about this result if he will endeavour to put the beds back into the position they must have had before the folding took place. This has been done for a part of the section in fig. 153, where the letters ABC show what was the original position of the points denoted by the correspond- ing letters in fig. 152. The five lowest groups must have been to some extent folded and denuded before the forma- tion of the Nagelfluh began, because the latter does not everywhere rest on No. 5, but is at different points in contact with Nos. 4, 3, and 2. On the surface so formed ISTos. * 6 and 7 were afterwards laid down in hori- zontal beds, and fig. 153 shows what must have been the rela- tive position of the several groups when this step of the process was com- pleted. Then ensued a period of disturbance, by which the contortion and inversion were produced. The nature of the displacement will be more fully seized on if we fix our at- ^ tention on any one bed singly, say the band of Keuper to the north of C. This was dipping gently to the north before disturbance, as in fig. bb 4 8 4 Geology. 153. It must have been gradually tilted till it became vertical, and then actually pushed over so as to make it slope in a direction exactly opposite to that it had to begin with ; in fact the angle through which it has been turned is very nearly two right angles. While this crum- pling went on denudation was at work, and by its action all the sheet of Nagelfluh and Marl has been carried away except that portion which is squeezed into the middle of the fold, where it is protected by the beds that have been bent over it. When we see these startling results as they are now, they look at first sight almost beyond reasonable explanation ; but if we try in imagination to put back the rocks into their original position, to follow them through the successive foldings they have undergone, and bear in mind at the same time how much has been removed by denuda- tion, we are able to realize some at least of the steps of the process, though the forces and the machinery by which the movements were produced may still be beyond our grasp. In fig. 154, for instance, the Fig. 154. dark portion represents a section across a mountain-chain, on the flanks of which inversions are repeated over and over again. Cover over with a bit of paper cut to shape the lighter part of the diagram, and see what we could learn about the order of the beds from the darker part, that is from the mountain -side as it now stands. Take only two beds, the dotted one and the one marked by a thick black line ; at the sum- mit the first overlies the second, a little way down they occur in the reverse order, and still further below they come back to their original relative position ; and these changes are again repeated lower down. In such a section no one could say which was top and which was bottom, and the true sequence of the beds could be ascertained only by following them to some district where they are less disturbed. And if we think only of what we can now actually see, that is if we still keep the paper covering on, we are puzzled to imagine how this repetition could have been produced. But if we take away the paper cover, we then see how matters stood before the ground had its present shape given to it, and our difficulties are materially lessened. At one time the surface may have been in some such position as AB, and the rocks beneath it had been puckered up in a series of zigzag folds. Then out of this block of crumpled strata denudation carried away Displacement of Rocks. 485 everything down to the uneven surface CD, and so carved out the mountain-chain. When we try to make out the structure of such ground we are at first bewildered, because we see only the portions of the folds that have survived ; but our difficulties vanish when we com- plete each fold by restoring the portion which has been carried away, and so are able to understand how what are now isolated portions of each bed were once connected. We have for distinctness' sake spoken of the folding of the beds and the carving out of the mountain-chain as two independent operations, the first of which was finished before the second began. In reality denudation was going on at the same time as the crumpling, and it con- tinued to act after the latter process had come to an end. It is clear that, unless great caution is exercised, fatal mistakes might be made on many points of geological importance in very con- torted districts. Perhaps the most serious risk we run is of vastly over-estimating the thickness of the beds. In fig. 154 for instance as we walk along the saddle-shaped summit of the mountain the beds all seem to dip in the same direction and at the same angle. If we con- fined our observations to the surface alone, we should certainly carry away the impression that in passing from right to left we had been continually coming on to higher and higher beds, and we should be very liable to conclude that a line drawn perpendicular to the bedding from one end of the section to the other gave the thickness of the group of strata. In reality this would give a thickness at least four times too large. With the whole section before us of course there is no danger of such a blunder, but on the ground these sharp tilted folds often give scarcely a hint of their existence, and what we see looks like a succession of beds all dipping steeply in the same direction. Again if the strong black line were a Coal-seam, we might easily fall into the mistake of fancying there were six different seams, because we found Coal cropping at six spots, the searn at each crop dipping appar- ently in such a way that it must pass under the next outcrop. In reality all the outcrops belong to the same seam. We may guard against such slips by picking out a bed or beds of marked character and watching whether they are repeated. In the case of the Coal a careful observer could hardly go wrong ; at three out of the six outcrops it is inverted, and the inversion would be shown at once by the Seatstone being above the Coal. Outlier and Inlier. Tilting and bending, combined with sub- sequent denudation, have often -resulted in the production of isolated patches of rock, and of these there are two kinds, Outliers and Inlier -s. In an outlier the detached mass is surrounded on all sides by beds geologically beloiv it, in an inlier by beds geologically above it. Instances of outliers are seen in fig. 147, where three detached patches of the bed No. 5 occur in the northern part of the map, across one of which the section on fig. 148 is carried. Towards the south end of the same map a larger basin-shaped outlier of the bed No. 6 is seen, which is crossed by the section in fig. 1 49. 486 Geology. Again in figs. 183 and 184, p. 519, there are two outliers of the bed d and one of the bed b. An example of an outlier is given in figs. 155, 156, and 157, which Fig. 155. GEOLOGICAL SKETCH-MAP OF SHUTLINGSLOW. Scale, 1 inch to a mile. Dotted lines are faults. are a sketch-map, section, and view of a hill called Shutlingslow, a very conspicuous object in the moorlands of North Staffordshire. The outlier, which is formed by the isolated patch of the bed No. 5, is very small, but the rock of which it is composed is a hard massive Gritstone, and in consequence of this character it has given to the Faulted Inlier of 1 23 45 Fig. 156. SECTION ALONG THE LINE AB IN FIG. 155. summit an outline so bold and characteristic that a trained eye at once recognises from the shape of the. peak its general geological structure. Outliers are the remains of a broad sheet of the rock which once spread far and wide over the country, but the greater part of which Displacement of Rocks. 487 has been carried away by denudation. Thus in fig. 1 84 the outlier Fig. 157. VIEW OF SHUTLINGSLOW. of the bed b on the hill to the right must once have been connected Fig. 158. GEOLOGICAL SKETCH-MAP OF CRICH HILL. Scale, 1 inch to a mile. Dotted lines are faults. with the strip of the same bed which crops out along the flanks of the 488 Geology. hill to the left, as shown by the dotted lines ; and the bed d, of which only two outliers remain, was once equally extensive. In many cases the dislocations called faults, which will be described in the next section, have contributed to the formation of outliers Thus in figs. 158 and 159 we have an outlier of the bed No. 5, bounded on the east side by a fault, which has upheaved and brought in contact with it the lower bed No. 1. Inliers often result when beds have been thrown into a dome, and the upper part of it has been shaved off by denudation. In this way a rounded area of the lowest bed which reaches the surface is laid bare, and the bed next above mantles round it in a ring. Such has been the case in Simon's Seat (see fig. 144), where an inlier of the bed No. 2 occurs at the top of the hill. Inliers of this kind were called by the older geologists "Outliers by Protrusion." But faults have also had a share in the production of inliers. Thus in tigs. 155 and 156 there is a small tri- angular area of the bed No. 1, overlaid on the north-east and south-east by the bed next above, and with the same bed brought against it by a fault on the west. Crich Hill in Derbyshire is a good case of a faulted inlier. Figs. 158 and 159 are a map and section of it. The patch of the bed 1 satisfies the defini- tion of an inlier; on the west and south-west it is bounded by faults, which have let down higher beds against it, while to the north-east it passes with a regular dip beneath the bed 2 immediately above it. SECTION IV. FAULTS. Eocks have been subjected to still more violent usage than the folding we have already spoken of. In many cases they have been torn across by rents and the parts which were originally continuous now lie at different levels on opposite sides of the fissure. Faults. 4 8 9 Such displacements are known as Faults, Tkroivs, Troubles, Heaves, Slips ; other local names are also applied to them. Fig. 160 is a section of a group of Coals, Shales, and Sandstones intersected by two faults. Fault. Fault. Fig. 160. If we look at the fault on the right, we see certain measures on both sides of it exactly the same in number, thickness, and composi- tion ; but on the left hand they are bodily in a lower position than on the right hand. This fault is said to throw down, or to have a down- cast, to the left. Similarly the other fault throws up in the same direction, or down to the right. The amount of the throw, or the size Fig. 161. FAULTS UNACCOMPANIED BY DISTURBANCE OR CONTORTION. of the fault, is measured by the vertical distance between the ends of the same bed on opposite sides of the dislocation ; thus the dotted line AB is the throw of the fault to the left in fig. 160. The throw of faults varies from a few inches up to thousands of feet. Sometimes beds show little or no change of dip on approaching a fault, as in the section on fig. 161. This does occasionally happen in the case of faults of considerable size. More frequently however the 490 Geology. beds are steeply tilted or violently contorted in the neighbourhood of a fault, as in the section on fig. 162. The amount of contortion does not necessarily bear any relation to the size of the fault, being some- times very conspicuous where the throw is small. A fault is sometimes a single clean-cut fracture, but it oftener happens that, as we draw near a large fault, the beds are broken by a number of smaller dislocations, as in the section on fig. 163, where we Fig. 162. CONTORTED BEDS IN THE NEIGHBOURHOOD OF A FAULT. cross several such before reaching the main fault. These minor throws are frequently parallel to the main fracture. In other cases faults branch off at large angles from a main throw, and decrease rapidly in size as they recede from it till they die out altogether. Fig. 164 is a ground-plan of a main fault and a group of associated smaller faults, some of which are rudely parallel to the principal dislocation and others branch off from it at various angles, while all show considerable Small Faults. Main Fault. Fig. 163. changes in the amount of throw. Very frequently the branches have a large throw at the junction and gradually grow smaller as they recede from the main fault till they die away altogether. The fissure of a fault is now and then narrow, clean cut, and of a uniform width. It oftener happens however that the walls of a fault are uneven, alternately approaching and receding from one another. The spaces thus formed are filled up with fragments of the adjoining Faults. 491 rocks, mashed and jumbled together, in some cases bound into a solid mass called " fault-stuff " or "fault-rock." Where a fault traverses clayey rocks, its fissure is often lined by a layer of extremely dense, tough, leathery fault-stuff, called in some districts a "leather coat." Xow and then a fault is filled in with crystallized minerals ; and if among them metallic ores occur, it becomes a mineral vein (see p. 546). Slickenside. The walls of a fault are fre- quently grooved and some- times highly polished, as if the rocks on opposite sides had ground against one another. Such mark- ings are called Slickensides. In many cases they are undoubtedly due to the cause just mentioned, and it often looks as if the heat produced by the fric- tion had baked and har- dened the rock, and coated the walls of the fissure with a glazed lining; in other cases a thin glaze of some mineral seems to have been deposited on the polished surface increasing its smoothness* Surfaces marked by slickenside are often found traversing beds in every direction in the neighbourhood of a fault, as if the whole body of the rock had been shattered and the bits rubbed against one another by the motion which produced the dis- placement. Cases also occur where inclined beds have slipped upon one another and marked their faces with slickenside, and we may even see the faces of joints traversed by hori- zontal polished grooves, such as would be produced by horizontal motion and grinding. On approaching a fault some rocks, especially Sandstone, lose their bedding and become shattered and traversed by a number of cracks roughly parallel to the plane of the fault, which are sometimes called * See Journal of Royal Geol. Soc. of Dublin, x. 96. See also Quart. Journ. Geol. Soc. xxxi. Ill, 113, 386. Fig 164. GROUND-PLAN OF A MAIN FAULT WITH BRANCHES AND PARALLEL FAULTS. Main fault shown by a double line, other faults by single lines. Each fault has a small cross-mark placed on the down-cast side, and the amount of the throw written alongside in feet and inches. A cipher is placed where a fault dies out. 49 2 Geology. " Ruttles " by quarrymen. Such changes, which are perhaps akin to cleavage, and the hardening often noticed adjoining a fault, point to pressure and other violent treatment during the production of the dislocation. Hade Of Faults. Faults are sometimes vertical, but by far the larger number are inclined at different angles to the horizon. The inclination of a fault is called its hade or underlie, and is measured by the angle between a vertical plane and the plane of the fault. Except in very disturbed and contorted districts the hade or slope is almost cdivays towards the doivnthrow side : exceptions to this rule are called "reversed faults." Cases of reversal are occasionally seen, but many of them are probably more apparent than real, and caused by some temporary bend in the direction of the fissure. I have seen such a case, where the fault, on entering a well-jointed bed, took the line of a joint with an opposite slope to its own, and so became for a short part of its course reversed, while if looked at as a whole it followed the general law. Genuine reversed faults, however, do occur among highly-contorted strata ; * they are found running along overturned anticlinals such as those shown at the eastern end of the section in fig. 150. Course of Faults. The course of a fault is rarely, if ever, absolutely straight, but in the majority of cases faults show a tendency to run in straight lines ; sometimes the deviations from a straight line are so small as to be scarcely noticeable, and in many cases, where they are quite sensible, there is still a general tendency to a rectilinear trend, the fault swinging first to one side and then to the other of a straight line which represents its average direction. The bendings in such a case are usually gentle curves, but occasionally very abrupt zigzags. *r Some faults on the other hand are decidedly curved. Many cases of apparent curvature however, which are derived from mining plans, are deceptive and have arisen from not distinguishing between a fault and its branches. A fault has been followed underground till a branch is reached and the workings have then been carried along the branch, and on the plan the fault and its branch have been represented as one fault bending sharply at the junction. Careful investigation often shows us in such cases that the original fault runs on without any alteration in its direction, and proves that what seemed a bend is really produced by a distinct fault branching out from the main fracture. Parallelism of Faults. It frequently happens that the faults of a district can be divided into two systems, and that all the members of one system show a general tendency to run parallel to the strike while those of the other system range roughly along the dip. The faults of the first system will be parallel to the longer axes of the larger folds in the rocks, and thus faulting and folding seem to stand * See Professor H. D. Rogers, Transactions Royal Soc. of Edinburgh, xxi. 443. Daubree, Geol. Experimental?, i. 344. t See a very jagged faulted boundary of the Carboniferous Limestone on Sheet 81, S.E. of the one-inch map of the Geological Survey of England and Wales, and its description in the Memoir of the Geological Survey on North Derbyshire and the adjoining parts of Yorkshire, p. 33. Faults. 493 in close connection. Two such systems are said to be conjugate to each other. Joints it will be recollected group themselves into two corresponding systems, and joints we saw were probably the result of strain or torsion. Strain and torsions would be necessary consequences of folding. Hence it is in the highest degree probable that jointing and faulting were caused by the forces which folded and contorted the rocks. It may be that more than one pair of conjugate systems exists ; or there may be, in addition to a pair of conjugate systems, a system of parallel faults without any system conjugate to it. A relationship between faults and antic! inals is also pointed out by the fact that the one sometimes pass into the other. The sharpness of the bend gradu- ally increased till at last the tension became greater than the rock could stand, and fracture accompanied by relative displacement of the severed portions resulted.* Changes in Size and Dying out of Faults. The amount of the throw of a fault very seldom keeps the same value for any long distance, and we will glance at some of the causes which produce changes in the size of faults. If a fault be perpendicular to the strike, its throw will remain the same as long as the beds on opposite sides of it have the same dip and strike. Changes however in the amount or direction of the dip will give rise to corresponding changes in the amount, and in certain cases in the direction of the throw. Fig. 165 shows a model which illustrates Fig. 165. CHANGES IN THE AMOUNT AND DIRECTION OF THE THROW OF A FAULT, PRODUCED BY CHANGE OF DIP. one way in which this is brought about. There is a bed of Coal shown by the thick black band, shifted by a fault A BCD. The measures overlying the Coal are supposed to be removed so that we see its surface. On the north side of the fault the bed dips steadily to the east, and its surface is the plane BCEF. On the south side the bed has been thrown into a series of folds giving it a wavy surface, A LKHGMNO. Between C and G the bed is lower on the south than on the north side, or the downthrow is to the south. The amount of throw however steadily decreases as we go towards 6r, and at that point the bed is at the same level on both sides of the fault, or the * The Bivelin Valley near Sheffield furnishes an instance. See Memoir of the Geological Survey of England on North Derbyshire and the adjoining parts of Yorkshire, p. 67 ; also Geol. Mag. vi. 507. 494 Geology. fault has no throw. Between G and H the bed is higher on the south than on the north side, or the fault throws down to" the north. The amount of the throw increases to the west of G for a while, then begins to decrease, and at H again comes down to nothing. Between // and K the fault resumes its former southerly downthrow, and at L another change in the direction of the throw occurs. The same result will evidently be produced when there are a number of branch faults springing out from a main fault. In the model in fig- 166 the measures overlying the black bed are, as in the last Fig. 166. CHANGES IN THE AMOUNT AND DIRECTION OF THE THROW OF A FAULT, PRODUCED BY BRANCH FAULTS. figure, supposed to be removed. A BCD is the plane of a fault, and BCEF the surface of the bed on the north side of it. At C this fault throws down 100 yards to the south, so that on the latter side the Coal is found in the position DFG. FGHK is the plane of a branch fault throwing down to the east 80 yards ; this fault does not affect the bed on the north side of the main fault, but on the south side the bed is brought by it into the position KHLM; there is now only 20 yards difference in the position of the bed on opposite sides of the main fault, or the throw of the latter is reduced to 20 yards. Still further to the west is another fault, LMNO, throwing up 10 yards to the west, and bringing down the size of the main fault to 10 yards. Lastly, the branch fault, PORE, throws up 40 yards to the west ; by this fault the Coal is raised up to QRS, and is 30 yards higher on the south than on the north side of the main fault, or the throw of the latter is now 30 yards down north. In this way by a succession of steps a fault which throws down to the south 100 yards is changed into a fault with a downthrow to the north of 30 yards.* It is easy to see that by proper adjustments a fault might be made to die out permanently in cases such as those just described. In fig. ] 65, suppose that the bed on the south side of the fault, instead of rising to the west of K at as steep an angle as in the figure, took the same inclination as it has on the north side. The bed would then be at JT, and would continue to be to the west of K, at exactly the same level on both sides of the plane A BCD, or the fault would disappear. The same result would be brought about in fig. 166, if the throw of the fault APQR was 10 yards. * Models such as those in figs. 165 and 166 are very easily made with card- board and are more readily understood than drawings. Faults. 495 Mining operations constantly afford proofs of the dying out of faults, and this must be brought about in some such manner as has been described. Frequently a large fault splits up near its termination into a number of branches, each of which gradually dies away. If the beds have a different strike on opposite sides of a fault, and the line of the latter is parallel to the strike on one side, its throw will necessarily vary in size. For instance, in the ground-plan in fig. 167 we have on the south Fault. Fig. 167. side of the fault a persistent easterly strike and a dip to the north ; on the north of the fault the strike gradually changes from an easterly direction on the one side to a north-westerly trend on the other. The fault runs east and west, or parallel to the strike of the beds on its south side. The consequence is, that as we go to the west we find 1 23 Fig. 168. SECTION ALONG THE LINE AB IN FIG. 167. the bed 7 brought against lower and lower members of the series, or the downthrow of the fault increases. The increase in size will be evident from comparing the two sections AB and CD ; in the first the bed 7 is brought on a level with the bed 3, in the second it has been still further let down so that it is on a level with the lower bed 1. It is 496 Geology. worth notice that if we had confined our investigations to the neigh- bourhood of the line AB, where the beds have the same strike on both Fig. 169. SECTION ALONG THE LINE CD IN FIG. 167. sides of the fault and the latter ranges along the strike, we should not have detected the existence of the beds 3, 2, and 1 ; and unless the fault had been actually seen, we might not have become aware of its existence, and might have supposed the beds on its south side passed under 4, instead of really lying a long way above that bed. Effect of Faults on Outcrop. It is very important that the practical geologist should clearly realize the effect which faults have in shifting the outcrop of a bed. In fig. 170, ABCDE is the surface of the ground, supposed for Fig. 170. SHIFTING or THE OUTCROP OF A BED BY A FAULT. simplicity's sake horizontal ; GELH, the plane of a fault ; DELF, GAKH, the position of the same bed on opposite sides of the fault, the measures overlying the bed being supposed to have been removed. ED, the intersection of the bed with the surface on the upcast side of the fault, shows its outcrop on that side ; but it is clear that after having been thrown down into the position A GHK it will not reach the surface till some way to the left of ED, such as at AG. That is to say, on the doivncast side the outcrop is shifted towards the rise. If we knew the angle of the dip, it is evident that by measuring the horizontal displacement GE we could calculate the vertical throw of the fault, and conversely knowing the throw we could find GE.* The * If t be the throw of the fault, 6 the full dip on the downcast side, the angle between the line of the fault and the strike on the downcast side, then the space through which the outcrop will be shifted, measured on level ground along the line of the fault, equals t cotan cosec 0. If measured perpendicular to the strike the shift is t cotan 6, and may be found Faults. 497 smaller the dip the greater will be the amount of the shift, and the only case in which the outcrop will not be shifted is when the beds are absolutely vertical. If the surface of the ground be undulating, the displacements of the outcrop will become more complicated ; but the above rule will always hold good except when the ground slopes in the same direction as the beds clip and at a larger angle ; in this case the horizontal shifting of the outcrop will be towards the dip. This case is illustrated by fig. 171. S R Fig. 171. SHIFTING OF OUTCROP BY FAULT. PQRS is the surface of the ground represented for simplicity's sake as plane, the other letters have the same meaning as in fig. 170. Both the ground and the bed slope towards the spectator, but the slope of the ground is the steeper. The shift of the outcrop is obviously to the dip. Miners use the term "heave "to describe the horizontal displace- ment of an outcrop by a fault. Faults will heave not only beds but any other divisional planes traversing the rocks ; thus if faults them- selves have been formed at different times and cross one another, the one last produced will heave all of earlier date which are not absolutely vertical. Mineral veins are " heaved " in a similar manner. It is perhaps in the case of a group of mineral veins that the effect of a "heave" becomes most conspicuous. The whole of the rocks of a district appear in such a case to have been bodily shifted in a hori- zontal direction. But the reasoning just gone through shows that there need not have been any horizontal motion whatever, and that the shift may be due to a displacement which was entirely vertical. If we know the amount of the vertical throw and the dip of the heaved bed or vein, we can calculate the amount of horizontal shift that would result ; and if the observed shift agrees with the calculated displace- ment, the throw of the fault must have been altogether in a vertical direction. But it must not be assumed that the displacements pro- by multiplying t by the value of n corresponding to the value of 6 in the table on p. 462. For instance suppose the throw is 50 yards and the dip 12: the value of n for 12 is 47, and the shift perpendicular to the strike = 50 x 47 = 235 yards. Conversely the throw will be found by dividing the shift measured per- pendicular to the strike by n. 2i 49^ Geology. duced by faults are always wholly vertical ; it is conceivable, nay highly probable in those cases where the rocks have been subjected to powerful horizontal compression, that horizontal motion may have taken place, and that beds may have been moved not only up and down, but also to and fro on opposite sides of a fault. A case of horizontal shifting of this nature occurs in the highly- inclined strata on the west side of Freshwater Bay in the Isle of Wight, and is shown in fig. 172. The rock is Chalk and the dark Fig. 172. HORIZONTAL DISPLACEMENT BY A FAULT, FRESHWATER BAY, ISLE or WIGHT. bands are lines of Flint. The central band is a tabular layer and is shifted 12 inches to the north. The band of nodules to the north is shifted 10 inches ; the other band of nodules has suffered a similar shift. The displacement dies away both to north and south. Indirect Evidence for Faults. We have sometimes the good luck to see faults in actual section, as in figs. 161, 162, and 163, which are all sketches from nature. But in many cases the geologist has to infer the presence of faults from circumstances connected with the lie of the beds which cannot be explained any other way. Thus the shifting of the outcrop of a bed is proof positive of a fault ; and by noting where the outcrops of successive beds are broken and heaved, we get a series of points on the fault and can lay down its line. There is another way in which we are often enabled to infer the presence of a fault that is nowhere actually seen, which will be under- stood by a reference to fig. 167. In the district of which that wood- cut is a geological map, the group of beds numbered 1 to 5 run up one after the other to the outcrop of the bed 7, and end off against it. Now this abrupt termination of the outcrops of the beds 1 to 5 may be produced in two ways : there may be a fault, as is actually the case in the instance before us, ranging along the line where the outcrops are stopped off; or there may be between the group 1 to 5 and the Faults. 499 bed 7 what is known as an unconformity, the meaning of which term will be explained in section vi. of the present chapter. But if we are sure there is no unconformity, then such behaviour of the outcrops as is shown in the figure can be explained in no other way than by a fault. We may also infer the presence of faults in cases such as the follow- ing, and if we take due care, our conclusions may be as safely relied upon as if we had actually seen the fault itself. If from a study of sections we establish the fact that a certain bed A, is always found beneath another bed J5, and then find at any spot A dipping so as to abut against or pass over B, and if by no possible contortion or inversion B could be got to pass wider A, then there must be a fault between them. Figs. 173 and 174 will illustrate the line of reasoning pursued in such cases. If we go northwards from Filey Brig along the Yorkshire coast, we find a beautiful series of sections in the cliffs, which show beds coming out one from under the other in the following order. 4. Sandy Limestone and Calcareous Sandstone. Coralline Oolite. 3. Blue Sandy Clay. Oxford Clay. 2. Calcareous Sandstone. Kelloway Rock. 1. Sandstones and Shales. We follow the lowest division, and find it forming a series of head- lands seen in the most distant parts of the view, and after passing these we enter the southern part of Cayton Bay, a sketch of which seen from the north is given in fig. 173. The darkly-coloured promontories about the middle of the coast-line are formed of the Sandstones (1) ; then comes a portion of the cliff more moundy in outline, where a thick mass of stony Clay descends to the sea-level and hides the bedded rocks from view. Between this obscure ground and the spectator rise the bold Lebberston Cliffs, which we recognise at a glance to be com- posed of our old acquaintances Coralline Oolite, Oxford Clay, and Kelloway Rock.* The dip is quite perceptible even from a distance ; and if we carry on the lines of bedding in the headlands beyond the clay-covered interval up to Lebberston Cliff, we see that, so far from the Sandstones of the former passing beneath the Kelloway Rock of the latter, as we found was the case in the normal unbroken section to the south, they would, if they retained the same dip, abut against the Oxford Clay. The first question we ask ourselves is, whether the Sandstones may not bend over rapidly to the north beneath the clay-covered ground, then resume their old dip, and so come into their proper position beneath Lebberston Cliff? But we can detect no symptoms of the abrupt changes of dip required by this supposition, and what is more, when we test the idea by actual measurement, we find that by no bend- ing, however abrupt, could the whole thickness of the group 1 be got in between the points where the clean-cut sections on the north and south terminate. One explanation alone remains, namely that the rocks of Lebberston * In the sketch the first is left nearly white, the Oxford Clay light with streaks of bedding, and a dark bed of Kelloway Rock sticks out at the bottom. 5OO Geology. Cliff have been let down against the Sandstones by a fault, and we w 'c O K H accordingly construct our section as in fig. 174,* and restoring by the * It will be observed that the deep hollow on the section to the south of Lebberston Cliff does not exist in the view. It is filled up by the stony Clay already mentioned, which caps all the distant cliffs, but is omitted in the section to avoid confusion. How the Displacements of the Rocks were produced. 501 dotted lines the parts which have been carried away by denudation, we find that the fault brings the base of the Kelloway Rock just on a level with that of the Coralline Oolite, or that its throw is equal to the combined thickness of the Oxford Clay and Kelloway Rock. Evidence as complete as that just giv.en may be always safely accepted as unquestionable proof of faulting ; but the observer must be on his guard against jumping too hastily to conclu- sions in such cases, and must not call in a fault to help him out of a difficulty till he has thoroughly satisfied himself that the relative position of the beds can be explained in no other way. Where a fault is only one way out of several of explaining observed facts, it may yet be the best way and its pre- sence highly probable ; but the observer must endeavour to obtain additional evidence sufficient to put the question beyond reasonable doubt before adopt- ing a fault as the final solution. SECTION V. HOW THE DISPLACE- MENTS OF THE ROCKS WERE PRODUCED. Such is an outline of the displace- ments which rocks have undergone. We may next inquire how they were pro- duced, and this inquiry naturally falls under two heads.* The first is purely geometrical, and asks what was the kind of motion by which they were brought about 1 The second is mechani- cal, and inquires what were the forces that caused that motion? The first question may be treated of here, the second falls to be considered in part in Chapter XIV. Character of the Movements. It is evident that the movements to which uplifting and tilting were due cannot have gone on everywhere to the same extent. Rocks have been raised higher and more violently disturbed at some spots than at others. The first question is, Was the disturbance sudden and confined to certain lines or centres, so that the rocks were snapped and raised at a * Corresponding to the two subdivisions of the mechanics of motion, Kine- matics and Dynamics, or the science of motion and the science of force. 502 Geology. bound into the positions they now occupy ? Or was the displacement widespread and varying continuously in amount from place to place, so that it reached a maximum along certain lines or around certain centres and died away gradually as it receded from them ? Experience is against the first supposition ; for although it seems likely that small upheavals have been suddenly produced in some instances,* by far the larger number of the cases of oscillation that have been observed extend over large areas and vary in amount and direction continuously. What is more, connected observations of the lie of rocks over a large area furnish evidence of the strongest kind in favour of the second supposition. Isolated measurements show us beds dipping here in one direction at one angle, and there in another direction at another angle. Now suppose that when we have amassed a sufficient number of such observations, we endeavour to determine from them what must be the underground course of the rocks in order that they may come out at each spot where they are seen with the observed dip and direction, and so to arrive at a general view of the geological structure of a country. Whenever we do this, we hnd that we can account for the observed facts only on one supposition, and that is, that the rocks have been folded into a series of troughs and arches or thrown into domes and basins. This is the great general law which governs everywhere the arrangement of the disturbed portions of the earth's crust. Faulting, or violent contortion and inversion, often complicate and obscure this structure and interfere with its symmetry, but never to such an extent as to prevent its being recognised as the great leading feature in the arrangement of the rocks. Disturbances such as these last are therefore of the nature of accidents, and if we eliminate them, and try to form a broad general view of the lie of the beds under a large area, it is the structure just mentioned that invariably comes out.t We may say then that wherever we find beds inclined to the horizon, we are somewhere on the slope of an anticlinal ; and where- ever the beds of a rock group that has undergone disturbance lie flat, we are on the crest of an anticlinal, or at the bottom of a synclinal, or on one of the horizontal portions of the minor bendings that are ever occurring here and there in the sweep of the grand curves. Whatever may have been the forces to which this arrangement has been due, it is quite evident that the movement which produced it cannot have been local, but must have prevailed as universally as the folding itself ; and the generally regular character of the result shuts out completely the idea of a violent, convulsive action, though faulting and contortion point to concentration of energy around the spots where they occur. * The recorded cases have been accompanied and are usually said to have been caused by earthquakes. It is far more likely that earthquake and upheaval, instead of standing in the relation of cause and effect, are both results of a common cause. t It is hardly possible for any one who has not gone through a course of practical geological work in the field to realize fully the grounds on which this assertion is based; but every geologist of experience soon comes to find out the truth of it. How the Displacements of tJie Rocks were produced. 503 . All the observed facts therefore are decidedly in the teeth of the first solution, and strongly in favour of the second. Folding would produce both Elevation and Dip. Folding such as we have described of course necessarily involves tilting and all the different forms of inclination which have been de- scribed as occurring in nature. And it is clear that it would also pro- duce both elevation and depression, the one when portions of the earth's crust are carried aloft on the summits of arches, the other when portions are sunk into troughs. But the reader must not jump to the conclu- sion that all hill-ranges coincide with anticlinals and all valleys with synclinals. We shall see in the next chapter that in most cases the reverse of this is true, and that the present surface of the ground is largely due to denudation acting while the folding was going on or after it was finished which has immensely modified the forms that would have resulted from elevation alone. Direction of the Folding Force. Our next inquiry is, In what direction did the force act which brought about the foldings and displacements that have been described in the preceding pages ? There are two perfectly-distinct methods by which these results might be produced. The very word elevation suggests the notion of a force that acted from below vertically upwards. In order to produce folds all that is necessary is that this force should not act with equal intensity over the area affected by it. Along anticlinal lines it must be at a maximum, and it must gradually decrease in intensity from there down to each synclinal line, along which it must have its least value. The way in which an action of this sort would produce eleva- tion and folding is shown in fig. 175. Fig. 175. FOLDING PRODUCED BY VERTICAL UP-THRUST. Let PQ be the surface of an undisturbed bed, and let a force tending to raise the bed vertically upwards be exerted from beneath in the direction of the arrows ; let A and B be points where the force has maximum value, C an intermediate point where it is at a minimum. Then while A and B are raised to a and 6, C will only reach to a less height c, the points PACBQ will be lifted into some such positions as pacbq, and two anticlinals with a synclinal between them will be formed. But all the results we have been considering might also be produced equally well in the following way. Suppose the bed AB to be subjected to a horizontal thrust acting in the direction of the arrows in fig. 176. The effect would manifestly 504 Geology, be to crumple it up into the shape acdeb, and we should again get a series of anticlinals and synclinals. ct Fig. 176. FOLDING PRODUCED BY HORIZONTAL THRUST. This method of displacement may be easily imitated by laying a strip of cardboard on a table and pressing the ends together by sliding the fingers along the table. M. Daubree devised a machine by which results exactly similar to those met with in nature are more perfectly reproduced. He operated on strips of various substances, confined between two plates which limited their motion in a vertical direction. Horizontal pressure was applied by means of a screw. When the plates were parallel symmetrical folds were produced ; they became more numerous as the pressure was increased. When the plates were inclined to one another contortion and inverted arches were obtained : in one experiment the strip was bent into just such an S-shaped fold as occurs in the section in fig. 152. Similar conditions to that of the inclined plates would obtain in nature if the rocks overlying the beds that were being contorted were thicker at some spots than others, for where the weight above was smallest the rocks would be freest to rise vertically. By taking strips of unequal thickness other forms of crumpling and contortion were produced.* We have to choose then between these two explanations, and to adopt as the most probable the one which accounts for the greatest number of observed facts. Now as far as the formation of symmetrical arches and troughs, like those in fig. 175, goes, one way is as good as the other; but when we come to anticlinals where the rocks are doubled under on the steeper side, to complicated contortions and puckerings, and to the inversions which are their results, vertical upheaval is manifestly quite unable to produce these, while on the other hand they are just the forms that would result from lateral thrust. It is impossible for instance that the arrangement of the beds in the section across the Appalachians in fig. 150 could have been brought about by a force acting vertically upwards. There, not only are some of the arches unsymmetrical and some tilted over, but in the first the steepest sides all face the same way, towards the west, and in the second the tilt has been in every case in the same direction. We also notice that the sharpest bends are at the eastern end, and that the folding grows gradually less sudden, and the curves open out as we go towards the west. These are just the results that would follow if a group of horizontal strata were crumpled up by a powerful thrust, t At the east end of the section then there can hardly be a doubt that horizontal pressure and not vertical upheaval has been the producing cause ; and there is such a gradual passage from the violent disturbances of that end into the * Geol. Experimentale, i. 288. t Silliman's Journ. 1st series, xlix. 284. OF THE UNIVERS How the Displacements of the Rocks were prot more symmetrical folds of the western end, that we must whatever caused the one must also have produced the other. Evidence like this and similar cases might be brought forward without number is strongly in favour of the second explanation. But though there can be little doubt that horizontal compression produced the folds in this and similar cases, it is not always possible to say on which side the thrust was applied. If we suppose it in this instance to have come from the east, we can then explain why the folding is sharpest and why the arches are inverted on that side, on the ground that the force produced its maximum effect there because it was closest to its source. As it was transmitted towards the west it became more and more weakened, and it gave rise to gentler folding. But it is possible to account for the fact equally well on the supposi- tion that the thrust came from the west, if we bear in mind that a ridge of old crystalline rocks lay to the east of the contorted beds. They were jammed up against this buttress, were squeezed closest in its immediate neighbourhood, and the resistance which it opposed to easterly motion caused the folds to be canted over to the west. The reversed faults also which seem to be the rule and not the excep- tion among violently-contorted rocks, could scarcely be produced in any other way but by horizontal thrust. Fig. 177 shows the way in which Fig. 177. FORMATION OF A REVERSED FAULT. an unbroken fold has been observed to pass gradually into a fractured and faulted anticlinal. The arch first becomes canted over ; the tilt gradu- ally increases and the beds become compressed between the axis planes ; then follows fracture and displacement. The hade of these faults is very large, so much so that they often seem nearly horizontal, and the beds on the downcast side have been shoved bodily underneath those on the upcast side. A very good instance occurs in the Somersetshire coalfield, where it is known as the Great Overlap Fault. Other cases occur in the Belgian and Westphalian coalfields. One-half of the anticlinal has slid over or under the other half in a direction not far from horizontal, and it is quite impossible that any force acting vertically upwards could have done this, but horizontal thrust is quite equal to such a task. Daubree indeed has produced Overlap Faults of exactly this nature by applying strong pressure to prisms in the direction of their length.* * Geol. Experimental, i. 321, 322. See also Hebert, Geol. Mag. [2] iv. 441. 56 Geology. Another test that readily suggests itself is this. In fig. 175 the bed AB must be pulled out to bring it into the position acb in fig. 176 it must be compressed. If then we have any means of learning whether folded strata have been stretched or compressed) we shall make some way towards deciding between the rival hypotheses. We may first consider whether the amount of stretching required by the hypothesis of vertical upheaval is such as we can reasonably suppose rocks capable of. In the case of broad open arches and basins perhaps no difficulty would arise on this ground;* but where rocks have been sharply bent into folds which follow one another in rapid succession, they would have to be pulled out to many times their original length to bring them into their present shape. Even supposing rocks as extensible as indiarubber, increase in length must be attended by a corresponding decrease in thickness, and therefore a group of rocks, when sharply folded, ought to appear to be more thinly bedded than in their undis- turbed position. But if we take a group of rocks w r hich lie undisturbed at one spot, and are violently contorted at another, we do not find the beds as a whole thinner at the latter than at the former. In reality however rocks are only slightly extensible, and it is utterly out of the question to suppose that they could .possibly have been dragged out to the extent necessary to bring them into their present form by vertical upthrust. f No such difficulty accompanies the squeezing hypothesis : a band of rock, which when horizontal was, say a mile long, is forced to occupy a smaller space, say three-quarters of a mile, and the only way in which this could be done is by puckering it up into folds. Again the phenomena of cleavage go altogether in favour of lateral thrust. All cleaved rocks are strongly contorted, and the planes of cleavage are parallel to the longer axes of the great folds. Now the structure of cleaved rocks gives proof positive that they have been compressed in a direction perpendicular to the cleavage planes, or, what is the same thing, to the axes of the folds. Lateral pressure therefore has acted on cleaved rocks, and it has acted exactly in the right direction to produce the existing folds. When we see that the rocks have been folded, and when we know that they have been acted on by a force competent to produce that folding, we cannot refuse to believe that the one has been the cause of the other. In the case of cleaved rocks then it is as nearly as can be a certainty that they were bent into their present form by lateral pressure. Another fact which has been repeatedly established in very con- torted districts points to the same conclusion. Where rocks have been bent into an arch like that in the middle figure of fig. 177, some beds are thick at the summits of the arches and the bottoms of the troughs and very thin along the legs of the folds : horizontal squeezing would * The reader will realize how very slight is the curvature of broad basins and arches by consulting section drawn to a true scale across districts where they occur, such as Horizontal Sections, Sheets 77 and 79 of the Geological Survey of England and Wales. t See a very ingenious paper, on the stretching which has taken place in dis- turbed rocks, by Mr. R. L. Jack, Geol. Mag. viii. 388. How the Displacements of the Rocks were produced. 507 evidently tend to produce this result. Heim quotes cases where this has occurred on a small and also on an enormously large scale.* He also points out that when the arch has been very much tilted over so that the axis-planes become nearly horizontal, the compression tends to drive the trough under to the left and the arch over to the right ; the beds in the leg between are rolled out between these two masses of rock as they move in opposite directions, and this has gone on in some cases to such an extent that some of the strata in the centre of the leg have been not merely very much reduced in thickness but actually squeezed out altogether.! The late Mr. W. Hopkins was one of the ablest supporters of the vertical-upheaval theory, and his papers^ on the subject are still well worthy the attention of the student of Dynamical Geology, though many of the geological opinions which he held have been long since given up. He tried to get over the objections stated in the last few pages by pointing out how contortion, and faults as well, might result from vertical upthrust. He supposed the uplifted area to be acted on underneath by a force, such as would be produced by the expansion of a body of highly-heated elastic vapour, and determined by mathe- matical calculation what would be the direction of the rents formed, when the rock was stretched up to the breaking-point, and fissuring took place. He found that in a rectangular area two sets of parallel fissures would be produced, and that the common direction of one set would be perpendicular to that of the other set. The directions of faults in nature do observe this law. Now suppose ABDC, fig. 178, to be a cross section of one of the Fig. 178. arches which has been fractured along the lines EF, GH, KL, MN \ then the pressure on these parts, such as GHKL, which are broadest below, would be greater than in such as EFGH '; the former would therefore be driven upwards, the fractured portions would be forced into some such positions as in fig. 179, and faults would be produced with a hade to the downthrow side, as is the general rule in nature. * See the beautiful instance in " Mechanismus der Gebirgsbildung, " Tab. xv. figs. 7 and 8, described in vol. ii. p. 52. t Ibid. i. 220. J Researches in Physical Geology, Cambridge Phil. Transactions, 1835 ; Eeport on Earthquakes and Elevation, British Association, 1847. 508 Geology. The same result would follow if elevation went on till the cracks gaped ; for then it would be the wedges, such as EFGH, which have their narrowest ends downwards, that would sink. As far as faults go then the explanation will do well enough, and Mr. Hopkins has pointed out that at a future stage of the process contortion as well might be produced in the following manner. Suppose that when the rocks had come into the position shown in fig. 179, the elevating force ceased to G Fig. 179. act and the shattered mass settled down ; a horizontal thrust would then be produced, which would increase indefinitely as the arch flat- tened. The broken portions would be jammed against one another and their beds crumpled up and contorted ; it might also well happen that a wedge like EFGH would be forced upivards by the nip of the two adjoining masses, and in this way reversed faults, such as accompany violent contortion, might result. In this way faults might be produced, and their direction and hade would be the same as in existing faults. Some degree of contortion might also be brought about. But the machinery would hardly be able to effect the amount of widespread and complicated contortion so frequently met with, specially in mountainous districts,* nor to pro- duce cleavage over areas hundreds of square miles in extent. We can scarcely conceive portions of the earth's crust, large enough to produce these results, being tilted bodily over in the manner this explanation requires. In fact, while Mr. Hopkins has clearly realized that contor- tion involves horizontal thrust, the means he proposes for generat- ing that thrust seem inadequate to produce it over areas sufficiently large. Faults then might follow from vertical upthrust ; it is not so easy to see how they would be produced by compression. Pressure would, it seems at first sight, have a tendency to close up any rents that existed rather than to open new ones ; and even supposing fissures were produced and the rocks on opposite sides of them displaced, the motion would be in such a direction that a " reversed " fault would be * Somewhat similar objections apply to an explanation put forward by Mr. Wilson in the Geol. Mag. v. 207; his figures show that his method would not produce crumpling enough. Mr. Wilson's explanation of the cause of faults is substantially the same as that given above. Hozv the Displacements of the Rocks were produced. 509 produced. For let AC Bach, fig. 180, be an arched stratum traversed by a fissure, DC, P the direction of the crumpling force ; then it is Fig. 180. clear that if P is approximately horizontal, its resolved component parallel to CD will tend upwards from D to (7, and the portion ACca will be pushed up, a displacement which would give rise to a reversed fault. It certainly looks as if stretching were necessary for the production of fissures, and as if the law, that a fault always hades towards the downcast side, could only be accounted for on the supposition that the depressed rocks had slid down the incline of the fissure. But the great weight of evidence against vertical upheaval prevents our accepting that as the cause of the stretching. And stretching would in the end result from lateral pressure if the process were carried far enough. As long as the length of the arc A CB did not exceed that of the bed in its unbent state there would be no stretching ; but as the underlying rocks were gradually arched up into the space acb, they might prevent the points a and b approach- ing one another, and still tend to drive the crest of the arch higher up, and their upward motion could go on only by means of the stretching, and at last rending, of the upper layers of the arch. This would give rise to two conjugate systems of fissures, such as we do frequently meet with in nature. Where, as is often the case, we find besides fissures which do not belong to either of these systems, or where we meet with a second pair of conjugate systems, these must have been formed by a subsequent crumpling along different lines. As far as hade goes we may imagine in a vague sort of way that the severed portions of the arch might get displaced relatively to one another in various ways. The displacement might be due to lateral pressure, in which case we have seen the faults would be reversed ; it might be due to portions being forced up by the upward motion of the interior of the arch, in which case there seems no reason why the hade should be towards one side rather than the other ; again, at a considerable depth the pressure would probably heat the rocks till they became plastic or half melted, and portions of the shattered upper part might sink down into the soft bed, in which case the fault would have the normal hade. It is impossible however to delude oneself into the belief that lame and crude explanations like these are satisfactory; probably we have yet very much to learn about faults before we can 5 i o Geology. frame a theory which will account on mechanical principles for their production and the law of their hade. Summary of the Evidence. The evidence then by which we must decide between vertical upthrust and lateral pressure stands as follows. Tilting and symmetrical folds would result equally from either. Vertical upheaval is capable of giving rise to faults, and indirectly to some degree of contortion ; and in the faults produced in this way the observed law of hade would generally prevail, those faults only being " reversed " which accompany great contortion. But no vertical up- heaval could bring about the widespread and excessive crumpling which so constantly presents itself, while this is just the arrangement that would follow from lateral thrust. Further, the amount of stretch- ing required by the hypothesis of vertical upheaval is far greater than can be admitted. Lastly, cleavage furnishes proof that rocks have been subjected to just the very pressure requisite to bend them into the folded forms they have assumed. The only displacements we cannot thoroughly explain by means of compression are faults. While therefore some of the observed facts can be accounted for equally well on either hypothesis, there are many which compression alone could produce ; indeed the only one of which the latter fails to furnish a perfectly satisfactory explanation, is the direction of the hade in a normal fault. The present state of our knowledge therefore decidedly tends to make us lean to the side of lateral thrust as the kind of force which has produced the displacements we are considering; and to believe that since the balance of evidence is so enormously in its favour, increased knowledge will remove the only bar that now exists to its being accepted without hesitation as a full and perfect explana- tion of the cause of these displacements. But the problem is far from solved. Numerous weak points in the explanations given will occur to the thoughtful reader ; and in truth we can so little realize the conditions under which the process of con- tortion went on, that the best explanation we can arrive at must neces- sarily be incomplete in particulars. What gave rise to lateral thrust is a question that falls to be con- sidered in Chapter XIV. Folding went on at great Depths. A very little reflection will convince us that rocks were not bent into their present shapes near the surface, but that when the process went on they were buried beneath a great thickness of strata which has since been carried away by denudation. In the first place we have positive proof that all folded rocks have suffered largely from denudation. The arches are never complete, but truncated by the removal of portions of the upper beds. Almost any of the sections in this chapter show this, and in figs. 184 and 194 the missing parts or " air-saddles " are some of them indicated by dotted lines. Faults tell the same tale; their course would be marked by lines of vertical cliffs, formed of the beds upheaved on the upcast side, if it had not been that these have been swept away by denudation and the surface pared down to a level. How the Displacements of the Rocks were produced. 511 And indeed it is on such a supposition alone that we can under- stand how rocks could have been bent as sharply as they have been without fracture. That they were consolidated in many cases when they were bent is certain. Thus Sir H. de la Beche points out that in Pembrokeshire a thick mass of Limestone the Carboniferous Lime- stone and a great deposit of overlying Shales and Sandstones the Coal Measures share in the same contortion, which therefore could not have taken place till after the deposition of the latter. But during the time and under the circumstances necessary for the accumulation of the upper group, the Limestone, if it ever was soft, must have become perfectly consolidated. Of the contortions shown in fig. 140, Professor Miall remarks, "The angles are sharp, but unbroken. You may easily test this by passing a finger over one of the bends. There is neither crack nor vein." And he disposes of the explanation that the rock was in an unconsoli- dated state when it was bent, by pointing out that some shells and corals preserved in it, which were certainly not originally plastic, are distorted by the folding. But there is really no necessity for supposing that rocks were soft at the time when they were folded. When we speak of rigid bodies we are apt to forget that rigidity is a matter of degree, and that no such thing as perfect rigidity exists in nature. The experiments of M. Tresca* show that even the substances which we look upon as types of rigidity can be bent, twisted, and even made to flow like liquids when sufficient pressure is brought to bear upon them. We shall see by-and- by that rocks have been subjected to pressures enormously greater than the most powerful we can . command, far greater than the pres- sures beneath which, in the hands of M. Tresca, steel was made to give proof of its imperfect rigidity, and under the influence of forces like these even those rocks which we look upon as most unyielding would be as flexible as sheets of cardboard. With the view of throwing light on the origin of contortion, Pro- fessor Miall carried on a series of ingenious experiments. He succeeded by means of pressure applied gradually for some length of time in bending thin plates of Limestone, but the bent slabs always cracked soon after the pressure was removed : this difficulty was partly over- come by embedding the pieces operated on in pitch. He very justly remarks that the frequent destruction by spontaneous fracture of bent plates, when removed from the machine, seems to imply that an inde- finitely-protracted and uniformly-contorting force is needed to produce unbroken curvature; that resistance on all sides diminishes the risk of fracture ; and that the results attained serve to strengthen the opinion that unbroken anticlinals and synclinals are formed only under a considerable weight of super] acent strata, t These experiments certainly seem to show that if solid rock is to be contorted without * Sur 1'ecoulemfnt des Corps Solides, Mem. des Savants Etrangers, xviii., xx. 75 and 137 ; Comptes Rendus, Ixvi. and Ixviii. : Liouville's Journ. [2] xiii. 379 and xvi. 308. t Geol. Mag. vi. p. 505 ; Proceedings of the Geological and Polytechnic Society of the West Riding of Yorkshire, new series, Part I. (1872); Popular Science Review, January 1872. 512 , Geology. fracture, there must be something to hold it together while the bending is going on ; and the necessary force of restraint would be supplied by the weight of a mass of overlying measures. In Sir James Hall's well-known illustration of contortion * a number of layers of cloth were laid on a table and pressed together by boards at either end. In this way they were forced into folds closely resembling the sharp contortions of rocks. But it was necessary to load them above by another board carrying a heavy weight, and this represents the mass of overlying strata which must have been present when rocks were undergoing folding. We saw reason, when considering the phenomena of metarnorphism, to believe that the process went on deep beneath the surface. Now Metamorphic rocks are almost always highly contorted. We have here then another reason for believing that the rocks were deeply buried when contortion was produced. Folding went on slowly. If the conditions under which the rocks were contorted were at all similar to those by which Professor Miall obtained his results, the bending must have gone on very slowly. Some experiments by Professor Thurston, of the Smithsonian Institute, also point to the same conclusion. He found that if iron, which had been forcibly bent close up to the breaking-point, was kept bent by pressure for seventy-two hours, it showed no tendency to return to its original form, but acquired a "permanent set;" and what is still more to our purpose, it then became capable of further bending, t This property is known as "elastico-viscosity." If bodies which possess it are bent slightly and kept bent for some time, they do not tend to return to their original shape. The molecules rearrange them- selves and become reconciled to their new positions and the body keeps the shape into which it has been bent when the deforming force ceases to act. We can realize from these experiments how by a repetition of small bendings rocks, apparently the most inflexible, could little by little be folded into the sharpest imaginable curves. Analogy leads us to the opinion that faults were not produced at one jump, but by a succession of small displacements. We do not know enough about the physical constitution of solid bodies to enable us to explain thoroughly how rocks apparently inflexible can be bent without fracture, but the following approach to a solution of the problem may be made. In solid bodies there is a force acting between each pair of adjacent molecules which resists any attempt to pull the molecules asunder ; it is called the Force of Cohesion. This force becomes- less and less as the molecules are drawn apart and vanishes when they are separated by a certain interval. When we break a solid the molecules along the surface of fracture are separated far enough to destroy the force of cohesion. Suppose that round each molecule as a centre a sphere is described * Transactions Royal Society of Edinburgh, vii. 85 ; Lyell's Elements, 6th ed. p. 50. t Journal of the Franklin Institute, January 1874. I am indebted to Mr. R. Hunt for calling my attention to this paper. How the Displacements of the Rocks were produced. 5 1 3 whose radius equals the distance that two molecules must be drawn apart in order just to destroy the cohesion between them, and call this the Sphere of Action of the Force of Cohesion for the solid in question. The radius of this sphere will always be very small, but it will be different for different bodies. In a brittle solid the radius may be supposed to be very small indeed ; so that if the molecules be separated by distances infinitesimally small, the body breaks : in a plastic body the radius is somewhat larger, and the molecules may be drawn further asunder without producing fracture. Now consider the case of a cube of rock deep in the earth. It is loaded by a column of rock far in excess of its crushing weight, and if unsupported at the sides it would crush. But the pressure of the rock on its four vertical faces holds the molecules in their places ; no two molecules can separate far enough to get outside the sphere of action of the cohesive force which holds them together. The weight of the load is greater than cohesion, and can therefore move the molecules about as freely as if the body were plastic or fluid, but the side pressures keep the molecules so close together that cohesion, though it is over- come, is not destroyed and no fracture takes place. Motion among the molecules becomes possible, but continuity is not destroyed, that is the rock becomes plastic. It will therefore yield to any pressure suffi- ciently great that is brought to bear upon it. Another view is that in a fluid, as soon as two molecules are forced apart, other molecules flow in to fill up the void. In a plastic body the same action takes place to a less extent, in a brittle body it does not take place at all : the stress therefore that breaks a brittle body only deforms a plastic body. Now in the case of our cube of rock the side pressures will tend to produce this influx of molecules and will give it the property which is the cause of plasticity. It may there- fore be deformed without fracture under a stress which would rupture it if it were unsupported at the sides. Contortions more frequent in Old than in Recent Rocks. When touching on the consolidation of rocks, it was noticed that as a general rule the older rocks were the more completely consoli- dated ; and it was pointed out that this was the case simply because they were older, and for that reason had been oftener and for longer periods subjected to the action of the forces which produce solidifica- tion. A similar statement holds good for contortion ; it prevails most largely, as a general rule, among the older rocks, and exactly the same explanation applies as in the case of consolidation. In the early days of geology this fact was held to prove that con- torting forces acted more energetically during far-distant periods of the earth's history than at present. But it is clear that the facts do not warrant this inference, and that they can be explained just as w r ell in the manner just stated. In the same way, when we see an old man more broken than the generality of young men, we do not infer that the wear and tear of life has necessarily been greater than usual in his case, but only that he has been exposed to it longer. 2K Geology. It is not intended to assert that there never was a time when the forces tending to produce contortion were more vigorous than now. Indeed, if there be any truth in the generally-received view of the earth's early history, this must have been formerly the case. But it is not the greater solidity and excess of contortion in the older rocks that lead us to such a conclusion, but reasons that will be explained further on. SECTION VI. UNCONFORMITY AND OVERLAP. What constitutes Unconformity. A question of paramount importance in geological investigations may be conveniently treated of here. It sometimes happens that in a group of stratified rocks the beds come on, one over the other, each with the same dip as the bed next below it ; any contortions or faults affect all the beds alike, and the same general circumstances of lie and positibn pervade the group from top to bottom. Such an assemblage of beds forms what is known as a Conformable Group. In other cases, in working our way across the rocks of a country, we find perfect conformity to prevail for a certain distance, and are then suddenly brought up by a decided break in the order of succes- sion, and of such breaks we readily distinguish two kinds. In the first case there is a sudden change in the dip and strike, or in one of them. An instance x of this kind is shown in fig. 181.* On Fig. 181. SECTION SHOWING UNCONFORMITIES ACCOMPANIED BY CHANGE or DIP. . Silurian Schists. 6. Pnddingstone, containing pebbles of a. \ rarhoTli fp rnil< , c. Shales and Sandstones, with thin beds of Anthracite. ) ^ ari e. Sandstone. I THIWWH. /. Magnesian Limestone. j du d. Igneous dyke. the right we have steeply-inclined beds of Schist dipping towards the right ; on the edges of these there rests a group of Shales and Sand- stones less steeply inclined and sloping in the opposite direction ; these again are capped by beds of Sandstone and Limestone lying perfectly flat. The different members of each of these three groups are perfectly conformable to one another, but in passing from each group to the one next above it we encounter an abrupt change of dip. In the other kind of break all the beds have the same dip, but they can be separated into two groups, the upper of which rests on a worn * Taken from General de la Marmora's Voyage en Sardaigne. Unconformity and Overlap. 515 and uneven surface of the lower, or abuts suddenly, without a fault, against a slope or cliff formed of the latter. A case of this kind is shown in the diagram in fig. 182, where the upper finely-bedded rocks lie in a trough, which has been worn out of the lower dotted group. A D 3 4 Fig. 182. SECTION SHOWING UNCONFORMITY UNACCOMPANIED BY CHANGE OF DlP. Meaning of Unconformity. -The occurrence of a break of either of these kinds is called an Unconformity, and the groups of strata separated by unconformities are said to be unconformable to one another. Such are the observed facts, and our next business is to ask what they mean, and what are the events by which they were brought about. Turning to fig. 181, we know that the Schists a at one time lay flat at the bottom of the sea in which they were deposited. They have been tilted from their horizontal position, and as they rose denudation pared off the edges of the strata and produced the surface on which the bed b rests. And all this must have been done before the deposition of the next group of rocks began. The tilting and denudation took time to effect ; in many cases a very long time would be required for the removal of the amount of rock which we can prove must have been carried away. The unconformity then we are now looking at is a proof that an interval occurred between the deposition of the two rock groups which it separates, and that during that interval no deposition of rock went on at the spot where the unconformity occurs ; or that if any rocks ivere formed there during that interval, they have been entirely carried away by denudation. We shall see by-and-by that the different rock groups of the earth's crust are in reality so many volumes in which is written an account of some of the events that went on during their formation ; and pursuing the metaphor, we may say that where an unconformity occurs, there are certain of these volumes missing, and that there is consequently a blank space in the chronicle. But just as an historian, when his investigations are checked by coming across an imperfect copy of a work, is sometimes enabled to make good the defect by going to 5 1 6 Geology. another library and recovering there the missing pages or volumes ; so the geologist, when he finds at one spot an unconformity and a corre- sponding break in the chain of events he is endeavouring to trace out, may sometimes pick up some of the lost links in other quarters, where the deposition of strata has gone on with less interruption. This is the case in the Sardinian instance. The bottom rocks are known, by tests which will be described further on, to have been formed at the same time as certain of the slaty rocks of North Wales, and the overlying group is of the same age as the beds from which we in England draw our supplies of Coal. Now with us between these two groups of rocks there is found a great mass of strata known collectively as the Old Red Sandstone, the formation of which went on during the interval which is represented only by an unconformity in the Sardinian series. The section of the latter shows us that the deposition of two groups was separated by an interval, it tells us thus much and no more ; from a study of English geology we learn what was going on elsewhere during that interval, and infer that it was of considerable duration because it allowed time enough for the accumulation of a vast thickness of strata. Exactly the same remarks apply to the unconformity between the middle and upper groups of the section before us ; and here again the rocks wanting in Sardinia are to be found in England and other parts of Europe. In the second kind of unconformity the lower group has not been tilted before the deposition of the beds above. Any displacement from a horizontal position that has taken place affects both groups alike, and must therefore have been produced after the deposition of the upper ; but the lower beds have been denuded before the deposition of the upper beds began, and as time would be necessary for this operation, we have here, quite as much as in the first kind, a proof that an interval, unrepresented at the spot where the unconformity occurs, intervened between the formation of the two groups which it separ- ates. That the denudation described must have taken place will be evident by a glance at fig. 182. The beds of the lower group could not possibly have been deposited so as to end abruptly on the slopes of the hollow which now exists in them. Each must originally have stretched across to the point where we find the corresponding bed on the other side, and the present interruption in their continuity must be due to the removal of portions of them. Unconformities of this class vary very much in importance. Some- times the erosion is such as might be brought about by a very trifling change in physical conditions, and so small in amount that no great time would be required to effect it. Such cases may be better described as "contemporaneous erosion and filling up,"* because they do not indicate the important break associated with the idea of uncon- formity. In other cases the denudation has been extensive, and the interval * An expression used by Professor Jukes. A case of this sort is shown in fig. 87. Unconformity and Overlap. 5 1 7 required for it of long duration, and these may be fairly spoken of as unconformities. An unconformity then of either kind shows us that at the spot where it occurs the process of rock formation did not go on con- tinuously ; that at a certain point of time a stop was put to deposi- tion by the upheaval and conversion into dry land of the sea-bottom ; that by this means the rocks which had been just laid down were brought within the range of denudation and portions of them worn away ; that the surface thus formed was afterwards lowered beneath water, and a new set of rocks deposited on the truncated edges of the lower group. On the other hand steady, uniform deposition would give rise to a conformable group of strata; but the converse proposition, that con- formity indicates the absence of any interval between the deposition of successive members of the series, is not necessarily true ; for we can readily imagine that after the deposition of any one bed the supply of sediment might cease, and a long time might elapse before it was renewed and the bed next above laid down, and thus there would be an interval between the formation of this bed and the one next above it; but if the lower bed remained undisturbed during this interval, the two would be perfectly conformable to one another. In fact the mere existence of a plane of division between two beds is proof of an interval between their formation, and this interval may have been a long one unless there is independent evidence to the contrary. In a word, an unconformity will bear but one interpretation that the process of rock formation was suspended for a time, and that during that time denudation took its place. We cannot be quite so sure of the meaning of conformity, because there are two ways in which it may have been produced. It may have, and in many cases has, arisen from a long, steady continuance of the same conditions ; but in itself it affords no certain proof that such was the case, because the formation of two consecutive beds of a conform- able group of strata may have been separated by an interval without any indication of the fact having come down to us.* To put the matter as shortly as possible, unconformity implies an interval, con- formity does not exclude it. Deposition on Sinking Sea-bottoms. Again conformity is not necessarily evidence that the sea-bed was absolutely at rest during the formation of the beds through which it prevails. We have many cases where strata, all of which must have been laid down in shallow ivater, are piled one over the other in perfect conformity to a thickness of thousands of feet. The only way we can account for this is by supposing that during the whole of the deposition of such a group the sea-bottom was slowly sinking, and that the space through which it ivent down in any given time ivas just equal to the thickness of sediment accumulated during the same time. By an adjustment of this sort the water would always be kept shallow, for as fast as subsidence deepened * The existence and meaning of unconformity were recognised first by the master mind of Hutton. See Theory of the Earth, i. 432, 458 ; Playfair's Works, i. 216, iv. 78. 5 1 8 Geology. it, deposition would fill it up again ; and if the movement affected the whole area over which deposition was going on, no unconformity would be produced. General Conclusions. We arrive then at the following canons. Conformity is produced when during the deposition of a group of strata there has been an absence of upheaval, depression, or denudation ; or when, if either of these operations has gone on, it has affected the whole area over which deposition took place. Uncon- formity exists when upheaval and denudation have removed a portion of one set of beds and another set of beds have been afterwards deposited on the surface so formed. Or, more shortly, a continuance of the same physical conditions gives rise to conformity ; unconformity has been produced by change in these conditions. A simple illustration will perhaps bring home more clearly to the reader's mind the facts of unconformity and their interpretation. Suppose that during a long, peaceful period the various events in a nation's history were noted down as they occurred, and the volumes piled one above another on the floor of a library. The heap so formed may fairly represent strata conformably deposited during a long con- tinuance of the same conditions. Suppose that a time of war and tumult followed, during which some of the volumes got disarranged, a part of the archives was destroyed, and throughout which the dis- turbed state of affairs prevented the carrying on of the work of the chronicler. This would cause a blank period in the history exactly corresponding to the gap indicated by a geological unconformity, and accompanied, like it, by disturbance and partial destruction of the record of what had gone before. On the return of more peaceful times the annalist might resume his labours, and if the volumes he produced were laid upon the disordered remains of the earlier records, they would correspond very closely to the upper group of strata which an unconformity shows resting on the edges of the lower beds. Illustration of Unconformity. In figs. 183 and 184 an attempt has been made to show the results of a strong unconformity. The first is a perspective view of a model, on the upper surface of which the outcrops of various strata are shown by different patterns, while along the sides we see, as we should see in a cliff, a section showing the course of the beds underground. The second figure is a geological section along the line marked on the model. We see at a glance that there are two rock groups, between the lie of which the most marked discordance exists. The lower, distinguished by a lighter tint, has been bent into a number of troughs and arches, which have been truncated by denudation, and a floor, marked by a stroke and dot line on the section, has been formed, on which the upper group rests in a nearly horizontal position. The latter has also suffered by denuda- tion, and only two detached outliers remain of the sheet of it which once spread over the whole area : the connection which originally existed between the beds of these outliers is shown in the section by dotted lines. Further proofs of denudation previous to the deposition of the higher beds are furnished by a fault and dyke, which traverse the lower group but do not penetrate into the upper. Another fault, which Unconformity and Overlap. 519 affects both groups, and is therefore of later date than the formation of the upper, is seen on the left hand. The clean-cut section shown by the cliff puts beyond question the existence of the unconformity ; but we will now go on to show how 520 Geology. the unconformity might be detected by simply mapping the country geologically, even if the cliff section did not exist, and without paying regard to the difference in the dip of the beds. If we trace across the country the run of the beds of either group, we find them always coming on, one over the other, in the same order. Among the lower set 1 is overlaid by 2, and this is always followed by 3, above which the other members succeed in the order of the numbers which they bear in the diagrams; and this is seen to be the case in what- ever direction we traverse the district, and whatever disturbance the beds have undergone. Similarly with the upper group ; it matters not where we descend the flanks of the hills which they compose, we always find them cropping out in the same order, d at the top, then c, then 6, and a at the bottom. But we meet with a very different state of things when we follow the line which parts the two groups. The bed a rests first on 2, it then stretches over 1, after leaving the latter it comes again to lie upon 2, and then creeps gradually on to higher and higher beds till it is in con- tact with the highest member 6 of the lower group; still further to the right it comes to lie upon 5 and 4 in suc- cession. Now this gradual passage of the upper group over the edges of the different mem- bers of the lower group can be caused only in two ways either by a fault bringing one against the other, or by an uncon- formity between the two. On the left hand there is a fault bringing Dyke. Fault. Fault. Unconformity and Overlap. $21 about this result ; on the right the wavy and indented nature of the boundary would be all but conclusive against it being faulted, even if the cliff section did not show the absence of any fault. Having assured ourselves then that there is no fault to cause the upper group to abut at different spots against different members of the lower, we may accept this behaviour as conclusive proof in itself of an unconformity between the tivo groups. Incidental Proofs of Unconformity. In some cases where such direct proofs of unconformity as we have just described are not forthcoming, we may detect its presence by evidence of a circumstantial kind. It frequently happens that the bottom beds of the upper of two unconformable groups are Conglomerates, the pebbles of which have been derived from the lower group. This shows denudation, and there- fore upheaval of the lower group before the upper began to be formed. This has been the case in the section in fig. 181, with the bottom bed b of the middle group. We may occasionally detect faults or igneous dykes which penetrate the lower group, but do not run on into the upper. Fig. 181 furnishes us with an instance of this ; the dyke d terminates at the base of the bed e, and the abrupt way in which it is cut off shows that it must originally have extended higher up than now, and that it has been truncated by the denudation that produced the floor on which e rests.* Similar cases are seen in the diagram fig. 184. When we find the lower of two groups of strata intensely metamor- phosed and the upper unchanged, there is a fair presumption that the operation took place before the deposition of the latter, and that there- fore an interval elapsed between the formation of the two. Deceptive Appearance of Unconformity owing to Un- derground Dissolution of Rock. In cases like these now under consideration, where the lower group is calcareous and the upper allows of the passage of water, the observer must be on his guard against too hastily inferring that an uneven junction necessarily means denudation of the former before the deposition of the latter. Under such circum- stances it may be that the two may have been originally laid down in perfect conformity, and that subsequently carbonated water percolated down to the Limestone, dissolved it away, and gave rise to inequalities and pot-holes on its surface, into which the upper beds have settled down. In such a case the irregularity in the junction of the two groups, having been produced after the deposition of the upper, is no proof of an interval having existed between the formation of the two.t Deceptive Conformity. In the examples we have given the evidence for an unconformity is so clear that there is no room for a mistake ; but unconformities are not always so easily detected, and in * Evidence of this kind is perhaps not always absolute proof of unconformity. Mr. Gilbert has described (Geol. of the Henry Mountains, p. 28) cases in which a dyke after penetrating vertically a certain number of beds is stopped off' abruptly by the under surface of the bed next above. Such cases however are probably rare. t See Quart. Journ. Geol. Soc. of London, xxii. 402. 522 Geology. some cases an apparent conformity exists locally between two rock systems which are really violently unconformable to one another. As an instance we may take the Magnesian Limestone and the Coal Measures of Yorkshire, the relative lie of which is shown in the sketch- inap and section in fig. 185. Section along the line AB. <3 Sandstone. I Outcrops of I Coal-seams. Coal Measures. Fig. 185. SKETCH-MAP OF THE YORKSHIRE AND DERBYSHIRE COALFIELD. A geologist who confined his observations to the neighbourhood of the point A would scarcely be able to detect there any signs of an unconformity between these two groups of rocks. The dip of both is in a general way to the east, and is so small that it would be quite impossible to say from such measurements as could be made in quarries or limited exposures of rock, whether there was any difference in the amount of inclination of the two formations. When we come how- ever to map the country in detail, we find that the outcrops of the Coals, Sandstones, and other members of the Coal Measures over a great part of the field trend pretty steadily north and south, but that Unconformity and Overlap. 523 at either end they bend round to the east. In fact the portion of the Goal Measures exposed at the surface forms a half-basin. Now over the central part of this basin the beds strike north and south or in the same direction as the Limestone, and as long as this arrangement prevails, the two formations exhibit no unconformity ; but at either end of the basin, where the easterly strike sets in, we find the Lime- stone, as we go either to the north or the south, resting in succession on lower and lower beds of the Coal Measures, and become aware how discordant the two formations are. We then realize that the Coal Measures have been thrown into a trough and largely denuded before the Limestone began to be formed. To give an idea of the amount of denudation, the Limestone at A rests on beds some 3000 feet higher up in the series than at B so that at the latter point at least this thickness has been swept away. Unconformities like these might escape the notice of a casual and hasty observer, but they would be certain to be revealed by the map- ping of a large tract of country. A case of deceptive conformity is shown in fig. 182. The section at the point A gives no indication of unconformity between the two groups shown in that diagram, but the section taken as a whole shows that a marked unconformity exists. The warning conveyed by such instances is, not to rely too much on apparent conformity, but to bear in mind that it may be only local, and to ascertain by widespread observations whether this is so or not, before concluding that a set of rocks form a conformable series. A single observation will often establish beyond doubt the existence of unconformity ; extensive research will be required before we can safely say beds are conformable to one another. Overlap. When the surface of a group of rocks has been worn by denudation into hollows, and these have been filled by the deposi- tion of a second set of strata, it necessarily happens that each bed of the latter will extend over a larger area than the bed next below, and will cover it up and hide it from view ; so that if we made a section across a country where this had occurred, we should find each bed of the upper group reaching across the bed immediately beneath to abut against the sloping sides of the hollow. This is called an Overlap, and each bed is said to overlap the bed below it. In diagram, fig. 182, we see very clearly, on the left-hand side the bed 1 overlaps 2 ; this again overlaps 3, and 3 overlaps 4, and so on. The existence of the beds below 1 would not have been known, if it had not been that on the right-hand side denudation has removed a portion of the rocks that once covered them, and has laid them open to view. Similarly, in section 4, fig. 187, 3 overlaps 2, and is itself over- lapped by 4. Occasionally we find some among the different members of a forma- tion occurring only in patches here and there, because the conditions necessary for their formation existed only at certain spots ; at the same time these different members overlap one another against the sloping surface of a group of older rocks. Great complication is thus intro- duced, which it requires the utmost care to unravel. A gcod instance 5 24 Geology. occurs on the north-eastern edge of the Lake district. If we turn to a geological map of England, we see that the old rocks of the Hill country, which are distinguished as Silurian, are flanked by a belt of a newer formation, the Mountain Limestone ; but every here and there between the two there come in detached patches of an intermediate group, called the Old Red Sandstone, which consist of Conglomerates and Sandstones. The fact that the latter does not form a continuous band between the Silurian rocks and the Limestone, but occurs only in isolated areas, is due partly to the circumstance that it was deposited only at certain spots and is therefore not present at all in some places, and partly to its being overlapped by the Limestone and so concealed from view at some places where it does exist. During the deposition of these rocks the order of events was as follows. The boss of hill country was upheaved and carved out by denudation into something like its present shape very long ago ; it was afterwards slowly lowered beneath water, and, as it went down, banks of shingle were piled up along the successive shore-lines formed by its gradual submergence. It is easy to see that these would reach their largest dimensions off the mouths of rivers, where the materials for their formation were supplied in greatest abundance. It is also not unlikely that the hills were then occupied by glaciers, and where one of these came down to the water, the load of rubbish on its back would yield plentifully matter suitable for the formation of shingly deposits. And thus it came about that the Conglomerate never formed an unbroken fringe all round the hills, but was deposited only in patches at certain spots. As depression went on, it was accompanied by a change in physical conditions ; the ice disappeared and the volume of the rivers decreased, so that they brought down fine Sand instead of their former coarse detritus ; thus there was next formed a group of Sandstones, less coarsely grained than the Shingle beds which preceded them, and these covered up the latter and extended some way higher up the Silurian slope. Lastly continued depression gave rise to a sea in which Limestone was formed ; this was mainly of organic origin, but it contains in the neighbourhood of the Lake hills beds of mechanically-formed sediment derived from the adjoining land. This last formation covered up the two preceding groups, and abutted at a higher level than they against the Silurian rocks. A section which crossed country where all these groups are present would run as in section 4, fig. 187. But the reader may ask how we become aware of the existence of the Sandstone and Conglomerate, which in the section just given are hidden from view by the Limestone. Their presence is revealed to us by the occasional removal of the overlying beds by denudation, in the way shown by the sketch-map in fig. 186. The district is traversed by river valleys : of these AB and EF run over ground beneath which all four rock groups are present, and cut down deep enough to show them all ; section 1, fig. 187, shows what may be observed in either of these valleys. The valley CD lays bare the Sandstone, but does not cut down to the underlying Conglomerate, the position of which is marked in section 2, fig. 187. The valley GH crosses a spot where the Conglomerate is absent, but shows Sandstone and Limestone ; Unconformity and Overlap. 525 section 3, fig. 187, runs along this valley. Lastly in the country to the right no valleys reach beneath the base of the Limestone, but it is * CO Copper compounds colour the flame emerald-green ; when moistened with Hydrochloric Acid, the colour becomes blue, or green edged with blue. Most of the natural compounds of Copper are soluble in Nitric Acid. A bit of bright Iron placed in the solution is coated by Copper : when Ammonia is added, the solution turns blue. Flame colouration suffices for the determination of Copper in most of its natural compounds : the green may be confounded with that of Boric Acid, but the production of a blue border when the assay is moistened with Hydrochloric Acid distinguishes the flame colouration of Copper from that of all other substances. As all the ores of Copper we have to notice give the above reactions, they will not be repeated for each ; only the points specially character- istic of each ore will be noted. All also give Metallic Copper with Carbonate of Soda on Charcoal. The student therefore if he cannot recognise any ore by its physical characters alone, will ascertain tirst that it is a Copper Ore by one of the tests just given, and will then have no difficulty in deciding which ore it is by the characters given below. Description of Metallic Ores. 5 29 (1 a) SULPHIDES OF COPPER. Copper-pyrites, Chalcopyrite, Kupferkies. A Sulphide of Iron and Copper, the relative proportions of the two metals being very variable. Percentage of Copper ranges up to about 34 per cent. G. 4'1 4'3. Mixtures of Copper- and Iron-pyrites are common, and may be styled Cupriferous Iron Pyrites ; in these the percentage of Copper may be as low as '2 per cent. Dimetric. ~ ='986, so that the unit octahedron is nearly a regular octahedron. Much more common massive than crystallized. H. 3'5 4. Brass to golden yellow : as a rule the larger the per- centage of Copper, the more golden is the tint of the yellow. Frequently iridescent on the surface (Peacock Copper Ore). Streak dark. Easily fusible B.B. to a magnetic globule. May be distinguished from Iron Pyrites by being easily scratched by the knife ; from Gold by not being malleable and by being soluble in Nitric Acid ; from Millerite (p. 539) by its strong Copper reactions. Blister Copper is botryoidal or mammillary Copper Pyrites, often dark olive-green externally. Bornite, Erubescite, Purple Copper Ore, Horse-flesh Ore, Buntlaip- fererz, Cuivre Panache. A compound or mixture of Sulphides of Iron and Copper of very variable composition. Often an alteration product. Percentage of Copper ranges from 44 to 70 (Rammelsberg). G. 4-4 5-5. The crystallized varieties are monometric, but the mineral usually occurs massive. H. 3. Colour of fresh fracture copper red ; tarnishes speedily, and is externally purple, brownish, and often iridescent ; streak dark. B.B. easily fusible to a magnetic globule. The purplish surface colour and the copper-red colour of a fresh fracture will generally distinguish Bornite from Copper Pyrites. Chalcocite, Redruthite, Copper-glance, Vitreous Copper Ore, Kup- ferglanz. Cuprous Sulphide, Cu.,S. 78 '8 per cent, of Copper. G. 5-55-8. Trirnetric, often forms hexagonal, star-shaped, or cross-shaped twins ; in micaceous spangles ; most commonly massive. H. 2-53. Sectile. Blackish grey, often tarnished blue or green outside ; streak shining. B.B. on coal boils and fuses easily. Its sectility is the distinguishing feature of this ore ; in pure massive varieties it is most marked ; impurities often cause a specimen to cut with a gritty feel, but it can always be cut. A blackish sectile ore which gives the Copper reactions must be Copper-glance. Tetrahcdrite, Grey Copper Ore, Fahlerz (Fawn-coloured Ore). 4(Cu,,:Fe:Zn:Ag 2 :Hg)S.(Sb:As:Bi) 2 S 3 . The metals of the formula are not necessarily all present. Antimony, Copper, and Iron may be said to be practically contained in all Fahlores, the other metals may or may not be found. G. 4 '5 5 -11. 2L 5 3 Geology. Monometric, strong tendency to assume regular tetrahedral forms, but often in very complicated and twinned crystals. Frequently massive. H. 3 4'5. Greyish or blackish, often tarnished outside. B.B. easily fusible. The greyish colour, and tetrahedral forms if crystallized, often lead us to suspect a mineral to be Tetrahedrite. Sulphur, Copper, and Iron are easily detected by any of the usual methods ; Antimony will give the sublimate of Sulphide of Antimony in a closed tube ; Arsenic, if present in any quantity, is recognised by its smell when the mineral is fused B.B. on coal. Mercury may be recognised by fusing the powdered ore with three times its weight of Carbonate of Soda in a closed tube. The metal sublimes and condenses in a whitish or metallic coat, and if this be gently swept together by a feather or chip of wood, it collects into small globules of mercury. For more complete determination of the constituents wet analysis must be used. Tetrahedrites rich in Silver are worked as Silver Ores. Tennantite is a black Fahlore, crystallizing in holohedral isometric forms. (1 b) OXIDES OF COPPER. Cuprite, Red Copper Ore, Eotliknpfererz. Cuprous Oxide, Cu 2 O. 88-8 per cent, of Copper. G. 5 -85 6 '15. Monometric, very common in regular octahedrons ; massive. H. 3 - 5 4. Deep red of various shades, often tarnished black out- side ; streak bright to brownish red. Crystallized it cannot be mis- . taken. The massive forms are a little like Haematite, but the red is more of a copper colour than the blood-red of Hematite, and any of the usual tests show it to be a Copper Ore. Earthy and sometimes mixed with Hematite it forms Ziegelerz (Brick Ore}. Melaconite, Schivarzkupfererz, Kupferschwarze. Cupric Oxide, CuO. 79 '85 per cent, of Copper, if pure. Mostly very earthy and pulveru- lent with large admixture of impurities ; generally coating Copper Pyrites or some other Copper Ore, by the alteration of which it has been formed. Black, soiling the fingers. Tenorite is crystallized Cupric Oxide. Trimetric. In scales on the walls of fissures in lava at Vesuvius. G. 6-25. H. 3. B.B. in- fusible. (1 c) CAEBONATES OF COPPER. Malachite. Green Hydrated Cupric Carbonate, CuCO 3 .CuH 2 O 2 . 57-33 per cent, of Copper. G. 3-74-01. Monoclinic ; good basal cleavage. Usually in kidney-shaped nodules or silky fibres ; also massive. Frequently incrusting other Copper Ores by the alteration of which it has been formed. H. 3'5 4. Bright green ; streak pale green. Fusible B.B. It is scarcely possible to mistake Malachite, but in cases of doubt, the student must recollect that it will dissolve with effervescence in dilute Hydrochloric or Nitric Acid, and the solution will give the usual reactions for Copper. Ignorant prospectors have mistaken Description of Metallic Ores. 5 3 1 Chlorite for Malachite, but the above simple test will at once distin- guish between the two. For distinction from Emerald Nickel see that mineral (p. 539). Azurite, Chessylite, Kupferlasur (Copper Lapis lazuli}. Blue Hy- drated Cupric Carbonate, 2CuCO 3 .CuH.jO 2 . 55'16 per cent, of Copper. G. 3 *5 3 '8. Monoclinic ; also massive and earthy. H. 3-54-25. Brittle. Brilliant blue, when not tarnished by exposure ; streak bright blue, but not so dark as the colour. Azurite can seldom be mistaken ; if any doubt exists, see if the mineral dissolves with effervescence in acid, and if the solution reacts for copper. There is however one not common mineral which closely resembles Azurite, viz. Lindrite or Bleilasur (Lead Lapis lazuli), PbS0 4 CuH 2 O.>. The powder of Liuarite is very pale chalky blue, and it does not effervesce with acid. (1 d) SILICATES OF COPPER. Chrysocolla, Kieselkupfcr. A hydrated Silicate of Copper which has arisen from the alteration of other ores, and has therefore a vari- able composition ; often mixed with Oxides of Iron and Manganese, free Silica, and other impurities. G. 2 2 -238. Generally in incrust- ing films, or stalactitic or botryoidal masses. H. 2 4. Conchoidal fracture. Green or blue, black or brown when impure ; streak white when pure. Infusible B.B. Dissolves in Nitric Acid, leaving Silica. The absence of effervescence with acids distinguishes it from Malachite ; its inf usibility is also characteristic. Dioptase, Emerald Copper. CuSiO 3 H 2 O. Hydrated Silicate of Copper. G. 3 - 28 3 -3 5. Rhombohedral, with rhombohedral cleavage. H. 5. Brittle. Emerald green ; streak green. B.B. infusible, but colours flame green. Decomposed by acids. It is harder than Chrysocolla, and has good cleavage when crystal- lized. (1 e) CHLORIDE OF COPPER. Atacamite. CuCl 2 .3CuH._,(X. Hydrated Cupric Oxychloride. G. 3-7613-898. Trimetric," brachydiagonal cleavage. H. 3 3'5. Various shades of bright deep green ; streak apple-green. B.B. fuses, colours flame green with blue margin. Soluble in acids. The blue margin to the green flame, obtained without moistening with Hydrochloric Acid, shows the presence of Chloride of Copper. Chlorine may also be detected by adding Nitrate of Silver to a solu- tion in Nitric Acid, when a curdy precipitate which turns grey or black when exposed to the light, comes down. There are besides Arsenates, Phosphates, and other natural com- pounds of Copper, but they are comparatively rare minerals. 2. ORES OF LEAD. The ores of Lead we have to deal with are all fusible B.B., and all give metallic Lead on charcoal with Carbonate of Soda, some of S3 2 Geology. them alone. Part of the reduced Lead is vaporized, and a coating of Plumbic Oxide is deposited on the coal, which is lemon-yellow when hot, and sulphur-yellow when cold. The bead is easily known to be Lead by its colour, its softness, and by its being malleable and sectile. For confirmation it may be dis- solved by heating in a small quantity of dilute Nitric Acid ; on adding a drop or two of dilute Hydrochloric Acid a white precipitate of Chloride of Lead comes down, when the solution has grown cold, in very slender needle-shaped crystals. This precipitate is insoluble in cold but soluble in boiling water ; it dissolves therefore when the solution is boiled, and comes down again somewhat slowly when it is allowed to cool. If the bead be very small no precipitate may form with Hydrochloric Acid, for Chloride of Lead is slightly soluble in Hydro- chloric Acid. It is then better to dissolve in Nitric Acid and evapor- ate to dryness ; dissolve the residue in a little water and add a drop or two of Sulphuric Acid when a white precipitate will come down ; or add a little Ammonium Sulphide, a black colouration or precipitate indicates Lead. The presence of Lead in a metallic ore may be ascertained by reducing it with or without Carbonate of Soda on charcoal, and noting the incrustation formed, which is yellow both when hot and cold. In the ores which are soluble the formation of a precipitate of Chloride of Lead when a Hydrochloric Acid solution cools, or when Hydrochloric Acid is added to a Nitric Acid solution, is also an easy and useful test. Having learned that a mineral is a Lead Ore, the student will deter- mine which ore it is by the characters given below. Galena, Sleiglanz.PbS. Lead Sulphide. 86 '6 per cent, of Lead. G. 7-257-7. Monometric. Cubes very common. Perfect cubical cleavage. H. 2-5 2-75; somewhat sectile. Colour and streak lead-grey, beautifully lustrous when freshly broken. When crystallized its colour and perfect cubical cleavage are alone sufficient to identify Galena. The massive granular forms may in cases of doubt be powdered, made into a paste with water, and treated B.B. on coal. Sulphurous fumes are given off and a bead of Lead is obtained, more easily with Carbonate of Soda than alone. Galena always contains some silver, in some cases enough to make it valuable as a Silver Ore. Silver is best detected by cupellation ; for the method reference must be made to any of the works on the use of the blowpipe. Anglesite, Bleivitriol, VitriolUeierz. PbSO 4 . Lead Sulphate. G. 6-126-39. Trimetric, crystals often tabular. H. 2 '75 3 ; very brittle. White when pure ; often stained various colours ; often splendid adamantine lustre ; streak uricoloured. Thin splinters fuse in the flame of a candle. Its very high specific gravity might lead a careless observer to confound tarnished and dirty Angle- site with Heavy-spar. But B.B. it is easily reduced to metallic Lead on coal and at the same time gives the Sulphur test. It cannot be Description of Metallic Ores. 533 confounded with the Sulphide Galena, and is therefore a Sulphate of Lead. Leadhillite and some other minerals also contain Lead Sul- phate; some of them contain Lead Carbonate as well, and therefore dissolve partly with effervescence in acids ; from all of them Anglesite is distinguished by its very easy fusibility. Anglesite is often found on the surface of Galena or other ores of Lead by the alteration of which it has been formed. Cerussite, Bleispatk, Weissbleierz, White Lead Ore. PbC0 3 . Lead Carbonate. 77'52 per cent, of Lead. G. 646 648. Trimetric, very common in slender needle-shaped crystals ; larger crystals often thin and broad, cross-shaped or star-shaped twins common. Often massive. H. 3 3*5. Very brittle when crystallized. White when pure, often with splendid adamantine lustre ; frequently tarnished by coating of ferric hydrate or other impurities. Decrepitates violently B.B., but with care, or by powdering and making into a paste with water, metallic Lead is easily obtained from it without Carbonate of Soda. Easily distinguished by dissolving with effervescence in Hydrochloric Acid and giving a precipitate of Lead Chloride. Arseno-phospkates of Lead. General composition, 3{Pb 3 (P:As),,O 8 }. Pb(Cl:F) 2 . Part of the Lead often replaced by Calcium. Hexagonal. Yellows, greens, browns, of very many strengths and tints. Fusible very easily, thin splinters in candle-flame. Varieties rich in Arsenic give strong garlic-smelling fumes when fused on charcoal. Physical characters will generally distinguish these minerals, but the following wet method may be used if necessary. Dissolve in Hydrochloric Acid, allow the solution to cool and stand for some time ; most of the Lead will come down as Chloride. Boil off nearly all the acid, dilute with water, and pass Sulphuretted Hydrogen through the solution till it smells strongly of the gas. Before smelling blow away the mixture of air and gas which hangs about the upper part of the tube or beaker. This will bring down the rest of the Lead, but the greater part of the precipitate consists of yellow Arsenous Sulphide, which is known by the strong garlic smell it gives when heated on charcoal. Filter, boil off the Sulphuretted Hydrogen, and test part of the filtrate for Phosphoric Acid with Ammonium Molybdate as directed on p. 142 under Apatite. Neutralize another part of the nitrate with Ammonia and test for Calcium with Ammonium Oxalate. Chlorine may be detected by dissolving in Nitric Acid and adding Nitrate of Silver ; white curdy Argentic Chloride comes down, which turns black or grey on exposure to sunlight. If Fluorine be present, heating in an open tube with Microcosmic Salt will set free Hydro- fluoric Acid, which turns moist Brazil paper yellow and may corrode the glass. The end of the interior of the tube must be kept moist, for dry Hydrofluoric Acid does not act on glass. The varieties containing little Arsenic are called Pyromorphite or Griinbleierz. Their commonest crystalline form is in hexagonal prisms terminated by basal planes. G. 6 '5 7'1. H. 3'5 4. Streak 534 Geology. yellowish. After fusion on coal Pyromorphite solidifies into a globule covered with crystalline facets. Arsenical varieties are called Mimetite or Braunbleierz. When crystallized they are common in hexagonal prisms terminated by hexagonal pyramids.* Faces very often strongly curved. H. 3 5. Streak white or nearly white. Mimetites rich in Lead have G. 7'0 7 '25, they are sometimes distinguished as Campylite: those rich in Calcium have G. 54 5 '5, they are sometimes distinguished as Heydyphane. Crocoisite, ltothesbleierz.P})Cr0 4 . Lead Chromate. G. 5'9 6'1. Monoclinic ; fair prismatic cleavage. H. 2 5 3. Slightly sectile. Various shades of bright or yellowish red ; streak orange-yellow. Melanchroite, Pkcenicocroite. 2PbCr0 4 .PbO. G. 5 "75. (?) Tri- metric, crystals usually tabular and intersecting in a network. Red turning yellow on exposure ; streak brick- red. H. 3 3 '5. Both minerals give with Borax or Microcosmic Salt a bead that shows the Chromium colours (see p. 545) ; this distinguishes them i'rom other Lead Ores. The streak and the tabular shape and network arrangement of the crystals of Melanchroite serve to distinguish it from Crocoisite, but the two cannot always be safely separated except by quantitative analysis. Minium, Red-lead, Mennige. 2PbO. PbCX = Pb 3 O 4 . Triplumbic Tetroxide. G. 4'6. Pulverulent. H. 2 3. Vivid red mixed with yellow ; streak orange-yellow. Crocoisite is the only Lead Ore with which it could possibly be confounded, but it gives no Chromium reaction. 3. ORES OF ZINC. Most of the ores of Zinc are infusible or nearly infusible. They can most of them however be reduced by treatment on charcoal in R.F. with a mixture of Borax and Carbonate of Soda. No bead is obtained, but the metal is volatilized and oxidized when it comes in contact with the air. The oxide condenses on the coal and forms a coating which is yellow when hot and turns white as it cools. If the incrustation be carefully moistened with Nitrate of Cobalt and heated in O.F. it turns green. Many Zinc Ores also turn green when moistened with Nitrate of Cobalt and heated in O.F. Zinc is precipitated from neutral solutions by Ammonium Sulphide, and, if no other metal be present, we obtain white Zinc Sulphide, which turns green when heated in O.F. with Nitrate of Cobalt. By one or other of these methods the presence of Zinc may be ascertained. The distinctive characters of the principal ores will now be given. Sphalerite, Zinc Blende, Black Jack. ZnS. Zinc Sulphide. 67 * These distinctions in crystalline habit are very general, but they must not be relied upon alone to distinguish the Arsenical from the Non-arsenical varieties. Description of Metallic Ores. 535 per cent, of Zinc. G. 3 '9 4 -2. Monometric, with very perfect dodecahedral cleavage. H. 3-54. Brittle. Very commonly dark brown or black and opaque, but also red, green, yellow, and translucent or almost transparent. Resinous lustre. Scarcely fusible ; dissolves slowly in Hydrochloric Acid, giving off Sulphuretted Hydrogen ; faster in Nitric Acid with separation of Sulphur. Its resinous lustre and cleavage are the two first points to notice ; the dark varieties resemble Cassiterite a little in their lustre, but Cassiterite is far harder, has no cleavage, and is insoluble in acids. An ore which gives Zinc incrustation and Sulphur reaction, is resinous, and has dodecahedral cleavage, must be Zinc Blende. The variety called Przibramite contains a small percentage of Cadmium. Treated carefully with Borax and Carbonate of Soda it gives a reddish-brown coating of Oxide of Cadmium before the Zinc incrustation appears. Smit/tsonite, Zinkspath. ZnCO 3 . Zinc Carbonate.* 52 per cent, of Zinc. G. 4 4 '45. Rhombohedral, perfect rhombohedral cleavage ; commonest in incrusting coats or kidney-shaped, botryoidal, or stalag- mitic forms, and massive. H. 5. White when pure, but generally stained by admixtures; delicate green sometimes; frequently rusty from Iron Oxide. Often very dirty from large admixture of rusty sand or clay. Dissolves with effervescence in Hydrochloric Acid, and after precipi- tating Iron by Ammonia, Zinc Sulphide is thrown down by Ammonium Sulphide. Hydrozindte is a Hydrated Zinc Carbonate. It differs from Smith- sonite in yielding water in the closed tube. Calamine, Electric Calcimine, Galmei, Kieselzinkerz. Zn 2 SiO 4 + Water. Hydrated Zinc Silicate. G. 3'16 3 '9. Trimetric, prismatic cleavage : often in stalactitic, botryoidal, or fibrous forms, and massive. H. 4 '5 5. White when pure ; often coloured by admixtures. Soluble in Hydrochloric Acid without effervescence, this distinguishes it from Smithsonite, the only Zinc Ore which it resembles ; the solution when evaporated deposits gelatinous Silica. Many specimens are mixtures of Smithsonite and Calamine ; these dissolve with effervescence in acid and gelatinize on evaporation. Zincite, Red Zinc, Rothzincerz. ZnO. Zinc Oxide. 80'26 per cent, of Zinc. G. 5 '43 5 '7. Hexagonal. In micaceous plates, laminae not so easily separated as in Mica, and brittle : also in combinations of hexagonal prisms and hexagonal pyramids with truncated summits ; also in grains and massive. H. 44-5. Brittle. Deep red or orange-yellow ; streak orange-yellow. Gives the ordinary Zinc reactions, and its physical characters dis- tinguish it from all the other Zinc Ores. Generally contains Manganese. Occurs with Franklinite. * Some authors call the native Zinc Carbonate Calamine, and the Silicate Smithsonite. 536 Geology. 4. TIN. Tin is best detected by reducing the mineral with Carbonate of Soda, or better with a mixture of Carbonate of Soda and Cyanide of Potassium, on charcoal. The metal goes into the coal. Scrape off the part of the coal beneath the assay and put it with the slag into an agate mortar and pound with water. Lower the mortar gently beneath the surface of a basin of water, and move it slightly backwards and forwards in the water. The finely-pounded charcoal floats away. Crush what remains and wash again, and repeat the operation till all the charcoal is washed away. Spangles of metal remain at the bottom. Boil these in a few drops of strong Nitric Acid, very soon a white powder (Stannic Oxide) settles to the bottom. This shows that the metal is either Tin or Antimony; it cannot be Antimony, for Antimony is brittle and would not flatten in the mortar ; it is therefore Tin. Ccusiterite, Black Tin, Tinstone, Zinnstein. SnO 2 . Stannic Oxide. 78-6 per cent, of Tin. G. 64 7-1. Dimetric. Occasionally the unit prism terminated by the unit octahedron is seen, but the crystals are more commonly very compli- cated and twinned. H. 6 7. Most commonly black or brown, but occasionally of other colours ; dark varieties opaque, light coloured sometimes nearly transparent. Resinous brilliant lustre. Infusible, practically unattacked by acids. Its great hardness, its resinous lustre which is most apparent on a fresh fracture, its infusibility, and its insolubility generally make it easy to recognise Cassiterite. In cases of doubt it may be reduced on charcoal. Wood Tin, Holzzinn. Stannic Oxide in kidney-shaped or botryoidal concretions, often with concentric and radiated structures. Colour brownish. Cassiterite is the only ore of Tin, but the following stanniferous mineral may be noticed. Stannite, Tin Pyrites, Zinnkies, Bell-metal Ore. Sulphide of Tin, Iron, Copper, and Zinc. G. 4 '3 4 '5 2 2. Usually massive. H. 4. Steel-grey to iron-black ; streak blackish. The appearance and colour of this mineral cannot very well be described in words, but must be learned from specimens, they are characteristic enough to enable us to recognise it with fair certainty. To confirm such a determination dissolve in Nitric Acid ; Sulphur separates and floats, white Oxide of Tin settles to the bottom, this may be separated by filtration and reduced B.B. on coal : to the filtrate, which is green, add Ammonia, it turns blue indicating Copper and rusty Ferric Hydrate comes down. Zinc may be detected by complete analysis, but a sulphide which contains Tin, Copper, and Iron, and has the physical characters of Tin Pyrites, may be safely put down to be that mineral. 5. ORES OF SILVER. The ores of Silver are easily reducible B.B. on charcoal either alone Description of Metallic Ores. 537 or with Carbonate of Soda. The bead can generally be identified by its physical characters ; it is whiter than Lead, and harder than both Lead and Tin. For certainty it may be dissolved in a small quantity of Nitric Acid and a little Hydrochloric Acid added, when a white curdy precipitate of Argentic Chloride comes down, which turns grey or black when exposed to sunlight. The precipitate is insoluble in boiling water, which distinguishes it from Lead Chloride, and is soluble in Ammonia, which distinguishes it from Mercurous Chloride. Argentic Chloride is slightly soluble in Hydrochloric Acid, and hence if the bead be very small, no precipitate may be formed when that reagent is added. It is better therefore when only a small quantity of metal is obtained to dissolve it in Nitric Acid, evaporate to dryness, and dissolve the residue in water. A drop or two of Hydrochloric Acid will then bring down the curdy precipitate. By these means a mineral may be ascertained to be a Silver Ore ; the distinctive characters of the commoner ores follow. Argentite, Vitreous Silver, Silver-glance, Silberglanz. Ag 2 S. Silver Sulphide. 87 '1 per cent, of Silver. G. 7 '196 7 "365. Monometric : also in branching and thread-like forms and amorphous. H. 2 2-5. Perfectly sectile, malleable. Black, metallic lustre; streak black and shining. Its marked physical characters, specially its malleability, distinguish it from all other Silver Ores. Accmthite has the same composition and the same physical properties except that it is trimetric. Stephanite, Brittle Silver Ore, Sprodglaserz, Schivarzgultigerz, Melanglanz. 5Ag 2 S.Sb 2 S 3 . Silver and Antimony Sulphide. 68 '5 per cent, of Silver. Often contains Iron and Copper. G. 6*269. Trimetric : also massive. H. 22-5. Colour and streak iron-black. Metallic lustre. Pyrargyrite, Dark-red Silver Ore, Dunklesrothgultigerz, Antimon- silberblende. 3Ag 2 S.Sb 2 S 3 . Silver and Antimony Sulphide. 5 9 '8 per cent, of Silver. Sometimes contains some Arsenic. G. 5 '7 5 '9. Hexagonal, crystals have rhombohedral terminations. Imperfect rhombohedral cleavage. Also massive. H. 2 2-5. Black tending to deep red, brilliant metallic lustre; streak deep bright red. Stephanite and Pyrargyrite agree in many of their reactions. Both are soluble in Nitric Acid with separation of Sulphur and white Anti- morious Oxide. Both give dense white fumes of Antimonous Oxide in the open tube. In the closed tube Stephanite gives only a feeble sublimate of Antimony Sulphide, the sublimate given by Pyrargyrite is larger. The streak however distinguishes them completely. Stephanite is not malleable, and this distinguishes it from Argentite. Pyrargyrite gives little or no reaction for Arsenic, and this distin- guishes it from Proustite. Freieslebenite, Schilfglaserz. 4PbS.3Ag 2 S.3Sb 2 S 3 = Pb 2 Ag 3 Sb 3 S 8 . Lead, Silver, and Antimony Sulphide. Silver 2 3 '8 per cent., Lead 30-5 per cent. G. 6 6 '4. S3 8 Geology. Monoclinic, prismatic cleavage. Prisms striated longitudinally like a reed (schilf). H. 2 2 -5. Greyish colour and streak. B.B. fuses easily on coal, and deposits crust, which is white outside from Antimonous Oxide, and yellow inside from Lead Oxide. The lead incrustation, its crystalline habit, and streak distinguish it from the two preceding species. Proustite, Ruby Silver, Light-red Silver Ore, Lichtesrothgultierz, Arsemilberblende. 3Ag.,S.As 2 S 3 . Silver and Arsenic Sulphide. 65 '5 per cent, of Silver. G. 5 "422 5 '56. Rhombohedral ; also massive. H. 2 2 '5. Bright-red colour and streak. Adamantine lustre. Fuses easily B.B., and on coal gives off fumes of Arsenous Oxide with strong garlic smell ; this distinguishes it from Pyrargyrite. In closed tube a small yellow sublimate of Orpiment (As 2 S 3 ) at a red heat. Cerargyrite, Horn Silver, Silberhornerz, Cklorsilber. AgCl. Ar- gentic Chloride. 75*3 per cent, of Silver. G. 5*3 5 *5. Monometric, but usually massive. H. 1 1 '5. Sectile, feels like cutting wax. Its physical characters are so marked that it can generally be recog- nised by them alone. It will dissolve in Ammonia, and on the addition of .Nitric Acid curdy Argentic Chloride is precipitated, and this is a further test. Mixtures of Cerargyrite with Native Silver or other Silver Ores occur : if they be treated with Nitric Acid, Argentic Chloride remains undissolved. ButtermilcAerz is a mixture of Cerargyrite and Clay. There are many other natural compounds of Silver for which refer- ence must be made to works on Mineralogy. 6. ORES OF COBALT AND NICKEL. These two metals are almost invariably met with in company ; it is scarcely possible to find a Nickel Ore that does not contain Cobalt, or a Cobalt Ore that does not contain Nickel. Oxide of Nickel gives with Borax in O.F. a bead that is violet when hot and reddish brown when cold ; with Microcosmic Salt the bead in O.F. is reddish brown when hot and yellow when cold. Oxide of Cobalt gives a beautiful blue bead in both flames with both Borax and Microcosmic Salt. Most of the mixed ores of Cobalt and Nickel are easily reduced by treating them with borax on charcoal in R.F. The globule of metal contains usually Iron as well as Cobalt and Nickel. It is cleared from the borax, some clean borax fused on charcoal, and the globule is treated with this borax-glass in O.F. The globule is again removed and again treated with clean borax, and the process repeated several times over. At first the borax is bottle-green from Iron Oxide : after a time all the Iron passes into the glass and the borax becomes coloured blue from Cobalt : the Cobalt by repeated treatment becomes oxidized and removed ; when it is nearly all gone the borax shows various and not Description of Metallic Ores. 539 very definite colours owing to admixture of Cobalt and Nickel colours in various proportions. At last the borax assumes the reddish brown of .Nickel ; the globule now contains this metal alone. To test further, a fragment of the globule is treated in Borax and Microcosmic Salt on platinum wire and the characteristic Nickel colours are ob- tained. Or the globule may be dissolved in Nitric Acid and Caustic Potash added ; a precipitate of Nickelous Hydrate comes down, which is white at first but turns gradually a pale apple-green. Niccolite, K up fer nickel* (Copper-nickel), Rothnickelkies. NiAs. Nickel Arsenide. 43 '6 per cent, of Nickel. Part of the Arsenic some- times replaced by Antimony. G. 7*33 7 "67. Hexagonal, but usually massive. H. 5 5 '5. Metallic lustre; pale copper-red colour, tarnishing black or grey; streak pale brownish black. Fuses B.B., and on coal gives off arsenical fumes. Treated as directed with Borax a bead of pure Nickel is obtained. This mineral is so marked in its look that it can scarcely be mistaken ; the simple blowpipe reactions mentioned may be used, if necessary. Millerite, Capillary Pyrites, Haarkies, Nickelkies. NiS. Nickel Sulphide. 35 '6 per cent, of Nickel. G. 4'6 5 -65. Rhombohedral, rhombohedral cleavage ; usually in slender hair- like crystals ; also massive. H. 3 3 - 5. Brass-yellow, often with grey iridescent tarnish; metallic lustre ; streak bright. Like Copper Pyrites, but by treatment with Borax or dissolution in Nitric Acid it is found to be a Nickel Ore. Pentlandite, Eisennickelkies. 2FeS.NiS. Iron and Nickel Sul- phides. G. 4-6. Monometric, octahedral cleavage ; usually massive. H. 3'5 4. Bronze-yellow; streak blackish. Not easily reduced B.B., and when reduced the metals are awkward to separate. Boiled in dilute Hydrochloric Acid Iron Sulphide is dissolved and Nickel Sulphide remains and will give Nickel reactions in a Borax bead. This method of treatment combined with the bronze-yellow colour will identify the mineral. Gersdoffite, Nickel-glance, Arsennickelglanz, Nickelarsenkies. NiS.,. NiAs 2 . Nickel Sulpharsenide. Normal mineral would contain 35*1 per cent, of Nickel, but the composition is very variable. G. 5 '6 5 '9. Monometric, fair cubical cleavage, and massive. H. 5 '5. White to greyish, tarnishing grey or dark grey ; streak greyish black. The usual reactions for Nickel Ores and in addition a yellow sublimate of Arsenic Bisulphide (As 2 S 2 ) in the closed tube. Zaratite, Emerald Nickel, Nickelsmaragd. NiCO 3 .2NiH,,O 2 .4H 2 0. Hydrated Nickel Carbonate. In incrustations and stalactitic masses, adjoining, coating, or pene- trating the ores by whose alteration it has been produced. * " Nickel" is a dirty body, a drab. When this mineral was first discovered, its colour led to the notion that it was a Copper Ore, and when the miners failed to get Copper out of it, they revenged themselves by giving it this contemptuous nickname, which may be Englished "Copper-coloured Slut." 54 Geology. H. 3 3 '25. Emerald-green, but a paler tint than Malachite. The paleness of its green will often distinguish it from Malachite. To confirm it may be tested for Nickel in a Borax bead, or dissolved in Nitric Acid and the solution tested with Caustic Potash. Malachite would give a blue precipitate which turns black on boiling. Morenosite, NickelvitrioL NiSO 4 .7H 2 O. Hydrated Nickel Sul- phate. About 26 per cent, of Nickel. G. 2 '004. In needle-shaped crystals or as an efflorescence or coating on other Nickel Ores by whose alteration it has been formed. H. 2 2 '25. Apple-green to greenish white ; streak faint greenish white. Soluble in water with a metallic astringent taste. Its solubility in water at once distinguishes it from both Malachite and Zaratite. The solution gives a white precipitate with Barium Chloride, and an apple-green precipitate of Nickelous Hydrate with Caustic Potash. Garnierite. 2(Ni:Mg) 5 Si 4 O 13 .3H,O. Hydrated Silicate of Nickel and Magnesium. This mineral, which seems likely to be a valuable Nickel Ore, is found near Noumea, the chief town of New Caledonia. It occurs in veins of Serpentine with Chromite and Steatite. Apple- green in colour, adheres to the tongue, not greasy to the touch. H. 2-53. It is an alteration product, and this will probably account for the wide variations in composition which it seems to show : one analysis gives 38 '61 per cent, of Nickel and scarcely any Magnesia, in another the Nickel amounts to only 3 '45 per cent, and there is 37 '38 per cent, of Magnesium.* Linnceite, Cobalt Pyrites, Kobaltkies, Kobaltnickelkies. 2(Co:Ni : Fe)S + (Co:Ni:Fe)S,. Cobalt, Nickel, and Iron Sulphide. G. 4-85. Monometric ; also massive. H. 5 '5. Pale steel-grey, tarnishing copper red ; metallic lustre ; streak blackish grey. A variety containing 30 to 40 per cent, of Nickel has been named Siegnite; a variety containing Copper also occurs. In all varieties there will be an absence of any marked Arsenic reaction and all will give the Sulphur reaction. If the amount of Nickel is small, a Borax bead will show Cobalt colour. If the colour of the Borax bead is indefinite, there is probably a large amount of Nickel, and the mineral may then be treated with repeated doses of Borax and the successive reaction for Iron, Cobalt, and Nickel obtained. In a mineral which has the physical characters of Linnae- ite these simple methods will usually suffice for its determination. In some cases wet analysis or the more refined blowpipe methods may be necessary. Smaltine, Grey Cobalt Ore, Tin-white Cobalt, Speisskobalt. (Co:Ni: Fe)As 2 . Cobalt, Nickel, and Iron Diarsenide. G. 6 '4 7 '2. Mono- metric, octahedral cleavage ; also massive. H. 5 '5 6. Tin-white to steel-grey when massive, with sometimes greyish or iridescent tarnish ; metallic lustre ; streak greyish black. * Jahresberichte fur Chemie. 1874, p. 1260; 1878, p. 1270. Description of Metallic Ores. 541 Except the specific gravity there is little in the physical characters of the massive forms of this group of minerals to distinguish them from the Cobalt Pyrites group. But both on charcoal and in the open tube they give strong arsenical reactions. In the closed tube a mirror of metallic Arsenic is usually obtained. Where Cobalt is the principal metallic ingredient, the name Smaltine is used : these will give a clear Cobalt colour in a Borax bead. Where there is very little Cobalt the mineral becomes a Nickel Ore and is called Chloantite or Weissnickelkies ; these give a sublimate of metallic Arsenic in the closed tube, and a residue of Copper-nickel (NiAs). An ore having the same composition but crystallizing in trimetric forms is called Rammelsbergite. The varieties rich in Iron are called Saffloritc. Cobaltite, Cobalt-glance, Bright-white Cobalt, Kobaltglanz, Glanz- kobaltkies. Normally Co(S 2 : As 2 ). Cobalt Sulpharsenide, but most variable in composition. In the variety Ferro-cobaltite Cobalt is largely replaced by Iron ; in some varieties Copper enters to a small amount. As a rule it seems to contain little Nickel. G. 6 6*3. Monometric, common shapes pentagonal dodecahedrons or combina- tions of that form and the cube. Cubical cleavage. Also massive. H. 5 '5. Silver-white with tendency to red or steel-grey with violet tinge. Ferro-cobaltite greyish black ; streak greyish black. It is not easy to distinguish the massive forms of this species from the preceding by physical characters alone, but they may be discrimi- nated thus. Cobaltite fused alone on charcoal or in the forceps gives off arsenical vapours, fused with Carbonate of Soda on charcoal a strong Sulphur reaction. The presence of both Sulphur and Arsenic will distinguish typical specimens of the mineral from both Linnaoite and Smaltine. Also Cobaltite gives scarcely any sublimate in the closed tube. Asbolite, Earthy Cobalt, Erdkobalt. A mixture of Manganese Peroxide (MnO 2 ), Cobaltous Oxide (CoO), sometimes Cupric Oxide (CuO), and Water. In a Borax bead it may show Manganese colour in O.F., and w r hen this disappears in R.F. the blue of Cobalt comes out. If the Cobalt is so large in amount as to mask the Manganese colour, that element may be determined by fusion with Carbonate of Soda and Nitre. In very dirty varieties clay and other impurities should be removed by washing before testing. Erythrite, Red Cobalt, Cobalt Ochre, KobaUbliithe (Cobalt -bloom). Co 3 As 2 O 8 . 8H 2 0. Hydrated Cobalt Arsenate. Cobalt often re- placed by Nickel and Iron, and sometimes by Calcium. G. 2 '948. Monoelinic, perfect clinodiagonal cleavage, thin laminae flexible ; usually in globular or kidney-shaped forms, or in earthy or pulveru- lent incrustations coating or adhering to other ores of Cobalt by whose alteration it has been formed. H. 1*5 2 '5. Sectile. Generally of reddish or pink colour, some- times greenish grey ; streak lighter than colour, powder deep lavender- blue. B.B. fuses easily and gives off arsenical fumes. Its very marked appearance, the Cobalt colour which it gives in a 542 Geology. Borax bead, and the arsenical reaction on charcoal completely identify it. 7. ORES OF ANTIMONY. The native compounds of Antimony when heated in the forceps give off dense white fumes of Antimonous Oxide (Sb 2 3 ) which have no smell and colour the flame pale yellowish green. On charcoal the same fumes are given off and a white coating is formed which is vola- tile under R.F. and colours it the same colour. In the open tube there are dense white fumes which condense into a white sublimate on the sides of the tube. Lead gives white fumes in the open tube, but they are not so dense as those of Antimony ; Lead colours the flame pale blue ; and the Lead incrustation is yellow. Stibnite, Antimonite, Grey Antimony, Antimonglanz. Sb-,S 3 . Antimonous Sulphide. 71 '8 per cent, of Antimony. G. 4 '5 16. Trimetric. Unit prism nearly square ; common in long prisms, more or less modified, with the lateral faces deeply striated transversely. Frequently in clusters of long slender rod-like prisms. Very perfect brachypinacoid cleavage in well-crystallized specimens. Also massive. H. 2. Sectile. Colour and streak lead grey, paler than Galena. Fuses easily in candle-flame. On charcoal B.B. and in open tube dense white fumes ; in closed tube volatilizes without decomposing, and forms a sublimate of Antimonous Sulphide which is black when hot and reddish- or orange-brown when cold. A little like Galena ; but its very easy fusibility and the dense white fumes which it gives off B.B. at once distinguish it. There are several Sulphides of Lead and Antimony which bear a superficial resemblance to Stibnite, but they are all harder and most of them less easily fused, and they give a globule of Lead on charcoal B.B., or a precipitate of Lead Chloride when dissolved in Hydro- chloric Acid. They are therefore easily distinguished from Stibnite, but most of them can be told from one another only by quantitative analysis. Antimony also enters into the composition of many other nlinerals, but they are not common. 8. ORES OF ARSENIC. The following are some of the most useful methods of detecting Arsenic in minerals. Many compounds of Arsenic when exposed for a moment to the point of the blue flame on charcoal give off fumes smelling strongly of garlic : many give the same result during fusion on charcoal. The flame is tinged pale greenish blue. In an open tube many arsenical compounds give off dense white fumes of Arsenous Oxide (As 4 6 ) which condense and crystallize in minute octahedrons on the sides of the tube. In the closed tube many Arsenides and some Arsenates give a brilliant metallic sublimate (arsenical mirror) above which is a black lustrous velvety sublimate. Heat moderately at first and gradually increase the temperature if Description of Metallic Ores. 543 necessary. Arsenous Oxide may be easily recognised by placing it in a closed tube arid a few fragments of charcoal above it. The charcoal is first raised to a red heat, and then by inclining the tube the assay is brought within the flame ; the Arsenous Oxide is volatilized and reduced by the charcoal to metallic Arsenic which condenses in a mirror above. Many Arsenates and some of the lower Arsenides will not give the above reactions. For these a tube with a bulb at one end is taken and the bulb half filled with a mixture of equal parts of Cyanide of Potassium and Carbonate of Soda : this is gently warmed and any moisture which condenses in the tube is removed by a roll of blotting- paper. When all the water is driven off, the powdered assay is introduced into the bulb and the contents strongly heated. The mineral is decomposed and a mirror of metallic Arsenic forms in the tube. In using closed tubes it is desirable to hold them in a good-sized pair of metal pincers and to grip them low down. The pincers con- duct away the heat and keep the part of the tube above them cool. Arsenic enters into the composition of many minerals ; the following may be fairly styled Arsenic Ores. Orpiment. As 2 S 3 . Arsenous Sulphide. 61 per cent, of Arsenic. G. 3-48. Trirnetric. Perfect macrodiagonal cleavage. H. T5 2. Colour and streak various shades of yellow. Burns on charcoal B. B. with a pale bluish flame giving off arsenical vapours. In closed tube alone melts, volatilizes without decomposition, and gives a yellow sublimate ; heated with Potassium Cyanide a sublimate of metallic Arsenic. Realgar. As 9 S.,. Arsenic Bisulphide. 70 '1 per cent, of Arsenic. G. 3-43-6. Monocliriic. H. 1*5 2. Red to yellow; streak red. Behaves like Orpiment, except that the sublimate in the closed tube is red. 9. BISMUTH. Bismuth fuses easily and gives an incrustation on charcoal which is orange-yellow when hot and lemon-yellow when cold. The colours do not differ sensibly from those of the incrustation in the case of Lead. There is however this difference, the incrustation of Lead Oxide gives a pale-blue colour to the flame when touched with the R.F., the Bismuth incrustation does not colour the flame. A more certain test is to treat the Bismuth compound on charcoal with a mixture of equal parts of Sulphur and Potassium Iodide, when a beautiful red incrus- tation is obtained at some distance from the assay. If a compound containing Bismuth be dissolved in Nitric Acid and the solution concentrated by evaporating nearly to dryness, the addition of water causes a white precipitate of Bismuth Nitrate. Bismuth occurs most commonly native, alloyed with small quantities of Arsenic, Sulphur, and other substances. It is sectile and brittle 544 Geology. when cold. If pure it volatilizes entirely on charcoal B.B., and may be distinguished from other volatile metals by its incrustation. Bismite, Bismuth Ochre, Wismuthocker. Generally in greenish-yellow or greyish-white crusts, pulverulent, or massive. G. 4'3611. Easily reduced on charcoal in R.F. to metallic Bismuth. Bismuthinite, Wismuthglanz, Bismuth Glance. Bi 2 S 3 . Bismuth Tri- sulphide. 81'25 per cent, of Bismuth. G. 6'4 6'459. Trimetric. Unit prism nearly square ; brachydiagonal cleavage : massive with foliated or fibrous structure. H. 2. Colour and streak lead-grey. Fusible in candle-flame. Soluble in Nitric Acid. 10. ORES OF MERCURY. Cinnabar, Zinnober. HgS. Mercuric Sulphide. 86 "1 per cent, of Mercury. G. 8 -998. Rhombohedral, crystals often tabular or modified hexagonal prisms ; cleaves parallel to faces of an hexagonal prism ; also massive and in incrustations. H. 2 2 '5. Bright red to black; streak vermilion. B.B. wholly volatile if pure. If carefully heated in an open tube sulphurous fumes are given off, and white vapours which condense into a coating which is white where thin and metallic where thicker. If this be gently swept together with a feather or splinter of wood it collects into small globules of mercury. Hepatic Cinnabar, Liver Ore, Quecksilberlebererz, is Cinnabar mixed with clay and carbonaceous matter. Horn Quicksilver, Quecksilberhornerz. Mercurous Chloride, Hg.,CL,. 84-9 per cent, of Mercury. G. 6 '482. H. 12. Sectile. May be distinguished from Cerargyrite by giving a sublimate of metallic Mercury when heated with Carbonate of Soda in a closed tube, and by being insoluble in Ammonia, but blackened by it. Mercury also occurs native in globules scattered through rocks. 11. GOLD. Gold occurs chiefly in the native state, either approximately pure or alloyed with Silver and other metals. It may be known by its being fusible, sectile, and malleable, and insoluble in all acids except Aqua regia. If the solution be diluted and Stannous Chloride added, a purple precipitate called the " Purple of Cassius " comes down. 12. PLATINUM. Platinum occurs native, alloyed with Palladium and other rare metals. It is infusible and malleable and insoluble in all mineral acids but Aqua regia. Ammonium Chloride added to the solution throws down yellow Ammonium Platino-chloride. Description of Metallic Ores, 545 13. CHROMIUM. Chromic Oxide dissolves slowly and colours strongly ; it gives beads of the following colours. With Borax. O.F., in very small quantity, yellow when hot, changing quickly to yellowish green w r hen cold. In rather larger quantity, reddish brown when hot, then yellowish, and green when cold. R.F., in very small quantity, green both when hot and cold. In rather larger quantity, the bead is reddish for a short time, and then turns emerald-green. With Microcosmic Salt. O.F., red when hot, turning first dirty green and then clear green as it cools. R.F., the same colours, but darker. Besides the minerals in which Chromium plays an acid part, some of which have been already described, the following may be noticed. Chrome Ochre, Cliromocker. Clayey matter coloured green by Chromic Oxide. Gives Chromium colours in Borax and Microcosmic Salt beads. 14. MOLYBDENUM. Molybdenum imparts a pale yellowish-green colour to the blowpipe flame; it is like the green of Barium, but when the two are seen side by side, the green of Molybdenum is seen to be the yellower. Molybdenite, Molybdanglanz. MoS 2 . Molybdic Sulphide. 59 per cent, of Molybdenum. G. 4*44 4 '8. In short tabular hexagonal prisms with very perfect basal cleavage ; more commonly in micaceous scales, which are bounded by cleavage planes, and are flexible but not elastic. H. 1 1-5. Sectile. Lead-grey; streak the same but with a greenish tinge. Infusible, but in the forceps tinges the flame pale yellowish green, specially if moistened with Hydrochloric Acid. On charcoal in powder, held well in the outer part of O.F., gives sulphurous smell, and coats the coal with Molybdic Oxide, which is yellow when hot and white when cold. The white coating, if touched for a moment with R.F., becomes azure blue : longer heating in R.F. gives a brown colour in the centre edged with azure blue ; the brown is not very distinct. Decomposed by Nitric Acid, leaving white Molybdic Acid. It may be confounded with Graphite, but the means of distinguishing between the two have been given under that mineral (p. 160). Molybdite, Molybdanocker, Molybdic Ochre. Mo0 3 . Molybdic Trioxide. 657 per cent, of Molybdenum. G. 4'49 4 '50. In fine hair-like crystals, and as an earthy powder or incrustation. H. 1 2. Yellow to yellowish white. B.B. on charcoal fuses and gives the same incrustations as Molybdenite. 15. TUNGSTEN OR WOLFRAM. Tungstic Oxide gives with Microcosmic Salt a bead which in O.F. is yellow when hot if strongly saturated. In R.F. the bead is dirty green when hot, and a clear blue on cooling. The blue is the 2M Geology. characteristic colour ; it may be often produced by exposing the bead to the tip of the inner blue part of the O.F. ; indeed if the yellow colour is to be obtained, the bead must be kept well in the yellow outer part of the O.F. A Borax bead is not coloured blue by Tungstic Oxide, and this distinguishes Tungsten from Cobalt. Iron interferes with the reaction. The method of detecting Tungsten in that case has been given under Wolfram (p. 152). Tungsten plays an acid part in Wolfram and other less common minerals ; the following is the only mineral that can be called an ore of Tungsten. Tungstite, Wolframocker. W0 3 . Tungstic Oxide. Earthy and pulverulent associated with minerals containing Tungsten, by whose alteration it has been formed. 16. TITANIUM. Titanic Oxide gives with Microcosmic Salt in O.F. a bead which is yellow when hot and colourless when cold, and in R.F. a bead which is yellow when hot, reddens as it cools, and finally assumes a beautiful violet colour. The assay should be very finely powdered, for Titanic Oxide dissolves very slowly, and as it colours feebly a large quantity is required to obtain the colour. A very little Iron completely masks this reaction ; the method of obtaining the Titanium colours in the presence of Iron has been given on p. 148. Some minerals in which Titanium plays an acid part have been described. The following may be looked upon as ores of the metal ; they are all three composed of Titanic Oxide (TiO 2 ) with some im- purities. Rutile. Dimetric, -= '6442. In modified prisms, and twins which consist of two or more prisms inclined to one another at large angles : frequently in long needle-shaped prisms. Also in masses without external crystalline shape, but showing cleavage. Prismatic and pinacoid cleavages. G. 4'18 4'25. H. 6 6'5. Reddish brown shot with red, brilliant lustre ; but also of various hues ; streak pale brown. Infusible. The peculiar reddish-brown colour shot with red is very suggestive of Rutile ; a mineral showing it should have its hardness determined, and trial should be made whether it is infusible. If it then gives the Titanium colour in a Borax bead, after being treated with Tin if the colour does not show without, it is Rutile. Anatase, Octahedrite. Dimetric, | = 1'7777 1. In unit octahedrons more or less modified, or in tabular crystals capped by the faces of flat octahedrons. Basal and octahedral cleavages. G. 3*82 3 '95. H. 5'5 6. Colours various, often brownish ; streak uncoloured. Crystallized it is easily distinguished from Rutile. When not in Description of Metallic Ores. distinct crystals, its streak is the most marked point also its hardness and specific gravity are both lower. Brooldte. Trimetric, no distinct cleavage. H. 5 '5 6. G. 4*12 4-23. Various colours ; brilliant metallic lustre ; streak uncoloured. 17. URANIUM. Uranic Oxide gives beads of the following colour. With Borax, in O.F., in small quantity yellow when hot and colourless when cold ; in larger quantity brownish red when hot, yellow when cold; in RF. bottle-green. With Microcosmic Salt, in O.F. yellow when hot, changing very quickly to yellowish green as it cools ; in K.F. dirty green when hot, clear green when cold. With Borax the colours are the same as in the case of Iron, and with Microcosmic Salt nearly the same as for Chromium, but the two reac- tions combined distinguish Uranium from either of these metals. Uranin ite, Pitch Blende, Uranpecherz. U 0^ 2 U 3 = U 3 8 . Uranoso- uranic Oxide, but almost always contains besides some of the metals Iron, Manganese, Calcium, Magnesium, Lead, Copper, Zinc, Bismuth, Arsenic. G. 6 '4 8. Monometric, but usually massive. H. 5 '5. Colour and streak very variable, mostly dark, sometimes greenish. Infusible or only slightly rounded on thin edges. The admixtures generally prevent the Uranium colours coming out distinctly in beads, but the following wet test can be applied. Dissolve in Nitric Acid, add Sodium Carbonate in fairly large excess, and boil. Carbonate of Uranium remains in solution, the Carbonates of the other metals are precipitated. Filter, to filtrate add Caustic Potash ; Uranic Oxide (UO 3 ) is precipitated and may now be tested in beads. Copper Uranite, Torbenite, Chalcolite^ Kupferuranite. Cu 3 P 2 8 . 2{3(U0 2 )P 2 O 8 }.24H 2 O = CuP 2 U 2 O 12 .8H 2 0, Double Phosphate of Copper and Uranous Oxide. G. 3 '4 3 '6. Dimetric, very perfect basal cleavage, along which it splits into micaceous plates that are brittle and not flexible. Generally in square tables bounded above and below by basal planes, with the edges replaced by faces of various octahedrons and the angles truncated by piuacoids. H. 2 2 '5. Sectile. Various shades of green; streak somewhat paler than colour. Fuses easily and colours the flame green. Soluble in Nitric Acid. Its brilliant green colour and the very marked shape of its crystals make it impossible to mistake this mineral. Autunite, Lime Uranite, Kalkuranglimmer. The same composition as Copper Uranite, but with Calcium in the place of Copper. Resembles it in every respect except colour and streak, which are some shade of yellow. 548 Geology. SECTION II. METALLIC DEPOSITS. Nature and Forms of Mineral Deposits. Metallic ores form but a very small part of the crust of the earth ; some few of them, such as Magnetite, are found finely disseminated throughout the whole of large masses of rock, but even in these cases they constitute only a minute percentage of the bulk of the rock ; as a rule they are confined to certain limited portions which are more or less distinctly marked off from the rock in which they occur. These portions may consist wholly of metallic ores of one or more kinds, usually they contain besides various species of ore a variety of other minerals. The ores and the associated minerals are in a large number of cases crystallized or crystalline. These collections of ores and other minerals may be called by the general name of Mineral Deposits. Mineral deposits put on a variety of forms ; we may for convenience group them under the following heads. The classification however is very largely one of convenience, there are no well-defined lines between the subdivisions, and we frequently meet with a mineral deposit which possesses characters belonging to two or more of the groups and are at a loss to know which to put it into. FORMS OF MINERAL DEPOSITS. Lodes or Veins, Flats, and Contact Deposits. Stockworks. Masses. Impregnations. Beds. Placers or Alluvial Deposits. 1. LODES. A Lode (German, Gang ; French, Filon) is a rent or fissure usually vertical or not far from vertical, but sometimes inclined at a moderate angle to the horizon. The fissure is filled with material called Vein-stuff, Matrix, Gang- gestein, Gangue, which consists very largely in most cases of crystallized minerals such as Quartz, Calcite, Fluor-spar, Heavy-spar. Fragments of the adjoining rock, clay, and other substances constitute a part or the whole of the vein-stuff in some cases. In this vein-stuff the metallic ores are distributed most irregularly in platy layers, strings, bunches, nodules, and other forms. Some parts of the vein contain no ore, and these barren portions are often very extensive ; in other parts there is practically no vein-stuff and the whole lode is filled with ore, these rich portions are unfortunately of comparatively rare occurrence. Different kinds of ore are frequently met with in different parts of the same lode. In a mine near Ronquillo in Spain I found the upper part of the lode was mainly filled with Iron Pyrites in which a little Metallic Deposits. 549 Galena was irregularly disseminated : as we descended lower, the Iron Pyrites grew less in quantity and the Galena increased till at last the latter became the only ore. Some Cornish lodes contain Copper in the upper and Tin in the lower parts. Comby Lodes. It frequently happens that the vein-stuff is made up of sheets or layers lying parallel to the walls of the vein ; each sheet consists of a single mineral, and when as is often the case the crystals of this mineral are prismatic in shape with pyramidal termi- nations, the longer axes of the prisms are perpendicular to the walls of the vein and their pointed ends are turned inwards. The section of each sheet bears a rude resemblance to a comb, and the lode is said to be " comby." In some comby lodes the sheets of the different minerals are arranged in corresponding pairs on opposite sides of a central plane, and if we have a sheet of any mineral on one side of this plane there is a sheet of the same mineral in a corresponding position on the other side. This is called a Symmetrical Comb. Thus in fig. 188 o u Fig. 188. SECTION ACROSS A COMBY LODE. the portions AabB, BbcC are both symmetrical combs. In the first we have a plate of Calcite on the left wall and a corresponding plate also of Calcite on the right : upon each of the plates of Calcite there lies a plate of Galena, and the comb is completed by a pair of plates of Quartz. The comb BbcC again is itself symmetrical, but the lode, though it consists of two symmetrical combs, is as a whole unsym- metrical. When the filling has not been complete, the empty spaces are called Druses ; Vughs and Tid-holes are Cornish names for such cavities. A somewhat similar structure, though produced in a different way, is sometimes found in lodes filled in with fine mechanically-formed debris such as clay. The material is apparently bedded, the beds 550 Geology. lying parallel to the walls of the lode. There is little doubt that this structure is really cleavage produced by the compression of the soft matter between the walls of the lode. Brecciated Lodes. When lodes contain a large number of angular fragments of rock this name is applied to them. The frag- ments have sometimes come from the rock which forms the wall of the lode, sometimes they are portions of combs or some previous infilling of the lode. In some cases the blocks are of considerable size, so large that they practically divide the lode into two or more portions ; they are then called Horses or Riders. In some cases the fragments are coated by a crust of crystallized mineral, either arranged in concentric layers or in prismatic crystals standing with their longer axes perpendicular to the surface of the blocks. Rounded pebbles and boulders have also been found in lodes.* They consist occasionally of rock different from that of the walls. Terms connected with Lodes. The inclination of the plane of a lode to a vertical plane is called its Hade or Underlie; it may be expressed in degrees, or by saying that the lode underlies one in so much, so many feet to a fathom, so many inches to a yard, and so on. Thus in fig. 189 the angle CAB is 19, AB is nearly three times BC, and if AB be a fathom BC will be nearly 2 feet. The hade is therefore 19, or the lode underlies one in three or 2 feet to a fathom. The two walls of a lode are thus distinguished. Suppose we are standing upright in a lode, one wall would overhang our head like a high-pitched roof, this is called the Hanging (German, Hangende t) Wall; the wall beneath our feet is called the Foot (German, Liegende t) Wall. The rock traversed by a lode is called by English miners the Country, by Germans Nebengestein. It often happens that a band of " country " on either side of a lode is more or less altered, or the lode is separated from the " country " by a layer of clay or some other substance. Such a belt separating a lode from the country is called a Salband. The line along which a lode comes to the surface is called its Out- crop, Outgoing, or Back (German, Ausgehende ; French, Affleurement, Crete). Gossan. The contents of a lode along the back are usually altered by the action of the air, and differ very much in appearance from what is seen in the lode at depths which the action of the air has not reached. Ferruginous minerals are so commonly present that these altered por- tions are in the majority of cases more or less rusty in colour. They are called Gossan, Iron Hat (Eiserner Hut) by the Germans, Chapeau de Fer by the French, Pacos, Colorados, Negrillos by South American miners. Besides the Oxide and Hydrate of Iron which give Gossans their rusty colour we find in them ores which have been formed by * Came, Journ. Geol. Soc. Cornwall, iii. 238 ; Salmon, Quart. Journ. Geol. Soc. xvii. (1861) 517 ; Smyth, Memoirs of Geological Survey of Great Britain, vol. ii. part ii. 604. t These terms are also applied by German geologists to bedded rocks in the sense of "overlying" and "underlying." Metallic Deposits. 55* the oxidation or other alteration of the ores in the lode. Thus the Gossan of a lode containing Lead Sulphide (Galena) may yield Lead Sulphate (Anglesite) and Lead Carbonate (Cerussite); in the upper parts of a lode of Copper Pyrites (Sulphide of Copper and Iron), Melaconite (Black Oxide of Copper), Malachite and Azurite (Hydrated Carbonates of Copper) often occur. A Gossan will thus often give some clue as to what will be found in the lode beneath ; and even where we are not able to derive much information of this sort from Gossans, they are of great practical im- Fig. 189. SECTION ACROSS A LODE. portance because they indicate the presence of lodes, and show the direction in which they are running. Gossans are not invariably present along the backs of lodes.* Heaving of Lodes. t Lodes are very generally faults. If after a lode has been formed and filled, the rocks are a second time sub- jected to disturbance and a second set of faults is formed, these will shift or " heave " the outcrops of the first set of lodes in exactly the * Smyth, Memoirs of Geological Survey of Great Britain, vol. ii. part ii. 656. t Rather an unnecessary amount of fuss has been made on this subject by writers on mining. The principle, which has been gone into on p. 496, is simple enough. The results however become very complicated when there are several sets of fissures formed at several periods, and each heaving all previously formed. For a good case see Quart. Journ. Geol. Soc. xxii. (1866) 535. 552 Geology. same way as we have explained that faults shift the outcrops of beds (p. 496), for a lode is nothing but a steeply-inclined bed. If therefore we find a lode "heaving" another we may be sure that the lode that heaves has been formed after the lode which is heaved. Before how- ever we can come to this conclusion we must be sure that the lode is really "heaved," for a deceptive appearance of "heaving" may be produced in several ways. For instance after one fissure (AB) had been formed and filled a second fissure might be opened say at right angles to it. When the second fissure struck AB say on the point C it would be very likely to run along AB say to the point D, for it might well be that the rending would take place more easily along an old fracture than across the solid rock beyond. After having followed AB up to D, a line of weakness in the rock might present itself, and the second fissure would take to this and continue across the country. If this second fissure became afterwards filled, it would have the appearance of having been heaved from C to D by AB, and we might fall into the mistake of supposing it older than AB. The apparent heaving of a lode then is somewhat dangerous evidence to trust to, but there can be no risk of mistake if we note whether the rocks which bound the lode are heaved as well. If they are, it is a genuine case of displacement, and the lode which heaves them must be of later date than any lode which is heaved with them. Direction of Lodes. Lodes are nothing but faults the fissures of which happen to have been filled to a greater or less extent with metallic minerals. Their directions accordingly follow the same laws as those of faults. As in the case of other faults we frequently find two conjugate systems (see p. 493); the lodes of each system are roughly parallel to one another, and one set range parallel to the general strike the other set parallel to the direction of the average dip. There may be besides other systems of parallel lodes with or without systems conjugate to them. In fact all that has been said about the directions of faults applies word for word to the direction of lodes. For instance in the north of England the beds have been bent into a great anticlinal whose axis runs north and south. One system of lodes runs in the same direction or parallel to the strike, a system conjugate to this trends east and west or parallel to the dip. There is besides a second pair of conjugate systems, called Quarter-point Veins, ranging about S. 75 E. and S. 35 W. Dimensions and Extent of Lodes. Lodes have been traced in a horizontal direction for several miles in some cases, in others we are told that they do not extend many yards. It must be remembered however that a lode is not usually followed after it ceases to yield ore in sufficient quantity to make its working profitable, and it by no means follows that the fissure terminates because it ceases to bear ore in paying quantity. Some lodes have been found to be productive to depths of more than 2000 feet, others .again either cease to bear ore or decrease in breadth so much as to be unworkable to a profit at very shallow depths. Here again however there is nothing to forbid the lode carrying ore or opening out again lower down, and actual exploration has in some cases shown that this does take place. Metallic Deposits. 553 The breadth of lodes varies from a mere fraction of an inch to several hundred feet in some instances. The breadth of the same lode also varies very considerably both vertically and horizontally ; and in some cases it is easy to see how this variation has been caused. Let the reader take a sheet of card- board and divide it by a cut like that shown in the right-hand figure on fig. 190. Fig. 190. Now let the sheet be held in a vertical position, and let the left hand be lowered so that the two pieces come into the position shown in the left-hand figure. Along the portions of the cut which are vertical we have broad open spaces ; in the inclined portions the two pieces are in contact and the fissure is closed. This is exactly what happens frequently in lodes. The parts where the fissure approaches the vertical are broad and carry ore because there is room for it ; the flatter parts are narrow, " nipped " as the miners say, and therefore unproductive. A similar variation in the breadth of the fissure would be produced in a horizontal direction, if the rock on one side of a curved rent were shifted horizontally. In soluble rocks, such as Limestone, variation in the breadth of a lode may have been caused by the unequal dissolution of its walls. Lodes frequently split up, and two branches after pursuing a separate course for a while run together again. In fact it seldom happens that a lode is a single fissure ; a number of rudely parallel fissures run side by side, or strings and branches start out from a main fissure. Lodes often terminate by splitting into a number of strings. In all these respects indeed lodes behave like other faults. Relation between Contents of Lodes and adjoining 554 Geology. Rocks. The amount and character of the ore yielded by lodes have in many cases been found to depend on the rock which they traverse. In the north of England Lead lodes cut through a group of rocks consisting of alternations of Limestone, Sandstone, and Shale. As a rule they carry most ore when both cheeks of the lode are Limestone, a smaller quantity when only one cheek is Limestone, and are narrow and barren in Shale. The reason for this may be partly mechanical ; in hard Limestone the walls of the fissure would stand and the fissure itself would be widened by the solvent action of percolating water ; there would therefore be space for the deposition of ore and the other contents of the lode. In soft Shale the fissure would soon close, owing partly to the crumbling in of the soft walls and partly to the readiness with which they yield to that horizontal pressure by which rocks have been brought into their present positions. The fact also that in many cases the fissures tend to be vertical in Limestone and inclined in Shale favours the accumulation of ore in Limestone. But this cannot be the whole reason ; if it were, hard Sandstones ought to be as good carriers of ore as Limestone, but this is not the general rule. The law how- ever is not without exceptions. At Grassirigton more ore was obtained from the lode in a bed of Sandstone called the Bearing Grit than from those parts which traversed Limestone. In Derbyshire sheets of contemporaneous Doleritic lava called Toad- stone are interstratified with the Limestone ; the lodes which contain Galena in the Limestone cease to carry ore when they enter a Toadstone bed,* but become productive again in the Limestone underneath. The Limestone itself can be divided into two parts, an upper thinly bedded, somewhat earthy, and containing Chert, and a lower thickly bedded, purer, and without Chert; a number of statistics which I collected seem to show that the lodes are richer in the upper than the lower division. In Cornwall the lodes are as a rule productive only in the neighbour- hood of the junction of Granite and Clay-slate or where they adjoin or cross dykes of Quartz-felsite known locally as El vans. The proximity of intrusive Igneous rocks seems in many cases to exert a favourable influence on the contents of lodes. The effect of the enclosing rock is also well illustrated by the Fahl- bands of Norway. The country in the neighbourhood of Kongsberg, where these occur, is composed of Gneiss with subordinate beds of Hornblende rock, Mica-schist, Chlorite-schist, Talc-schist, and Quartzite. Certain belts, which run parallel to the strike, are richly impregnated with Iron Pyrites, Magnetic Pyrites, Copper Pyrites, Galena, Zinc Blende, and other metallic ores, finely disseminated through the rock. These are called Fahlbands. The Fahlbands as a rule are not worth working, but there are lodes cutting across them and carrying Silver Ores which are rich when they cross a Fahlband and poor over the other parts of their course. It very frequently happens, that where two lodes intersect, unusually rich deposits of ore are found. * Hence the name, which is a corruption of the German Todtstein (Deadstone), the rock in which the ore dies out. Metallic Deposits. 555 Flats. Where lodes traverse bedded rocks, sheets of ore some- times spread out from the lode between the planes of bedding, which are thus designated (see fig. 189). Much more valuable and extensive deposits of a similar kind occur in soluble rocks such as Limestone. In this the dissolution of the rock has formed hollows and even caverns adjoining the lode, and these have been filled at the same time and in the same manner as the lode itself. The Pipe Veins of Derbyshire seem to have had a similar origin. They are rudely cylindrical pipes descending nearly vertically through the Limestone and filled with Lead Ore and other minerals. They are probably vertical portions of old underground watercourses. Contact Deposits. Tabular-shaped deposits of metallic minerals are often found along the surface of junction of two kinds of rock, and go by this name. In some cases the plane of junction is a fault, and they are then true lodes. In any case they have much analogy to lodes. 2. STOCKWORKS. The rock adjoining the walls of a lode is usually traversed by a net- work of countless cracks running in all directions which branch out from the lode and are filled with minerals similar to those found in the lode itself. It sometimes happens, that, without there being any one main trunk fissure, a mass of rock is interlaced by an exactly similar network of ore-bearing veins. Such a deposit is called in German a Stockwerk or floor, because the ore is extracted by a succession of floors excavated in the mass. The French call it Amas entrelace. The Mulberry Mine near Bodmin in Cornwall is a good instance. The rock is Killas (Clay- slate) dipping at 45 N. 22 W. and is traversed by a number of small fissures tilled with Cassiterite, Quartz, and a little Mispickel and Wolf- ram. The fissures trend N. 7 W. and dip to the west at 80 to 90 ; they range from mere cracks not thicker than a knife-blade up to 4 or 5 inches and rarely to a foot in breadth. They often run for considerable distances independently, and then bend and unite in the direction both of the dip and strike, and they are further connected by thin layers of ore running between the planes of bedding of the Slate. The belt of Slate through which these fissures extend is 30 yards wide and 300 yards long, and has been followed to a depth of 120 feet.* Deposits simulating Lodes but without well-defined Walls. These deposits differ from many stockworks in ranging along a fissure, and from lodes in the fact that the bulk of the ore is not contained in the fissure but is disseminated through its walls. Fig. 191, which is a generalized section of the so-called Great Flat Lode of Redruth in Cornwall, shows an instance. A is a fissure, called the Leader, filled partly by mechanically-formed debris and partly by crystallized Quartz, Copper Ores, and a little Cassiterite. B is a belt of Stanniferous Schorl rock 4 to 15 feet wide, consisting of Schorl and Quartz; it is made up of strings, veins, and spots of Quartz, * Foster, Quart. Journ. Geol. Soc. xxxiv. (1878) 655. 556 Geology, Cassiterite, Iron Pyrites, and Chlorite in a compact black matrix. It contains from 1 to 3 per cent, of Cassiterite, disseminated in little ^>f- D Fig. 191. SECTION ACROSS THE GREAT FLAT LODE OF EEDRUTH. grains, strings, or minute veins. It occurs sometimes above, some- times below, arid sometimes on both sides of the Leader. C, "Capel," "Grey back," "Black Granite." A Schorl rock, con- sisting of large grains of Quartz in a compact black matrix and very little Cassiterite. D, Granite. F, " Killas," Clay-slate. E, Schorl rock like (7, but foliated, the planes of foliation being parallel to the bedding of the Clay-slate F. Where the Leader has Granite on both sides, the Schorl rock both above and below is devoid of foliation. Except the Leader there is no hard line between the different kinds of rock. The Granite passes gradually into the amorphous Schorl rock (7, and this becomes richer in Tin Ore as it approaches the Leader. The Killas in the same way graduates into the foliated Schorl rock E, and in this also the percentage of ore increases as we draw nearer the Leader. There can be little doubt that the Leader has .been the channel by which the Cassiterite has been introduced into the rock, and that at the same time the Granite and Clay-slate adjoining it have been altered into Schorl rock.* An explanation of the way in which this may have been brought about will be given presently (p. 562). 3. ON THE WAY IN WHICH LODES AND STOCKWORKS WERE FILLED. The problem as to how the contents of lodes and the fissures of stock- works were introduced is one towards the solution of which but little progress has been yet made. On one or two points however geologists have been able to feel their way towards an explanation, and there are * Foster, Quart. Journ. Geol. Soc. xxxiv. (1878) 640. Metallic Deposits. 557 some theoretical views on the subject probable enough to call for notice. The following are some of the methods which have been suggested. By the Chemical Action of Percolating Fluids. The facts accumulated by mining experience form a huge chaotic mass which at first sight there seems no hope of reducing to law and order. A few general truths however stand out with a certain amount of distinctness, among which are the following. There are a certain number of the metals which occur more com- monly in lodes than in any other form of mineral deposit, and, when they occur in lodes, their most plentiful and widely- diffused ores are compounds devoid of Oxygen. The commonest ores of Copper, Lead, Zinc, Antimony, and Mercury are Sulphides : Silver Cobalt and Nickel are met with most frequently as either Sulphides or Arsenides. Chlorides, Iodides, Bromides, Selenides, and other non-oxidized com- pounds of some of these metals occur less frequently. Further when Sulphates, Arsenates, Carbonates, Oxides, and other oxidized compounds do occur, they are confined to the upper parts of the lodes, where oxidizing agents are able to penetrate ; they are in many cases known to have been formed by the oxidation or other alteration of Sulphides; and they are common only in the case of these metals which are easily oxidized, and rare in the case of Silver and Mercury which oxidize slowly and with difficulty. There seems very little doubt then that the metals we have men- tioned were deposited originally in the lodes as Sulphides or in some cases as Arsenides or other compounds devoid of oxygen. The same is probably true of Manganese in spite of the fact that it occurs almost always as an Oxide. A natural Sulphide is known, and the excessive readiness with which Manganese oxidizes will account for its almost universal occurrence in an oxidized condition. In fact we may almost say of Manganese that it is one of these rare cases where the exception does prove the rule. The problem then of the filling of the lodes which yield the metals enumerated, reduces itself to the discovery of a method by which Sulphides of these metals could be naturally formed and deposited in lodes. And since the lodes from which these metals are obtained form by far the larger part of all lodes, if we can answer this question, we shall have gone a long way towards furnishing a general solution of the problem before us. Metallic Sulphides have been produced artificially by a variety of processes, but we may first notice some cases in which by a great piece of good luck their formation by natural agencies has been actually witnessed. This has occurred at more than one locality ; the instances which have been most carefully inquired into happened at Bourbonne-les- Bains, Plombieres, and some other places in France where hot mineral springs come to the surface. For an account of the facts we are mainly indebted to M. Daubre'e.* The waters of Bourbonne have a temperature of 68 C. (154 F.) where they issue from the rock. They contain from 7 to 8 grammes per * Geol. Experimentale, i. 71. 558 Geology. litre of solid matter which consists mainly of Chlorides and Sulphates of Alkalies, Lime, and Magnesia, of an Alkaline Silicate, and traces of Arsenic and Manganese. Ammonia, Iodine, Copper, Lithium, Stron- tium, Rubidium, and Caesium are also present. The gases given off are chiefly Nitrogen with small quantities of Oxygen, Sulphuretted Hydrogen, and Carbon Dioxide. In cleaning out an old Roman well which went down to the source of one of these springs more than 4700 Roman coins of bronze, brass, silver, and gold were found buried in mud and sand ; many of these coins dated back to the time of Augus- tus, others were as late as Constantine and Magnentius. Assuming that they fell in not long after the date of the most recent these coins had been subjected to the action of the thermal waters for a period of some fifteen centuries. With the coins were other metallic objects such as tubes and rings, and also organic substances. In the bottom of the well was a bed consisting mainly of bits of sandstone. Many of the coins here had disappeared completely leaving only their im- pression, and the fragments of rock were cemented into a conglomerate by a crystalline cement with metallic lustre in which the following minerals were recognised. Cuprite in octahedrons, Copper Glance twinned as in natural crystals, Copper Pyrites in octahedrons and mammillated masses, Bornite in crystals, crystallized Tetrahedrite in great plenty, Covellite (CuS), Atacamite, and Chrysocolla ; Phosgenite (PbC0 3 .PbCl 2 ) in which were crystals of Galena, and Anglesite ; a coating of amorphous Oxide of Tin. On the dressed stones of a pave- ment were crystals and incrusting coats of Iron Pyrites, and in the cavities of bricks crystallized Calcite. In this case it is probable that the metals present had been first converted into Sulphates, and that these were afterwards reduced to Sulphides by the action of organic matter. This is a process which can hardly have gone on in the deep- seated parts of many lodes, and therefore we are not justified in assert- ing that there is any close analogy between the formation of these minerals and the method by which Sulphides were deposited in lodes. The facts however are very valuable, because they show that, where the requisite elements are present, there is a very strong natural ten- dency for them to unite in the wet way into the identical combinations which are met with in nature. It furnishes strong presumptive evi- dence that the wet way has been one at least of the methods employed in the formation of these natural compounds. Metallic Sulphides have been artificially formed directly in the wet way under ordinary temperatures and pressures. Bischof obtained small crystals of Galena by passing Sulphuretted Hydrogen very slowly through a very dilute solution of a lead salt. In this and similar experiments the process had to be continued for a long time, and one reason why comparatively few attempts of this kind have been successful is probably because they have not been continued long enough. Here, as in so many of our trials to imitate natural processes, scale is everything ; there is no parallel between what we can attain to even in the period of a lifetime and what Nature can effect in the ages through which she keeps patiently at work. And, as has been pointed out when speaking of metamorphism, Metallic Deposits. 559 there are two other conditions in which the most ingenious of our experiments fail utterly to come up to the operations of Nature. She has pressures and temperatures at her command far exceeding any we can avail ourselves of. That pressure and temperature increase enor- mously the solvent power of liquids we can realize experimentally ; even so intractable a substance as Heavy-spar has been by their aid dissolved in acidulated water and obtained again in a crystalline form ; what is the law that regulates the increase and to what extent it may go, we do not know, but experiments like that just mentioned justify us in saying that there is nothing to forbid the belief that the contents of lodes may have been introduced into them in solution by water or other fluids. Attempts have been made even to go further than this. Mr. Wallace has endeavoured to show that the lodes of Alston Moor are most productive in those parts which are most favourably situated for the free percolation of water from the present surface of the ground and that the parts where the underground circulation is checked are barren of ore.* He infers that the lodes were filled at a time geologically recent, and thinks it not unlikely that the deposition of ore is even now going on. That ore is now forming is perfectly possible, but the surface of that part of the country has been practically what it is now for such a very long period that even on his showing the lodes may have received their contents much farther back than he supposes. Assuming then that the contents of the lodes we are dealing with were brought in in solution, the next question is, Where did the metallic and other elements come from ? The fact that lodes bear so much more plentifully in certain rocks than in others has led to the conjecture that metals or metallic compounds exist very finely dis- seminated through these rocks and that they have been slowly dissolved out thence and conveyed by percolating liquids into the lodes. In some cases research has brought facts to light which make this explanation in the highest degree probable. In the Black Forest Sandberger has shown that where the Mica of the adjoining Gneiss contains very minute quantities of Cobalt, Arsenic, Copper, Bismuth, and no Lead, the lodes yield Smaltite, Cobaltiferous Fahlore, Copper Pyrites, and Bornite, but no Lead Ores ; where the Mica contains Silver, Arsenic, Bismuth, Cobalt, Nickel, but little Copper comparatively, and no Lead, the lodes carry Arsenical Ores of Silver, Cobalt, and Nickel, but ores neither of Copper or Lead. It is very much to be wished that similar investigations could be carried on in the case of Alston Moor, where the difference in the ore-carrying properties of the different beds is so marked. But even if we admit that the ores came out of the adjoining rocks, this carries us back only a single step, for we must next ask, How did they get into these rocks 1 Where we have to deal with intrusive Igneous rocks, we do see our way towards an answer. Some ores, such as Magnetite, Ilmenite, and Iron Pyrites, are common in Crystalline rocks and occur in such a way that they must have been formed during their cooling contein- * On the Laws which regulate the Deposition of Lead Ore in Veins. 560 Geology. poraneously with their other constituent minerals. And other ores less common have been artificially produced by the action of heat, and may therefore have been formed during the cooling of the Crystalline rocks. Among these are Copper Glance, Galena, Antimonite, Bismuth- ite, Proustite, Pyrargyrite, Bornite, and others. Again observations like those of Sandberger's have shown that a large number of metals are present in the Micas, the Augite, the Hornblende, and the Olivine of the Crystalline rocks. The percentage is extremely small, but quite large enough to furnish, when concentrated, metallic deposits, for the percentage which these form of the whole mass of the rock is no larger. By the decomposition of these minerals metals are set free and metallic compounds formed in the great underground laboratory ; at high temperatures and under great pressure many, perhaps all, of these compounds are soluble ; in solution they travel about through the body of the rock and may be deposited in any fissures they come across. Clastic deposits formed by the denudation of such Crystalline rocks, will necessarily contain some of these metallic compounds ; and when such a rock is permeated by liquids these compounds may be dissolved out and deposited in the fissures which cross it. Some such general explanation as this we may safely accept, and though we may never be able to say for certain what was the exact chain of reactions by which any given deposit of ore reached its present home, we can easily picture to ourselves many very possible processes by which such a result may have been brought about. For instance many Crystalline rocks contain Magnetite in large quantity ; we know by experiment that if Sulphuretted Hydrogen is passed over red-hot Magnetite, Iron Pyrites is formed ; Sulphuretted Hydrogen is largely given off by Solfataras ; at no very great depth the temperature would be high enough for this transformation, and we should have the materials and the conditions for the formation of Iron Pyrites. Experiment also proves that at these depths Iron Pyrites would be soluble in acidulated water and even in water alone ; in such situations then everything that is required for the filling of fissures with this mineral is present. When upheaval and denudation brought the rock to light, it would be traversed by lodes of Iron Pyrites. If we want to go further back yet and ask whence came the minute quantities of metal that occur in the minerals of the Crystalline rocks, we get on very theoretical ground but we are not quite at a loss for an answer. The inside of the earth is certainly composed of matter much denser than that which forms its crust, and it is likely enough that these metals exist there in quantity and that this is the storeroom from whence they originally came. Among the metals we have been dealing with Iron was not included, because, though its sulphide is one of the commonest of minerals, it occurs also in many other combinations to probably quite as large an extent. Iron is so universally present and the changes which its com- pounds are incessantly undergoing are so numerous that it is scarcely possible in many cases to conjecture what was their original condition when they first reached the surface, and what and how many modifica- tions they have there undergone before they acquired their present Metallic Deposits. 561 form. One or two of these however may be usefully noticed here, though they belong only in part to the special matter now before us. By the action of carbonated water Ferrous Carbonate is formed out of some iron compound in a rock and is carried away in solution. It is a very unstable salt and rapidly oxidizes where air has access, and Ferric Oxide or Ferric Hydrates are precipitated. These may be again reduced by organic matter and brought back to Carbonate. Oxidation might not take place underground, and as long as the water retained its Carbonic Acid the Iron Salt would remain in solu- tion. It could however be precipitated in this way. The decomposi- tion of Felspars and suchlike minerals yields soluble Alkaline Silicates ; if water holding these in solution come in contact with carbonated water, they are decomposed, the Alkalies combine with the Carbonic Acid of the water, Alkaline Carbonates are formed and carried away in solution, and Silica may be precipitated. If then a solution of these Alkaline Silicates come in contact with a solution of Ferrous Carbonate in carbonated water, the result will be the abstraction of the Carbonic Acid from the water, and consequently the precipitation of the Ferrous Salt ; Silica at the same time will be thrown down. By such a reaction a fissure may be filled with Spathic Iron Ore and Quartz, an association which frequently occurs in nature. To sum up, we may look upon the inside of the earth as the original home of the metals. Ascending thence, we do not know how, they are carried still higher by those masses of fused matter which from time to time force their way upwards through the crust. When by the cooling of these masses Plutonic or Volcanic rocks are formed, they either combine in very minute quantities with other substances to form the minerals of these rocks, or they crystallize out in distinctly metallic compounds like the Magnetite and Iron Pyrites so common in Crystalline rocks. Then follows a chain of chemical reactions, which we can do no more than dimly imagine, by which metallic compounds are formed, transformed, carried in solution, and at last deposited, in many cases as Sulphides, in fissures. By the oxidation of these Sulphides other compounds are formed at or near the surface. It remains to notice that the filling of lodes must in many cases have stopped and then gone on again, and this many times over. We see a proof of this in Comby Lodes. The two combs of every pair were laid down on the walls of the fissure at the same time ; but there must have been an interval and a change in the contents of the solution between the formation of two successive pairs of combs. When the lode is unsymmetrical, it must also have been opened afresh after the first comb was completed. Thus in fig. 188, if we suppose the left- hand set of combs the older of the two, the original fissure extended only from A to B and was filled by the deposition of the comb AabB : then the fissure BbcC was torn open and the new rent was filled by the deposition of a second comb within it. In one very striking case one-half of the lode was filled by a set of crystallized combs, and the other half by a breccia. It is likely that the deposition of ores in lodes has in many cases gone on at considerable depths. But the occurrence of rolled pebbles 2N 562 Geology. in lodes shows that there must in some cases have been communication with the surface, and the fact that these pebbles are coated with ore proves that deposition of metallic compounds went on after they were introduced. When these coatings are Sulphides, they probably came in dissolved in water as Sulphates, and were reduced by organic matter which might readily find its way down at the same time as the pebbles. By Sublimation from below. Galena, Zinc Blende, and other metallic Sulphides have been produced by sublimation both experi- mentally and in smelting-furnaces, and this has suggested the notion that lodes containing such ores may have been filled in the same man- ner. In those lodes when the Vein-stuff is Calcite or Heavy-spar, deposition from solution is to say the least a far more probable method, for these minerals have been formed in the wet way and have not been obtained by sublimation ; and the explanation is scarcely admissible in those cases where the walls show no trace of the alteration which the ascent of heated vapours would be likely to produce. liut there is one case in which the action of heated vapours fur- nishes a very probable explanation of the observed facts. In the manner of its occurrence, Tin contrasts in two respects in a very marked way with nearly all the other metals. It occurs practically always as an Oxide, for Tin Pyrites, the only natural Sulphide into which Tin enters, is a mineral of rare occurrence. It is also met with to a comparatively small extent in true lodes ; by far the larger part of the Tin Ore raised is obtained either from Stockworks or from deposits of the type of the Great Flat Lode of Kedruth, described on p. 555 ; and in these cases it is not found prin- cipally in the fissures themselves, but disseminated through a band of rock on either side of them. In all these cases the Tin Ore has been obviously introduced into the rock since its formation, it is a secondary product, and during its introduction the rock has undergone great alteration. This is proved, where that rock is Granite, by finding pseudomorphs of Cassiterite and Quartz after Orthoclase, the Felspar being sometimes wholly and sometimes only partly replaced. The Orthoclase is also very largely changed into Kaolin. The Granite in the Cornish case just quoted has been altered into Schorl rock in which pseudomorphs of Tourmaline after Orthoclase occur. At Zinn- wald in the Erzgeberge the tin-bearing rock is Greisen consisting of Quartz and Lepidolite. It contains Tourmaline and pseudomorphs of Quartz after Orthoclase, and passes into Granite. The Zwitter rock or Stockwerke-porphyr of Altenberg in Saxony is another famous receptacle of Tin Ore. It consists of Quartz shot with Micaceous Iron Ore, and contains, besides Cassiterite, Chlorite, Mispickel, and grains of crystalline Quartz. It passes into Granite. The adjoining Granite is traversed by numerous veins of Quartz, and alongside each there is a Salband of Zwitter which merges into Granite. There can be no doubt then that the Cassiterite was introduced into the rock by some agent which travelled along fissures and produced the metamorphic changes described. The question is, What was that Metallic Deposits. 563 agent ? One point in the composition of the associated minerals helps us to conjecture. Tourmaline and Lepidolite we have already men- tioned ; in addition the following are found, some or all of them, in stanniferous deposits all the world over ; Topaz, Apatite, Fluor- spar, and in Greenland Cryolite ; all these minerals contain Fluorine, some of them Boron as well. Axinite is common in deposits of Tin Ore, and it contains Boron. Quartz is also exceptionally abundant in those parts of the rock which yield Tin Ore. Other common associ- ates are Wolfram and Mispickel. The most conspicuous fact among those just mentioned is that of the minerals constantly associated with Tin Ore and evidently intro- duced at the same time with it into the rock, a large number contain Fluorine. This suggests that the mineralizing agent was a compound of Fluorine and Tin, and of such compounds Stannic Oxyfluoride seems the most probable. It is volatile and is not decomposed at a high temperature ; it could therefore ascend fissures in a state of vapour. If its vapour comes in contact with steam, it is decomposed ; Stannic Oxide is formed and Hydrofluoric Acid set free. Sn 2 F 4 O 2 + 2H 2 O = 2SnO a + 4HF1. Thus we get Cassiterite, and the changes that have accompanied its deposition would seem to follow naturally. The Hydrofluoric Acid would act on the silicates of the adjoining rocks. It might thoroughly decompose them, carry away their silica as Silicon Fluoride, and after- wards this compound might be decomposed in the presence of water and Quartz might be deposited. Or the Hydrofluoric Acid might work only partial alteration, converting for instance, if Iron were present, Orthoclase into Tourmaline ; and if Lime were present, it might form with it Fluor-spar. This view accounts so thoroughly for all the facts connected with the occurrence of Cassiterite, that it may be looked upon as a highly-pro- bable explanation of the mode of its formation. It is further sup- ported by the experiments of Daubree, who obtained crystallized Cassiterite by decomposing the vapour of Stannic Chloride by steam at a white heat.* Other metallic ores have been produced experimentally in a similar way. By the action of steam on Chlorides of Iron and Titanium, Specular Iron Ore and Rutile h'ave been manufactured. By bringing together volatile metallic Chlorides and Sulphuretted Hydrogen many natural Sulphides have been obtained. Methods like these may well have been employed by nature, but I do not know of any instance in which the evidence in favour of this having been so is so strong as in the case of Tin. The production of Specular Iron Ore by sublimation in the fissures of lava has been noticed on p. 362 : the Tenorite of Vesuvius has doubt- less been formed in a similar way by the alteration of Cupric Chloride, which is known to come up in a state of vapour. By Electro - chemical Action. By the action of weak * Geol. Experiraentale, i. 28. 564 Geology. galvanic currents on metallic solutions, Becquerel formed Sulphides of Silver, Copper, Tin, Lead, and Iron, and Oxides of Copper and Zinc. Mr. Fox by similar methods obtained in a mass of clay sundry metallic compounds deposited in fissures traversing the clay in a way closely resembling their occurrence in lodes. He also ascertained that galvanic currents are now active in actual lodes. From these facts he was inclined to attribute the deposition of ores in lodes to the action of these currents, to a kind of natural electrotyping in short.* Mr. R Hunt has repeated some of Mr. Fox's experiments with similar results, t There can be no question as to the existence of galvanic currents in lodes, and they very probably play an important part in modifying and rearranging their contents ; but whether they were in any case the means by which the ore was originally deposited is doubtful. It seems more likely that it is the presence of different kinds of ore in different parts of the lodes that sets up the currents which have been observed. Different ores may correspond to the elements, and underground mineral waters to the exciting liquids of a battery ; in fact in some of Mr. Fox's experiments different ores were actually used as elements. Some experiments of Reich are in favour of this view. When he connected two different points in a lode both containing ore by a conducting wire, he obtained a deviation of the needle of his galvanometer ; but he could not detect the slightest trace of a current by connecting points barren of ore.J 4. MASSES. The deposits we class under this head are like lodes sharply marked off from the rocks in which they occur, but instead of being narrow in comparison with their length they are approximately circular, elliptical, or lenticular in horizontal section. In some cases too instead of running down to inaccessible depths, they terminate down- wards at a moderate distance from the surface. Iron Pyrites Deposits of Andalucia. Among the most striking of metallic Masses are the great deposits of Iron Pyrites in Andalucia and the adjoining parts of Portugal. The rock of the country is Clay-slate, with a general north-easterly strike and dipping at high angles : here and there are intrusive masses of Diorite and other Crystalline rocks, they are generally lenticular in horizontal section, with the longer axes parallel to the strike. The Pyritous masses occur frequently in the neighbourhood of these intrusive rocks ; they are also lenticular in horizontal section and their longer axes run parallel to the strike ; they dip down in the same direction and approximately at the same angle as the Slates. They are very large : at the Tharsis Mine there are four masses, two of which are 2600 yards long and 200 yards across in the broadest part ; the deposit at Rio * Phil. Trans. 1830 ; Report of the Royal Polytechnic Society of Cornwall, 1836. "r Memoirs of the Geological Survey of Great Britain, i. 450. % Poggendorf, Ann. xlviii. 287. Metallic Deposits. 565 Tinto is equally large, and some of the other masses are little inferior in size. No bottom has been reached in the larger masses : some of the smaller ones wedge away downwards to nothing. The ore is granular Marcasite, with a small admixture of Copper Pyrites, traces of other ores of Copper, Lead, and Zinc, and about 4 per cent, of Silica. There are large " horses " of rock, and a few veins of Quartz here and there, but with this exception there is no vein-stuff, the mass is practically all ore. The masses are surrounded by a Salband con- sisting mainly of very hard porcelainized Clay-slate ; the belt however immediately in contact with the ore has been often corroded and softened, probably by the action of Sulphates produced by the oxida- tion of the Pyrites. Breccias of porcelainized rock and Quartz with a ferruginous cement also occur in the Salband. The walls of the mass are highly polished and slickensided. A gossan, called locally a Colorado or Montera de hierro, overlies each mass : it consists of Ferric Hydrate, Red Clay, and fragments of Slate : it reaches in some cases a thickness of 160 feet. I cannot say that I have ever seen my way to a reasonable ex- planation of the manner in which these huge masses were formed. They seem to be connected with the intrusive rocks, and these 'con- tain very often a large quantity of Iron Pyrites. I hardly dare suggest the possibility of their being themselves intrusive masses, but something may be said in favour of such a view. The materials for the formation of Iron Pyrites have been in many cases present in the fused magma of an intrusive rock in small quantity, and that mineral has crystallized out and 'now forms one of the original minerals of the rock. It is within the bounds of possibility that the elements of Iron Pyrites have in some cases formed the bulk, instead of a small part, of the magma ; and that in cooling they separated from the rest of the magma and hardened into distinct and independent bodies of this mineral. The metamorphism of the surrounding Slate and the friction- breccias of the Salband would be explained on this hypothesis. Deposits of Haematite in Cumberland. These are large irregularly-shaped masses of Haematite enclosed in Carboniferous Limestone. The mineral is mainly massive and compact. Kidney ore is a frequent form and crystallized Quartz and Iron-glance coat the walls of cavities in the mass. It is surrounded on all sides and often roofed in also by solid Limestone. There can be no doubt here that the ore fills in caverns and under- ground watercourses, and the only question is where it came from and under what form it was brought in. It probably w r as introduced from the surface, for I believe none of these deposits have been found below a certain moderate depth. In some deposits a bedded arrangement and the presence of interstratified bands of Shale point to deposition by water.* The overlying rocks give us a hint as to the origin of the mineral. * In one case at least the roof of the cavern is formed, as is often the case in caverns, by the under-surface of a bed of Limestone. This together with the interbedded Shale bands give the deposit the appearance of a true bed, which is probably deceptive. 566 Geology. They are Sandstones stained deep red by Ferric Oxide ; they belong to the red rocks that were deposited in closed lakes whose waters were strongly charged with Salts of Iron (see p. 244). Some breccias at the base of this group of Sandstones, which rest directly on the Limestone and which have very much the character of old screes, contain a large amount of Haematite and pass downwards into deposits of that mineral that fill up pipes and hollows in the Limestone underneath. We may therefore reasonably suppose that the Haematite was derived from the Iron-charged waters of these lakes, and carried down by per- colating water through swallow-holes into hollows and caverns. The pressure of the overlying rocks converted what was originally Haematite mud into a solid form, and part of the material subsequently gathered together into kidney-shaped nodules. If, as is likely, the Iron Salts came from a volcanic source, the water might well contain Silica in solution, and this may be the origin of the beautiful crystals of Quartz which accompany the Haematite.* Where the roof of a cavern full of Haematite has been carried away by denudation, exposure to the weather has reduced the ore to a soft Haematite mud largely made up of filmy micaceous scales. This is used for forming the beds of puddling-furnaces and is hence known as " Puddling Ore," the solid Haematite being called " Blast Ore." 5. IMPREGNATIONS. Sometimes, instead of being collected in lodes or stockworks or masses, metallic ores are disseminated through the whole body of beds or masses of rock. The particles may be crystals or rounded grains visible to the eye or they may be so small as not to be separately distinguishable. In other cases the grains of a rock are coated over with films of metallic compounds. Alderley Edge. A good case occurs at Alderley Edge in Cheshire. The rock of which this hill is composed is sandstone and conglomerate. In certain beds the grains of sand are coated with Malachite and Azurite to such an extent that the rock is worth raising and treating for Copper. Other ores also occur in the rock, Galena, Cerussite, one of the Copper Phosphates, Wad, Spathic Iron Ore, Iron-glance, and Heavy-spar : ores of Cobalt are also present in work- able quantity. It seems not unlikely that the ores in this case are mechanically- formed debris derived from the denudation of lodes. Originally most likely Sulphides, they have been converted into oxidized ores by the action of percolating water. Kupferschiefer. A still more striking instance is the Kupfer- schiefer or Copper-slate of Mansfield in Thuringia. The general section of the Permian rocks in which it occurs is Zechstein .... Dolomite, Rock-salt, and Gypsum. i i i * f Finely-laminated hard bituminous Marl-slate and Copper-slate . ^ rf ^ thickness . * Harkness, Quart. Journ. Geol. Soc. xx. (1864) 152. See also The Iron Ores of Great Britain (Memoirs of the Geological Survey), part i. p. 20. Metallic Deposits. 567 Weisslegendes . . "White Sandstone. Rothtodtlegendes . Red Sandstone and Conglomerate. The ores are found in the Marl-slate, chiefly in the lower part which is specially distinguished as Copper-slate, either finely and invisibly disseminated, or in layers, fissures, and little nests. The principal are Copper Pyrites, Bornite, Copper-glance, Iron Pyrites, Zinc Blende, Argentite, Copper Nickel, and Cobaltite. The upper beds of the Weisslegendes are sometimes penetrated by strings of ore. In the bottom division, the Red-dead-underlying rock, ths ores dis- appear. It seems likely that here the ores were formed in part at least by chemical precipitation at the time the rock was deposited. Fossil fish are abundant, and the state of their preservation and their large number shows that they must have been killed and buried suddenly. The whole group of rocks is of chemical origin and must have been formed in closed lakes. During the formation of the Marl-slate the water became occasionally habitable and tenanted by large shoals of fish. Then water strongly charged with salts of Copper and other metals was poured in and every fish was instantly poisoned and quickly buried in mud. The soluble salts were reduced by the decay- ing organic matter and thrown down as Sulphides. Much rearrange- ment of the metallic compounds has evidently gone on since their formation, but the accounts do not seem to imply that they show any signs of having been originally mechanically-formed debris. The large amount of organic matter present has prevented their oxidation and the majority of them are still Sulphides. Plumbiferous Sandstone of Bleiberg in the Eifel. Here in the Bunter Sandstone there is a bed of coarse friable Sandstone through which are disseminated sandy concretions about the size of a pea containing crystallized Galena, rarely Cerussite, and still more rarely oxidized Copper Ores. The rocks are inland-sea deposits and partly of chemical origin, and the water in which they were formed must have had some Lead-salt in solution. We have already seen that under these circumstances Sulphuretted Hydrogen can be made to throw down Galena, and the ore was doubtless formed by this or some similar reaction. I believe it will be found to be universally the case that where metallic impregnations occur, the rocks that contain them belong either to the class of chemically-formed deposits or are volcanic. Native Copper of Lake Superior. The deposits of Native Copper in the north-western peninsula of Michigan are found some of them in true lodes, others may be regarded as impregnations. The rocks in which these latter occur are Sandstones and Conglomerates with interbedded flows of Amygdaloidal Doleritic lavas. The Amygdaloidal kernels are coated with thin sheets of Copper deposited on the \valls of the cavities or over crystals of some previously-formed mineral, and in addition thin plates of the metal fill chinks and rents in the rock. Copper is distributed through the whole thickness of the beds but tends to be concentrated along the bottom of each bed. In the Con- 568 Geology. glomerates the metal incrusts the pebbles, sometimes completely covering them. In the case of one Conglomerate "the cement is entirely metallic, the Copper forming closely-fitting shells over the pebbles, and at times permeating them to such an extent as to form with the siliceous mass a kind of Copper-concrete." Here again it is the lower part of the bed which is cupriferous. With regard to the origin of this Native Copper we must bear in mind that though certain parts of the deposit are now very rich in metal, if the amount present were uniformly distributed through the body of the rock, it would not form perhaps as much as 1 per cent, of the whole mass. Now minute quantities of Copper have been detected in Orthoclase, Idocrase, Olivine, and other constituent minerals of the Crystalline rocks, and compounds of Copper have been found in small quantity in lavas. It is also possible that the lava originally contained finely- disseminated Copper Sulphides as is the case occasionally with Copper slags. By the decomposition of the Copper-containing minerals, or by the alteration of the Copper compounds, soluble salts of Copper may have been formed, and these may have been reduced and metallic Copper precipitated by the action of organic matter. Such a reaction has been observed to take place by Bischof. The concentration of Copper towards the lower part of each bed is strongly in favour of its having been brought into its present position by a liquid. In the Sandstones and Conglomerates it is possible that comminuted Sulphides of Copper, derived from the denudation of lodes, may have been mechanically deposited along with the sandy and pebbly sediment ; oxidation may have converted these Sulphides into soluble Sulphates, and the Sulphates may have been reduced to the metallic state by the action of organic matter.* Ore-bearing Crystalline Rocks. Magnetite is, as we have several times had occasion to notice, frequently minutely disseminated through certain of the Crystalline rocks ; and Iron Pyrites, Copper Pyrites, Ilmenite, and other metallic minerals occur under similar con- ditions. Occasionally ores are found disseminated in Crystalline rocks in such quantity and in lumps of such a size as to make the mass worth working. Magnetite, Ilmenite, and Iron-glance are the commonest, but metallic Sulphides also occur in this fashion. Chro- mite is also found, generally in connection with Serpentine. The ores here are no doubt original constituents of the rock. Metals were present in the fused mass, and instead of entering into the com- position of the Silicates that were formed during its cooling, as happened in many instances, they collected into compounds more exclusively metallic in character, somewhat in the same way as Magnetite is formed in puddle-furnace slag and Copper Pyrites in the slags of copper-furnaces. One of the effects of contact metamorphism has been in some cases the formation of metallic ores along with other minerals in the altered rocks, specially when these are Limestones. * Bauerman, Quart. Journ. Geol. Soc. xxii. (1866) 448. Metallic Deposits. 569 6. BEDS. Deposits of ore sometimes occur in true beds regularly interstratified with sedimentary rocks. In such cases they have been formed in several ways ; they may have been deposited either as mechanical debris or by chemical precipitation at the same time as the rocks among which they are found ; or a bed of rock, a Limestone for instance, may have been altered and converted more or less completely into a metallic ore. Bog-iron Ore. These deposits, which have been already noticed (p. 299), furnish a good instance of bedded deposits of metallic ore now in process of formation. Clay Ironstone. This mineral occurs constantly in Shale either in thin beds or more frequently in nodules ranged along the planes of bedding. It consists of Ferrous Carbonate mixed with more or less clay : the Shales which yield it generally contain the remains of plants and much finely-disseminated vegetable matter. We may suppose that the water in which the Shale was laid down contained Ferrous Car- bonate in solution. The salt was precipitated by the escape of Carbon Dioxide from the water, and its oxidation was prevented by the vegetable matter present. The beds of mud which were deposited while this precipitation was going on would be charged throughout with Ferrous Carbonate, and in course of time the salt separated out and collected into nodules by the process which we are pleased to call Concretionary Action. Cleveland and Northamptonshire bedded Iron- stones. We have already mentioned that the Cleveland Ironstone is a bed of Limestone in which Calcium Carbonate has been replaced by Ferrous Carbonate, a pseudomorph of Siderite after Calcite in fact on a large scale (p. 412). Professor Judd has explained on a similar principle the origin of the ores of the Northampton Sand, and has traced out the series of changes by which the deposit has been brought into its present form.* Bedded Iron Ores of Canada. In the highly-altered Lauren- tian and Lower Silurian rocks of North America true beds of Iron Ore, varying in thickness from a few inches to 200 feet, occur. Magnetite is the commonest ore, but Iron-glance and Haematite also occur. These deposits are usually associated with or are not far removed from beds of crystalline Limestone. The alteration which the rocks of the district have undergone is so great that it is quite impossible to say what the original nature of these ores may have been. That they were deposits formed in lakes seems very likely : possibly they were produced by the oxidation and precipitation of dissolved Ferrous Carbonate : they may have been Bog-iron Ores ; the occurrence of Apatite in large quantity among them would be explained on this hypothesis, for these ores contain a large amount of Phosphoric Acid. It is not clear from the descriptions whether the great deposits of * The Geology of Rutland (Memoirs of the Geological Survey of England and Wales), chap. vi. 57 Geology. Magnetite at Danemora and elsewhere in Sweden are true beds or masses. They occur in highly-altered rocks, and the crushing and dislocation have been so great in many such cases that a bed of hard material is often forced through adjoining rocks of more yielding composition and a deceptive appearance of intrusive behaviour is produced. Ore also which originally formed a true bed may be carried in solution and deposited in cracks and fissures formed during the dis- turbance of the rocks, and if the bed be tilted nearly into a vertical position, a close resemblance to a lode will be produced. Under such circumstances great caution is necessary before pronouncing on the character of a metallic deposit. 7. PLACERS. This form of metallic deposit calls for only short notice. When rocks are denuded and the waste is carried away by running water, the material is sorted according to its bulk and specific gravity, the heaviest falling to the bottom first. If the rocks contain metallic ores or native metals, these will be among the first to come down, because they are heavy and also because in some cases they are hard and not easily broken small. If further such a deposit be rolled about and rearranged many times over, the bulk of the metallic fragments will at last find their way to the bottom. In the gravels that are formed on the beds of rivers which run across metalliferous rocks mechanically-carried fragments of ore or metal are thus accumulated in sufficient quantity to be worth raising. Cassiterite (Stream Tin), Gold, and Platinum are the substances obtained most largely from such deposits. In a large number of instances alluvial deposits of ore are of com- paratively recent date ; but sometimes old metalliferous gravels have been sealed up beneath lava-flows or otherwise protected from denu- dation, and have thus survived down to the present day.* * Selwyn on the Gold-fields of Victoria, Quart. Journ. Geol. Soc. xiv. (1858) 533. CHAPTER XIII. HOW THE PRESENT SURFACE OF THE GROUND HAS BEEN PRODUCED. "Now concerning the exaltation of the mountains above the valleys it ap- peareth to come to pass by the water in former times, whose property is to wear away by its motion the most loose earth, and to leave the more firm ground and rocky places highest." A DISCOVEEY OF SUBTERRANEAL TREASURES (GABRIEL PLATTER, 1738). " That mighty trench of living stone, Where Tees, full many a fathom low, Wears with his rage no common foe ; Condemned to mine a channeled way Through solid sheets of marble grey." SCOTT. SECTION I. PROOFS THAT THE SHAPE OF THE SURFACE IS DUE TO DENUDATION. WE have now made ourselves acquainted with the processes by which the materials that compose the ground on which we live and move were brought together, compacted into their present form, and placed in their present position ; our next step will be to inquire how the surface of that ground has had its present shape given to it how mountain-chains, tablelands, hills, valleys, and plains, and all the lesser inequalities that diversify the face of the earth, were produced. Surface due to Denudation. We saw in the last chapter that the crust of the earth has been from time to time crumpled and folded into troughs and arches, and nothing would be more natural, when we see a mountain or hill, than to suppose that it is one of the arches, and that the valley which lies at its foot runs along the line of one of the troughs ; that in fact if we were to cut a deep trench Fig. 192. SECTION SHOWING WHAT WOULD BE THE GEOLOGICAL STRUCTURE OF A COUNTRY IF THE HILLS COINCIDED WITH ANTICLINALS. across hill and valley, we should see in its sides that the rocks were arranged underground as in fig. 192. 5/2 Geology. Nothing certainly could be more natural than to suppose this would be the case ; but a very little examination suffices to show us that no supposition could by any possibility have been made so utterly contrary to fact. In a very large majority of cases we find that the rocks that form a hill lie in a trough, instead of being bent up into an arch ; and that a line of valley, instead of coinciding with a trough, runs along the crest of an arch on the rocks below. In other cases the hills and valleys have apparently not even so much connection as this with the folds into which the rocks have been bent. And even in those cases where the rocks forming a moun- tain are arched and plunge down on each side in the same direction as the slope of the ground, if we draw a section across the hill, and are careful to put in the inclinations of its sides and of the beds composing it, as they occur in nature, we see at once that the arch is incomplete, that large portions of it have been carried away, and that, though the formation of the hill may have begun with a bending up of the rocks, some other cause must have operated on it to give it its present outline. The section in fig. 193 would give a truer idea of the relations between the shape of the ground and the lie of the rocks beneath it. On the left we have two hills, the rocks of which lie in troughs, and valleys between cut out of the crests of arches in the rocks ; then follows a tract where the beds are folded into sharp curves, but the surface, instead of following these curves, has been planed away till it cuts across them in every direction. On the right are lofty moun- tains, from the summit of which the beds dip away on each side in the same general direction as the slope of the ground, and where surface outline does follow to a certain extent the flexures of the rocks, and a broad valley between in which the arrangement of the rocks is trough-like. But look at this last case a little more closely ; the surface is nowhere formed for any space by a plane of bedding, the arch is more or less truncated and defaced, and in order to see it as it was when the rocks were first folded, we should have to put back the portions shown by dotted lines, which have evidently been carried away. It is easy to see for instance that portions only of the bed marked A remain on the mountain-tops and in the valley ; if it had originally the same thickness throughout, it must once have reached up to the dotted continuation of its upper surface, arid the parts between that line and the present surface are gone. The group of beds marked B also so exactly correspond on opposite sides of the chain, that we feel sure the portions now so widely disconnected must have once formed parts of an unbroken sheet of strata, that this has been bent in the direction shown up in the air by the dotted lines, and the portion between the two present outcrops has been removed. The reader will see by-and-by that it does not necessarily follow from this that the mountains have ever been as high as the restoration of the missing parts of the beds would make them ; all that is asserted is, that the portions between the dotted lines and the surface have been removed. Surface formed by Denudation. 573 The inequalities of the surface of the ground then are not due, or are due only in a minor degree, to the folds into which the rocks beneath it have been thrown ; some words used a little way back point unmistakably to the cause to which they are mainly due. Completing the curves in fig. 193, and restoring the arch to its original shape, .,; we find that parts of it have been car- \{ ried away. Again, why does the hill on \\ \ltbES'E the left stand up so conspicuously ? The \ rock at the summit and those on its flanks did not originally terminate where they do now, but stretched right and left, as shown by the dotted lines. These dotted parts have been carried away. The reader has doubtless before this said to himself, " Yes ; and what carried them away can have been nothing else but denudation;" and he will be right, for, as far as we know, there is nothing else that can have done it. The conclusion then we come to is, that in most cases valleys have been carved out by denudation, and hills are what denuda- tion has spared; and that even in those cases where hills and valleys may have originated in a bending up or bending down of the rocks beneath them, their outline is still very largely due to denuda- tion. It will be enough to refer the reader to two of the countless instances in which not a doubt can exist that striking hills are merely remnants that have escaped denudation. No better proofs of the truth of this assertion can be found than the mountains on the west coast of Suther- land and Ross, figured in " Siluria," p. 170, and so eloquently described by Hugh Miller (The Old Red Sandstone, p. 56) ; and the Scur of Eigg, described by Pro- fessor A. Geikie (Quart. Journ. Geol. Soc. xxvii. 303). The truth that the present inequalities of the surface are mainly due to denuda- tion was first clearly seized upon by Hutton. His conclusions are thus ele- gantly summed up by Playfair. " It is where rivers issue through 574 Geology. narrow defiles among mountains that the identity of the strata on both sides is most easily recognised, and remarked at the same time with the greatest wonder. On observing the Potomac, where it pene- trates the ridge of the Alleghany Mountains, or the Irtish, as it issues from the denies of Altai, there is no man, however little addicted to geological speculations, who does not immediately acknowledge that the mountain was once continued quite across the space in which the river now flows; and if he ventures to reason concerning the cause of so wonderful a change, he ascribes it to some great convulsion of Nature, which has torn the mountain asunder and opened a passage for the waters. It is only the philosopher, who has deeply meditated on the effects which action long continued is able to produce, and on the simplicity of the means which Nature employs in all her operations, who sees in this nothing but the gradual working of a stream, which once flowed over the top of the ridge which it now so deeply intersects, and has cut its course through the rock in the same way, and almost with the same instrument, by which the lapidary divides a block of marble or granite."* Amount of Denudation. It is desirable at the outset that we should clearly realize how enormous has been the amount of the matter carried away to form the present surface of the ground. For this end the reader cannot do better than turn to Professor Ramsay's paper, " On the Denudation of South Wales and the adjacent Counties of England " (Memoirs of the Geological Survey of Great Britain, vol. i. p. 297). To illustrate the methods employed to calculate what is the quantity of rock that has been removed, one of the sections of that paper is reproduced, with trifling modifications, in fig. 194. The part drawn with strong lines represents the rocks below the surface ; it is constructed by first obtaining an accurate profile of the ground by levelling ; the different beds that come out to-day along the line are then examined and their dips measured, and they are then drawn in, each with its proper dip. In this way, starting on the south, we pass over four groups of rocks, called respectively Coal Measures, Carboni- ferous Limestone, Old Red Sandstone, and Silurian, which come out one from below another in the order in which they have been named, with a steady rise to the north. About A the southerly dip begins to decrease, a. little farther to the north we reach a point where the beds are observed to lie flat, and after passing this point a dip to the north sets in and gradually increases in amount. This shows us that A lies on the crest of an arch, or anticlinal, into which the rocks have been bent. If we continue the same kind of observations, we find that this arch is succeeded by a trough, and the trough again by a second arch, on the northern flank of which the dip is steadily to the north up to the end of the section. In the Silurian rocks there is a well-marked and easily-recognised bed of Limestone marked by a black band. This bassets at E and C, but the bed cannot have originally ended at these points as it does now. Before the strata were folded into their present form, it must have spread out as an unbroken sheet through the body of the Silurian * Works, i. 116. S.urface formed by Denudation. 575 rocks ; and if we carry on its under and upper boundaries, bending them so that their dip may be always the same as that observed in the rocks beneath, we shall see how much of the Limestone, and of the beds under it, has been swept away in the course of the formation of the present surface. In the same way we determine the original con- nection of the bassets of this bed at C and D. Again the Old Red 576 Geology. Sandstone appears at both sides of the arch, but it cannot, any more than the Limestone bed, have ended originally as it does now. Let us carry on its upper and under boundaries, being careful to keep them everywhere parallel to the curves drawn for the Limestone bed, and we shall obtain the outline of the belt which once connected the de- tached outcrops on the southern and northern flanks of the arch. The lines showing the former connection of the rocks, obtained in the manner just described, are dotted, and the tinted space between them and the surface shows how much has been swept off by denuda- tion. The scale of the section is the same for heights and distances, so that everything is in its true proportion, and a glance will show how insignificant is the portion of the rocks that now remains, com- pared with that which has disappeared. The tinted portion in the figure is more than 2 square miles in area, but if we take this as the measure of what has been carried away, we shall rather exaggerate the amount, for the arch would open in gaping cracks at the summit. We may say however that, for every mile in the length of the anticlinal, not far from 2 cubic miles of material must have been swept off to give us the present surface enough to cover the whole of Great Britain to a depth of nearly a foot. And this is not all, for probably not only the Old Ked Sandstone, but the Carboniferous Limestone and the Coal Measures as well, were once continuous over the area, and they are wholly gone along the greater part of the line. SECTION II. THE SHARE OF EACH DENUDING AGENT IN PRODUCING THE SHAPE OF THE SURFACE. *> The surface of the earth then has been carved into its present shape. and denudation is the instrument that did the work. We have already seen that a number of different agents take part in the process of denudation, and we must now inquire how the task has been portioned out among them. In a former chapter on denudation we dealt mostly with the character of the waste resulting from its action ; we have here to look to the kind of surfaces that each of its cutting tools gives rise to. Share of the Sea. We will begin with the sea. A very little reflection will convince us that even at moderate depths the sea can do very little denuding work of any kind. We have seen that running water by itself is not able to cut or wear the rocks it flows over ; but that, if the current is strong enough to carry in suspension or roll along coarse sediment, a large amount of erosion is produced by the aid of the latter. Now the circulation of the depths of the ocean is carried on by currents in all probability of very moderate velocity ; and the water, if it hold anything in suspension, must be charged with fine mud or ooze instead of the rough sediment which enables rivers to exert so powerful a cutting action. That this is the nature of a very large proportion of deep sea-bottoms has been abundantly shown by soundings, and even in the cases where the sea-bed is strewn with coarser detritus a large fragment is of rare occurrence. Professor Wyville Thomson gives two such cases met with in dredging to the Denudation. 577 north of Scotland. In one haul the largest pebble weighed 421 grains or seven-eighths of an ounce, and may have been about the size of a walnut, and no other was met with anything like so large. In another case 718 fragments were brought up from a depth of 1443 fathoms; one weighed 3 grains, the rest being from one-half to a quarter of a grain in weight.* The deep portions of the sea therefore do not possess the conditions necessary for denudation, and we may conclude that the only change that can happen to a surface buried beneath them will be the gradual filling up of any inequalities that may exist by the deposition of fine sediment. But it is otherwise when we come to the coast-line. There we find abundant implements of destruction furnished by the piles of broken rock and rubbish, which atmospheric disintegration and the under- mining of the waves are always detaching from the cliffs. These the breakers, as they are driven in by violent gales, hurl against the rocks of the shore, and in this way incessant destruction of the latter goes on, the land is slowly worn back and the sea advances steadily inland, t But this takes place only between the limits of high and low tide, and practically marine denudation is confined to this zone. The sea then acts powerfully in working back the coast-line, but it does not exert any appreciable wearing action below the level of the lowest tide ; the result therefore of marine denudation must be to wear down a country submitted to its influence to an even surface coin- ciding approximately with the level of the lowest tides. When it has done this, it can do no more in the way of destruction, and it suddenly changes its part to that of a conservative agent, for its waters protect the plain so formed from the action of other denuding forces. Of course it is not intended to assert that the sea everywhere advances at the same rate ; its progress depends on the hardness and structure of the rocks opposed to it, as we shall see more fully in a subsequent section of this chapter. It is this irregular advance that gives rise to bays and promontories and other, inequalities of the coast-line ; but, given time enough, even the boldest headland will be at last cut back. Isolated pinnacles, stacks, and skerries often hold their own for a long time against marine denudation, and stand up as landmarks to show the space over which it has worked its way, but in the end these are undermined, topple over, and are cleared away. Plain of Marine Denudation. The even surface that would result from the action of marine denudation alone is called a " Plain of marine denudation." But in order to get such a plain we must not have any denuding forces at work besides the sea, for a very short exposure to subaerial denudation would soon destroy the uniform flatness which is its characteristic feature. Such a thing then as an unmodified plain of marine denudation never can have existed ; and if there ever had been * Depths of the Sea, App. C. t For details the reader may turn to Professor A. Geikie's Scenery and Geology of Scotland, chap. iii. ; Lyell's Principles, 10th ed. chaps, xx. and xxi. 2o 578 Geology. such a thing, we cannot expect to find any cases where it still retains perfectly its original character. But by careful attention we can yet detect, even among the wonder- fully-diversified features of the present surface, traces of the horizontal planing of the sea by which that surface began to be formed. If we draw a section on a true scale across a country free from great moun- tain-chains, it will in many cases be something such as is shown in fig. 195. There will be hills and valleys, but it will be found possible to Fig. 195. SECTION SHOWING THE PROBABLE RELATION OF THE PRESENT SURFACE TO AN OLD PLAIN OF MARINE DENUDATION. draw a straight line AB gently inclined seawards, that will touch, or nearly touch, the tops of most of the hills, while none of them will rise above it. If we took a raised map of the country and laid a flat board upon it, the same would be true for the board. Now it is likely that the board represents the flat surface which marine denudation, if it had acted alone and had not been interfered with by other agents, would have produced. A very striking instance where such a plain as we have described can still be very distinctly recognised, was brought before the writer's notice during a short journey on horseback over the wild country in the west of Andalucia. At first view this region seemed to be a gently-undulating expanse, stretching out as far as the eye could reach, over which it looked as if one could ride straight away without check or hindrance. A very short time sufficed to show how different the reality was from the appearance. Steep-sided valleys, sometimes deserving the name of ravines, stretched across the route in quick succession, down which the horses had warily to pick their way and out of which they had laboriously to toil, and for a great part of the way the rate of progress did not practically exceed a foot-pace. The conviction was forcibly brought home to the mind that the history of the formation of the surface was something like this. The country had been first smoothed away by some horizontally planing force to an even surface, and afterwards the valleys had been cut down below its level by a trenching process that acted vertically. After what has been said the reader will recognise the sea as the first of these agents, and he will shortly see that the excavators of the valleys have been rivers that ran in them. Share of Subaerial Denuding Agents. Rivers. We may next turn our attention to subaerial denuding agents, and first among these we will take rivers. The coarse sediment that is swept along the bottom wears away the bed, and therefore rivers, as long as they have sufficient fall, are constantly deepening their channels. The banks are also undermined, and from time to time portions, which have been thus deprived of support, break off and fall into the stream, Denudation. 5 79 and the channel thus becomes widened. But its sides, from the way in which they are formed, will always tend to be steep ; their inclina- tion will depend, just as in the case of a railway cutting, on the angle at which the material of the banks will stand ; but it will always be considerable, unless some other denuding agent comes in to modify the results which would be produced by river-action alone. Rivers therefore are denuding tools which tend to cut steep-sided trenches across a country ; and these trenches they are continually deep- ening as long as they have sufficient fall. It will be at once objected to this generalization, that this is not the character of the river- valleys we are most of us acquainted with ; but the reason for this is, that we have very few of us seen a valley that is due to river-action alone. In the formation of most river- valleys other denuding agents besides the stream that flows in them have had a share, and the shape of the valley is the result of the joint action of all. But a case will be given immediately in which the river has not been interfered with, and here we shall see that the result has been exactly such as we described. We will first point out how most river-valleys lose the trench-like form with which they must have started. Rain and the action of the weather round off the edges and break down the sides of the trench, and thus the steep-sided gorge gradually opens out into a broad val- ley, and the widening goes on as long as the slopes are steep enough to allow the disintegrated matters to be washed down into the stream. One test of the correctness of this explanation readily suggests itself. If it is true, the width of the valley ought to depend on the ease with which the rocks on its flanks yield to atmospheric wear. This is found to be the case. Many river-valleys show along their course alternations of broad flats and narrow steep-sided gorges ; in such cases it is always found that, where the valley is broad and open, the river is running across easily-denuded strata; but that wherever a ravine occurs, its banks are formed of unyielding rocks. An instance of this is shown in fig. 196 ; the portion of the valley in the foreground slopes gently up from the river-banks, but when the river crosses the range of hills in the distance the valley contracts into a ravine. On the right-hand side is a section, such as would be given by a very deep railway cutting, which lays open the geological structure of the country; and this shows us that the part where the gorge occurs is formed of hard, thickly-bedded Limestone, while the more undulating portion is underlaid by soft Shale. Examples of this kind are common enough round the border of the Carboniferous Limestone of Derbyshire ; as for instance where the river Derwent enters the Limestone tract about a mile above Matlock. The gorge here was originally only just broad enough to admit the river ; and when the highroad was carried along the valley, the gap had to be widened by blasting away its rocky wall. We also frequently meet with river-valleys whose section is like that in fig. 197, broad with gentle slopes in the upper part, and a deep steep-sided trench, in which the river flows, in the middle. In such a case we find on examination that the upper beds are soft, and have been largely worked back by atmospheric causes ; but as 5 8o Geology. soon as the river had cut down to the more indestructible rock at the bottom, the trench which it ate out retained more nearly its original shape. Canon of Colorado an Example of River-Action. But Denudation. 581 these are matters that will have to be considered more fully further on \ let us now see if we can anywhere hear of a river- valley where Soft Rock. Fig. 197. SECTION ACROSS A RIVER-VALLEY, WITH GENTLE SLOPES WHERE THE BANKS ARE FORMED OF SOFT ROCK, NARROW AND STEEP-SIDED WHERE THE STREAM HAS CUT DOWN TO HARD PtOCK. the stream has been let alone to do its work without the interference of other denuding agents, and learn what the result of that work has been. No better instance can be given than the well-known one of the Colorado River of the West, which empties itself into the Gulf of California. This stream flows for nearly 300 miles of its course in a profound chasm, sometimes not more than 50 yards wide, the walls of which are approximately vertical and vary in height from 3000 to 6000 feet ; that is to say, the gorge is in places more than a mile deep. No one can deny the trench-like character of such a channel ; but are we sure that it has been cut by the river ? The first explanation to suggest itself is, that this mighty chasm is a rent torn open by an earthquake or some similar convulsion, which the river has appro- priated to its use. A little examination shows that this has certainly not been the way in which it was formed. The beds on opposite sides correspond perfectly ; and the rock at the bottom, though deeply eaten into, is nowhere fissured and broken. But what completely settles the question is the fact that the country on both sides is chan- neled in every direction by innumerable narrow, steep-sided, winding chasms, which differ from that of the river only in size. These minor chasms all spring from the main gorge, and divide and subdivide as they recede from it. Their arrangement is so exactly like the branch- ing network of a river and its tributary brooks, that there cannot be a shadow of a doubt that it was from such a system that they took their rise. Each has been made by a stream eating its way lower and lower down, till this extraordinary assemblage of ravines, which are known by their Spanish name of canons, has been produced. This explanation is further confirmed, when we find at various points along the canon patches of River Gravel lodged far above the level of the highest floods, and great sheets of similar Gravel, spread- ing over the flat bottom of the valley where it opens out below the gorge, the pebbles of which are formed of the rocks at the top of the wall of the canon. There can be no doubt then that the Great Caiion, and the innu- merable ravines that spread out from it, have been formed by streams that run or once ran in them ; and what is more, the other condition we were in search of is also satisfied here, no other denuding force has had a share in their formation. For the district is practically rain- less ; and this is the reason why the caiions are so markedly trench- 582 Geology. like there has been no atmospheric wear to round off their edges arid work back their walls.* We have found then here exactly what we wanted, a case of river- action pure and simple ; and we learn from it that rivers are denuding tools that act vertically, and that the channels they cut, when they are left to themselves, are steep-sided trenches. The line of reasoning we have been pursuing will be perhaps made somewhat clearer if we consider a little more in detail the formation of the Great Canon and its tributaries. The country traversed by it is an elevated plateau, varying in height from 5000 to 8000 feet above the sea ; from this tableland a wall of mountains rises on either side, the Sierra Nevada on the west, and the Rocky Mountains on the east. The two bounding ranges run together both on the north and south, and in this way the tract becomes hemmed in on all sides by lofty mountains, and a great basin is formed. The western barrier is breached by three great openings, through which the drainage is discharged by three great rivers, of which the Colorado is one. The district is all but rainless ; and as all the water that is brought into it by the rivers flows along the bottom of profound canons, the surface is dry and parched and in great part desert land. But this was not always the character of the region. The system of canons shows that it was once traversed by a network of streams which flowed on the surface, and there are other reasons for believing that it was formerly well watered and fertile. It is not unlikely that originally the basin was occupied by a lake or lakes, the waters of which were dammed back by the western mountain barrier, and whose overflow escaped through shallow depressions at the same spots where that range is now cut through by the openings already mentioned. At that time most likely the country stood much lower than now ; but after a while elevation set in, and as the land rose, the notches in the mountain-range, through which the water ran out, were worn deeper and deeper, and the level of the lakes was lowered till the basin was at last laid dry. Thus was formed a tract of land, the drainage of which passed out through the gorges which had in their infancy given exit to the water of the lakes. Elevation still went on, and in consequence the gorges, and the river-channels that emptied through them, were continually being cut deeper and deeper till their present enormous depth was attained. One thing more was wanted to give the country its present peculiar character ; rain would inevitably wash in the sides of the chasms and convert them from canons into broad valleys. This result was prevented by a decrease in the rainfall, which may have been brought about thus. Before the region reached its present height, though it was surrounded by a belt of hills, these were of too moderate an elevation to intercept the clouds that passed over them, and moist winds therefore were able to reach the interior ; but when the encircling hills became converted into lofty mountains, the wind, from whatever quarter it blew, was robbed by them of all its moisture before it reached the central plateau, * The fact that the strata are horizontal has also conduced to the preservation of the trench-like shape of the canons. There is not the tendency to landslip- ping that would have existed if the strata had been inclined. See p. 583. Denudation. 583 and the latter became in consequence a rainless area. The gradual elevation of the land then had a twofold result : the rivers were enabled to go on deepening their channels, and rain was kept away. The canons owe their formation to the first of these results, and their preservation to the second.* Other Subaerial Denuding Forces. The whole army of subaerial denuding agents assists in the work of widening the trench- like excavations to which rivers give rise, and in destroying in count- less ways the uniformity of the plain of marine denudation. A description of the mode of action of each has been given in Chapter III., from which the reader will be able to gather how each contributes its share to the general result. The peculiar surface features due to moving sheets of ice will be treated of in a separate section. Landslips. Among the many ways in which subaerial denuding forces bring about the widening of valleys, one of the most important is by the formation of landslips. When the top of a hill or the summit of a steep ridge is capped by hard, massive, heavy rock, beneath which lie softer and more yielding beds, the weight of the rock atop tends to crush down and drive out- wards along the hill-face the soft strata below. In this way portions of the capping are deprived of support, break off, and slide f or topple over down the slope. Very frequently this goes on till the whole hillside from top to bottom is strewn with slipped masses piled one on the top of another in wild confusion. In every case where the above conditions are present, there will be a tendency to the formation of landslips, and sundry other circumstances will increase this tendency and render their formation the more easy. Firstly, the breaking off of the upper rock will take place the more readily, if it be traversed by large open joints. Again, if the dip of the beds be from the hill- side into the valley, that is towards the side on which there is no support, the surfaces of the beds form inclined planes down which detached portions of rock tend to slide ; on the other hand, if the dip be into the hill, there will be no tendency to slide, and landslips can be formed only by the crushing out of soft underlying strata. Further, if the cap be an open porous rock and the beds below impervious, we have one of the most important aids in the formation of landslips. The water which sinks into the upper bed descends till it reaches the impervious stratum below ; being there unable to penetrate lower, it runs off along the plane of junction, and the moistening of the upper surface of the bottom stratum either makes the inclined floor, on which the mass above rests, slippery, or in some other way renders motion easier than it would be if the surface were dry. * For details about this extraordinary region, see the Report of the Exploring Expedition of the American Government (Washington, 1861). There is a good abstract in Nature, i. 434, and a description of the plateau in Sir Wentworth Dilke's Greater Britain. See also Professor Hayden's Sun Pictures of the Rocky Mountains, chap. vii. t It is perhaps hardly correct to use this word, for there would be too much friction to allow of pure sliding. By what means exactly the loosened portions are enabled to move is not very certain, but move they do. 5 84 Geology, The section on fig. 198 illustrates an actual case where all the con- ditions tending to the formation of landslips are found together. The Fig. 198. SECTION TO ILLUSTRATE THE FORMATION OF LANDSLIPS. 1. Massive, jointed, pervious Sandstone. 2. Soft, impervious Shale. 3. Landslips. left bank of the valley is crowned by a thick bed of Sandstone, which is massive and heavy, traversed by large open joints, and pervious to water ; beneath it are beds of Shale nearly impervious, and so much softer than the Sandstone that water can easily reduce them to a state of mud ; the dip also is down into the valley. This combination has produced the result that might be expected, and the whole flank is covered with large landslips. On the opposite side, any tendency there may be to the formation of landslips is counteracted by the dip of the beds into the hill, and not a single slip has occurred. One or two additional cases of well-known landslips may be noticed here. Enormous slips occur round the basaltic plateau of the north- east of Ireland. The cap of this tableland is a sheet of massive Basalt, 700 or 800 feet in thickness ; beneath this comes Chalk, which rests on Marl or Shale. These lower beds are softened by percolating water, and crushed out by the weight above. Landslips on a large scale take place on the Dorsetshire coast. The section of the cliffs is 4. Chalk. 3. Sandstone with Chert. (100 to 150 feet.) 2. Loose incoherent Sand, called Fox Mould. (150 to 200 feet.) 1. Lias Clay (impervious). The three upper beds are pervious, but the water is stopped at the Lias Clay. The loose Fox Mould is under no circumstances very well able to support the weight of the beds above, and when it becomes soaked with water it is still further weakened, and portions of it washed out along the face of the cliffs ; the dip also is seawards. There is everything therefore favourable to the production of landslips, and they occur on an enormous scale.* The picturesque undercliff of the Isle of Wight owes its wild and rugged outlines to the piling one upon another of landslips, which have from time to time broken off from the cliffs arid hillsides above. The section of the solid hill-face shows (3. ( 2. Sandstone, Sand, and Sandy Clay. (150 feet.) Impervious. 1. Gault Clay. The beds dip seawards, and the surface of the Gault Clay is rendered so unctuous and slippery by the water which reaches it through the * See a detailed account of a very large slip at Axmouth, by Messrs. Conybeare and Buckland (Murray, 1840), and Lyell's Principles, 10th ed. i. 536. , 3. Chalk. Pervious. Denudation. 585 overlying strata, that sliding readily goes on. The cause is so obvious that the Gault goes locally by the name of the " Blue Slipper." The masses detached by landslips are more or less shattered, and hence fall a prey to atmospheric destruction more readily than when they formed part of a solid rock thus landslipping becomes a very efficient aid in widening those valleys along whose flanks it goes on. Basin-shaped Lie of Outliers. One more fact in connection with landslips calls for notice. It will be found to be very generally the case that where a hill-top is capped by an outlier of rock, the dip is on all sides into the hill. The reason of this is that the inward dip hinders the formation of landslips, and so contributes to the preserva- tion of the outlier. An outlier whose beds dipped from the centre outwards would, if other conditions were favourable, shed off landslips all round, and would thus be soon carried away altogether. The above rule is so very general that the mere occurrence of an outlier on a hill-top affords strong presumptive evidence that the beds of the hill lie in a basin.* History of the Idea of Subaerial Denudation. The theory that all the lesser inequalities of the earth's surface are due to subaerial denudation is now very generally adopted in this country, and is gaining ground among Continental geologists. But though this view is by no means new, it is only of late years that it has met with any- thing like general approbation. Men for long refused to believe that results apparently so great could follow from causes seemingly so insignificant, let them act as long as you will ; or rather, they preferred to save themselves the trouble of investigating the nature and capabili- ties of these forces by attributing the formation of valleys to causes the existence of which was purely imaginary, or to agents which a little inquiry would have shown were totally inadequate to the task. The earlier speculators supposed valleys to be rents and fissures torn open by convulsions, the like of which had never come within man's experience ; and in spite of its manifest contradiction to observed facts, the notion for a long time held, and in some quarters still holds, its ground. Mountain-chains were imagined to have risen with a bound from the sea-bed, and thrown off gigantic waves, which ploughed deep into the ground and scattered its debris far and wide as they rushed madly over the country. A step was gained when these wild dreams were abandoned and it was realized that the inequalities of the surface had been carved out by denudation. But even then only one-half of the truth was seen, for geologists for a long time persisted in attribut- ing the whole of the work to the sea. A very slight amount of obser- vation would have taught that marine denudation tends to efface rather than produce surface inequalities ; but the supporters of this view were quite content with a vague idea that the sea had done it, and did not trouble themselves to explain exactly how. Thus a host of vain imaginings was for a long time preferred to the simple explanation to which a study of nature leads us. The whole truth was first thoroughly seized upon by two of the master minds of the science by Hutton in * Ruskin, Modern Painters, vol. iv. chaps, xiii. xiv.; Topley, Geol. Mag. 586 Geology. 1795, and by Scrope in 1826 ; and the latter, by an appeal to the district of Auvergne, triumphantly refuted the objection that subaerial agents were not competent to perform the task assigned to them. That country has been formerly the scene of volcanic activity on a large scale, and many of the cones are still standing in a fair state of preserva- tion. Now these cones are composed of such friable materials, that sub- mergence beneath the sea would inevitably sweep them away alto- gether. It is therefore quite certain g that the country has never been overflowed by the sea since the eruptions took place, and that any >H changes in its surface configuration, fa which can be proved to have been 33 produced since that date, must be due to subaerial action alone. Such > changes can be proved to have oc- a I curred in numerous instances ; for E | example Mr. Scrope pointed out o o cases where an old valley had been | dammed across or filled up by a g lava stream, and where the barrier had been cut through or the valley excavated afresh. By reasoning of this kind he established beyond demur, in that particular case, the power of subaerial agents to do all that the theory requires of them ; and what they can do in Auvergne, they are just as well able to do else- where. S3 o * * & *J SECTION" III. HOW THE CHAR- ACTER AND LIE OF THE UN- DERLYING ROCKS AFFECT THE I SHAPE OF THE GROUND. We have already had to notice that the relative power of rocks to resist denudation is an important element in determining some of the leading features of the surface. In this section we will treat this part of the subject more fully. Relative Hardness. The character which exercises more influence perhaps than any other in this aspect is relative hardness and softness. Hard rocks are able to hold out against the wearing How Recks affect the Shape of the Ground. 587 action of denudation better than soft rocks. Hence districts formed of hard rocks stand up more or less boldly and ruggedly in high ground. The country occupied by softer rocks is lower, tamer, and more uniform in outline. This is well brought out in fig. 199, which shows the main features of the country along a line from Snowdon to the east coast of England. On the west rises the mountain district of North Wales, formed of old, very much hardened rocks, named Silurian. Then follows a broad, gently-undulating tract of low ground occupied by softer strata, known as the New Red Sandstone, which have been only slightly tilted from a horizontal position. To the east of this plain a boss of lofty ground marks the position of the Derbyshire hills ; these owe their elevation to the fact that they are composed of a hard group of rocks, known as the Carboniferous, which have been brought up from beneath the New Red Sandstone in a broad anticlinal fold. Descend- ing the eastern flank of the Derbyshire plateau, we find its beds dipping beneath the New Red Sandstone, and pass on to a flat identical in character with that formed by the same formation on the west. After a while the New Red Sandstone begins to be covered up by other formations, known as the Oolitic and Cretaceous; and where the harder rocks of these groups come to the surface, the ground rises into long terraced ridges. The section shows three tracts of lofty uneven ground, and two districts of low and flat ground ; and in each case elevation and ruggedness go along with hardness in the underlying rocks, and a low level and evenness of outline with a substratum of soft rook. The view in fig. 196 also illustrates this general truth; the softer rocks in the foreground give rise to a low undulating tract, while the hard Limestone stands up in a line of bold hills. Another very striking instance of the way in which hard rocks give rise to projecting eminences is shown in fig. 200, which is a view of two hills called Park and Chrome Hills, near Longnor on the borders of Derbyshire and Staffordshire. Here the main mass of the Carboni- ferous Limestone rises at a steep angle from beneath a body of very much softer Shale, and forms a tableland, the face of which overlooks like a wall the flat country occupied by the Shale ; in fact what we usually get along the line of junction is just such a view as is shown in fig. 196. In the case now before us however this very simple type of landscape is diversified by the presence in the middle of the Shale flat of the two conspicuous peaks shown in the sketch. A portion of the Limestone wall is seen in the background, and well in advance of it the hills stand up like outworks in front of a rampart. I can recollect being very much puzzled, when I first saw these hills from a distance, to account for their isolated position, but a closer examination made all clear. Each consists of a mass of Limestone, roughly triangular in plan, which has been brought up by faults in the middle of the Shale. The soft rock has been washed away by subaerial wear all round, and two pyramid-shaped eminences of Limestone have been left standing up. At the bottom of the figure there is a section across both hills ; the surface, when subaerial denudation began its work, may have been 588 Geology. somewhere about the dotted line ab all between that line and the present surface has been removed, and it is easy to see that the occur- rence of the two isolated hills is due to the fact that the soft Shale has been carried away to a much larger extent than the hard Limestone. Other Qualities which enable Rocks to resist Denu- dation. But it is not always the hardest rocks that best resist How Rocks affect the Shape of tlie Ground. 5 89 denudation. Chalk for instance is by no means a hard rock, but it stands up boldly in conspicuous hills above Clays almost, if not quite, as hard as itself, in a way which shows that it has something about it that enables it to hold its own against the wear and tear of atmos- pheric agents better than the Clays. Probably the property which produces this result is the extreme porousness of the rock all the water that falls upon it is at once sucked in, and there is scarcely any flow over the surface to produce erosion ; the Clay on the other hand, which has suffered so much more largely, admits no water, and hence a large portion of the rain which its surface receives is available for denudation. There is a fact pointed out to me by Mr. C. E. Homer- sham which bears out this view. When we pass off the Chalk on to the adjoining district of London Clay, we find that the bridges become all at once larger, and that where a road crosses a flat liable to floods the flood-arches are more numerous and wider. The contrast is very striking, arid proves how much larger the surface-flow of water is over the London Clay than over the Chalk. Another illustration of the principle we are now considering is fur- nished by the section on fig. 199. It will be noticed that there is a slight depression in the middle of the Derbyshire hills ; the boss is higher at the edges than in the centre. But the rock which comes to the surface over this sunken space is by far the hardest of the group that makes up the high ground. The probable reason why it has not the superior elevation to which its hardness would seem to entitle it, is that it is a Limestone ; it is therefore dissolved away chemically as well as worn away mechanically, while the beds above it are Sand- stones, whose destruction is mainly effected by mechanical means alone. Examples like this teach us that the rock which makes the boldest feature is not necessarily the hardest, but the one which can best resist denudation, to whatever quality that power is due. Difference between Results of Marine and Subaerial Denudation. So far we have dealt mainly with denudation in general in this section ; but it is instructive to note how the results which depend on the qualities we have been considering vary accord- ing as the agent employed is the sea or subaerial forces. The effects of marine denudation are seen in the shape of the coast. It cuts hori- zontally, and those rocks which are best able to resist it show their power by running out into promontories and headlands, while the more yielding strata are worked back, and give rise to bays and inlets. Subaerial denudation acts on inland districts, and cuts vertically ; and by it the easily-denuded rocks are worn down into plains and valleys, while the strata which give way less readily stand up in hills and ridges. Both cases are seen side by side in the Isle of Wight. If the reader turns to a geological map of the island, he will see a belt of Chalk, a rock which we have seen resists denudation, running across the middle of it from east to west ; the rocks overlying the Chalk to the north, and those underlying it to the south of the belt, are both of a more yielding character. The Chalk shows its superior powers of resisting 590 Geology. the horizontal planing of marine denudation, by jutting out farther to sea than the beds above and below it into the bold headlands of Culver Cliff and the Needles. Inland it gives proof of its ability to hold out against the vertical action of subaerial denudation, by standing up higher than the beds on either side in a bold ridge that stretches athwart the island between these two projecting points. Effect of Natural Planes of Division. Joints and other How Rocks affect the Shape of the Ground. 591 natural planes of division exercise an important influence on the shape of the surface. They admit water and determine the lines along which it acts with greatest efficiency ; and when their fluid contents are expanded by frost, it is along them that portions break off. Hence, in well-jointed rocks, valleys will tend to become gorges and hillsides to become precipitous. This is well illustrated by the view in fig. 201, which represents a valley in the Millstone Grit district of Derbyshire. The bold line of mural precipices which crown the flanks on either side are composed of a hard Gritstone, while the gentler slopes below are occupied by softer alternations of Shale and Sandstone. This difference in the character of the rocks would alone lead to consider- able difference between the inclination of the upper and lower parts of the sides of the valley ; but the cliffs at the top owe their marked steep- ness and buttressed faces to the fact that the capping rock is traversed by two sets of long regular joints, nearly at right angles to one another. As the Grit is undermined by the weathering out of the soft underlying beds, portions become deprived of support and break off along these joints, and hence the upper part of the hillside assumes the form of a vertical cliff, the face of which is from time to time renewed and kept always sharp and clean. That this is the method of the formation of the feature is clearly seen when we examine the sides of the valley. At the foot of the present cliff we find a talus of blocks which have evidently been detached very recently ; but these proofs of the work of destruction are not confined to this part of the hillside ; the slopes all the way down are thickly strewn with huge masses of Grit, perfectly angular, and except that they show a little more signs of weathering, in every way similar to the freshly-fallen blocks at the top. These, there can be no doubt, are the ruins of old escarpments, and they indicate successive positions of the cliff while it was being worked back to its present line. These loose blocks are so numerous that they furnish ample materials for walls and buildings, and do away in great measure with the labour of quarrying. They are distinguished from the stone raised in quarries by the name of " Day-stones." It is denudation, guided by natural planes of division, that has given rise to the isolated pinnacles of rock that occur so frequently both inland and on the sea-coast. We have chosen two instances, one formed by subaerial, the other by marine denudation. The first, shown in fig. 202, is a tall spire of Limestone standing in one of the Derby- shire dales. Fig. 203 shows the way in which it was formed : the rock is traversed by two sets of joints ; carbonated water passing through these dissolves the Limestone and widens the fissures. By the enlarge- ment of one set of joints a number of buttress-shaped projections jutting out from the hillside are produced. A similar operation acting along the other set of joints cuts up these buttresses into pinnacles. Some of the buttresses shown are in the first stage, the most distant has already begun to be subdivided into pillars. . Our other instance is the Needles of the Isle of Wight, fig. 204. The Chalk of which these are composed has been tilted till its beds are nearly vertical, and it is also traversed by points perpendicular to the bedding. The waves, aided by subaerial agents from above, have worked their 59 2 Geology, way along these two planes of division, and severed several blocks of Figs. 202 and 203. DETACHED PINNACLE AND BUTTRESSES OF LIMESTONE, DERBYSHIRE. the rock from the main mass in the cliff. The faces on which the light falls are formed by planes of bedding, the slight deviation of How Rocks affect the Shape of the Ground. 593 which from the vertical gives the Needles their overhanging position ; the faces in shadow are joints. Effect of the Lie of the Beds on the Shape of the Surface. It is easy to see that the inclination of the beds is an 2p 594 Geology. important element in determining the shape of the ground. Consider two areas of equal extent, over one of which the beds lie flat, while in the other they are inclined to the horizon. In the case of the first, when marine denudation ceased to act, the surface was formed every- where of the same rock ; it would therefore be lowered everywhere at the same rate by subaerial denudation, and the result would be a same- ness of feature and a tendency to the formation of a plain or table- land. If the rock composing this flat is hard, the valleys cutting across it will keep a narrow, steep-sided cross-section, will be trenches in fact that from a broad point of view will not interfere with its general plateau-like character ; but if the underlying rock be soft, the river-trenches will be gradually opened out into broad valleys, and this widening may go on till the plateau-like character of the ground becomes entirely destroyed, and its former existence can only be in- ferred by noting that the ridges separating the valleys are all very nearly of the same height. If we now turn to the other area, where the beds are tilted, we see that the surface is formed of a succession of rocks, differing in hard- ness and other qualities ; it will therefore be lowered unequally by subaerial denudation, and the result will be variety of feature and a tendency to the formation of hills and valleys. Steps in the Formation of the Surface. Let us now, with these general principles to guide us, try if we can picture to our- selves the steps by which a mass of rock formed beneath the sea is converted into a land surface diversified by hill and valley. First came the upheaval, and this we have seen was effected by a bending of the originally horizontal strata into a series of arches and troughs, the former of which gave rise on emergence to tracts of dry land. But the result was not accomplished without a struggle ; whenever one of these broad-backed masses reached the surface of the water, it came within the range of marine denudation. The waves attacked it and pared it away, and as it was slowly lifted up, slice after slice was planed off the top. Thus a constant battle went on between the two opposing forces, the one striving to raise the submerged mass beyond the reach of the waves, the other wearing it away down to the sea-level as fast as it got its head above water. But at length the up-arching movement gained the mastery, and a tract of dry land was established. Since this land had been formed by the planing away, one after the other, of horizontal slices from the back of the arch, its surface must have been nearly level ; but since it owed its existence to a bending up of the beds, it would probably be slightly higher in the middle than at the margins ; and the slope either way would be in the same direction as the dip of the beds. Hence the surface of the new-born land would consist of two inclined planes, meeting along the crest of the arch, and sloping thence gently down to the sea-level ; and the inclination of the surface on either side would be in the same direction as the dip of the beds. This was the first step of the process, and when it was completed, a section across the country would be such as is shown in fig. 205. The tract of dry land thus established is placed beyond the reach of Hoiv Rocks affect the Shape of the Ground. 595 the sea, only to be subjected to the action of subaerial denudation, and we must next inquire how it will be modified by atmospheric wear and Fig. 205. SECTION ACROSS A COUNTRY AFTER THE FORMATION OF A PLAIN OF MARINE DENUDATION. art. Sea-level. tear. First, rain streams over it, and seizing on any little inequalities becomes collected into channels. Since these channels must follow the slopes and the ground slopes either way with the dip of the beds, it is easy to see that the earliest watercourses will run in the same direction as the general dip of the country. Moreover, because these channels are formed by river-action, they will tend to be trench-like in shape. Fig. 206. DIAGRAMMATIC PLAN AND SECTION OF A RIVEE-TRENCH, CROSSING STRATA OF UNEQUAL HARDNESS. N. B. The bands representing the different strata in the sides of the trench are so violently foreshortened as to give the idea that the beds are vertical. The reader must please bear in mind that the beds are dipping at a moderate angle from A towards B. The second step then consists in the formation of a series of trench-like river-channels running in the direction of the dip. The valleys thus formed cut across the outcrops of the beds, and are hence called Transverse Valleys. Fig. 206 is a bird's-eye view of a country, showing in a diagram- 596 Geology. matic form one of these trench-like valleys (AB), cutting across the strike of the beds, which come to the surface along the lines cd, CD, ef, EF, and dip towards the spectator. But these first-formed valleys cannot long keep their trench-like shape. Atmospheric action we have seen gradually broadens them, and it has been further pointed out that the process goes on much faster in some rocks than in others, so that along the outcrop of certain beds the valley is widened at a more rapid rate than along that of others. Hence arises that alternation of broad valley and strait gorge which is so constant a feature in valley contour ; and from a continu- ation of the same process still more important modifications result, which we now proceed to notice. In fig. 206 suppose that AB represents a transverse trench, and that among the beds which it cuts across, cd and ef are more easily denuded than CD and EF; the widening of the trench will go on faster in the first pair than in the second ; where it crosses CD and EF the steepness of its sides will be destroyed very slowly, but where its walls are formed of cd and ef, its edges will be more rapidly worn back, and little recesses will be formed in the face of the trench. The continual washing in of the soft strata will deepen and extend these recesses, and they will creep step by step outwards along the outcrop of such beds, assuming in succession positions such as those marked by the dotted lines 1, 2, 3, 4. A very little reflection will show that this becomes in the end equi- valent to the formation of two branch valleys running along the out- crop of the stratum ef, the streams draining which become feeders of the original transverse river. Here then we have arrived at the third step in the process of valley excavation. It consists in the formation of valleys branching out of the first-formed, transverse valleys, and running along the outcrop of the more easily-denuded beds. The valleys thus formed, because they follow the strike, are dis- tinguished as Longitudinal Valleys. We have for distinctness' sake spoken of the three steps, the Forma- tion of the Plain of Marine Denudation, the Excavation of Transverse Valleys, and the Wearing back of Longitudinal Valleys, as having taken place one after the other. In reality the last two, and to some extent all three, go on together. The result however will be evidently the same as if each step had been finished before the next was begun. The result of the process we have been following will be the pro- duction of a country composed of alternations of hard and soft beds dipping at moderate angles, and the general physical features of such a country will be as follows. There will be two sets of valleys. The larger will run across the strike and will slope in the same direction as the general dip of the beds. The streams that feed the rivers of these transverse valleys will run parallel to the strike and along the outcrop of soft beds. Each of these longitudinal valleys will be bounded on either side by a ridge formed of a hard rock. Two points would speci- ally strike the eye on taking a bird's-eye view of such a country or on looking at a raised map of it : the long lines of parallel ridges formed How Rocks affect, the Shape of the Ground. '597 by the outcrops of hard beds, and the way in which the great trans- verse valleys cut across these ridges. Mr. Topley has shown that in the case of the Weald of Kent and Sussex flexures in the strata have had some effect in determining the lines of the great transverse valleys and have in other ways influenced the shape of the surface. The main transverse valleys that run with the dip follow the lines of broad secondary synclinals whose axes are transverse to the main axis of upheaval. We can realize how bends of this character would be very liable to produce the slight depressions which are necessary to start a valley. He also quotes cases in which local transverse anticlinals seem to have initiated smaller transverse valleys that run against the dip and empty into longitudinal valleys. He also finds that passes are apt to occur where escarpments are faulted.* There is nothing in this inconsistent with the views just enunciated. While we feel confident that denudation has been the main factor in bringing about the shape of the ground, it is perfectly conceivable that folding and other displacements may have played their part in the operation by determining the lines along which denudation acted most efficiently, and by modifying and guiding its action in other ways. Escarpment and Dip-slope, Such are the broad general facts. When we come to examine more in detail the shape and char- acter of the features of a country composed of alternations of hard and soft rocks dipping at moderate angles, we find that there is in most cases a marked difference between the slope of the two sides of a ridge formed by the outcrop of a hard bed. On one side it presents a steep face, on the other it falls away in a long gentle incline. The upper part of the steep side is very abrupt, sometimes a vertical face of rock, below this the inclination of the ground is somewhat less steep. The next thing we notice is that the steep faces are all turned one way, and the gentle slopes all the other way; the first all look towards the quarter to which the beds rise, the latter are inclined in the same direction as the dip, and frequently at almost the same angle. Hence the latter go by the name of Dip-slopes, while the upper abrupt portions of the steep sides are distinguished as Escarpments. In the view and section in figs. 207 and 208 we have a very marked instance of the kind of feature just described. There are two ridges running roughly parallel to one another, arid in each the side turned away from us is steep, and the side facing us is a broad flat surface sloping gently down towards the spectator. The section shows how these features are related to the lie and character of the rocks ; it runs from the highest point of the distant ridge across the summit of the nearer ridge on which a group of trees is perched. We see at a glance that the steep fronts look in the direction of the rise of the beds, and that the long gentle slopes fall away in the same direction as their dip and nearly with the same inclination. Further, the rocks 2 and 4 which form the escarpments are hard Sandstones ; and beneath each of these lie more yielding Shales 1 and 3, over the outcrop of which * The Geology of the Weald (Memoirs of the Geological Survey of England and Wales), p. 280. 598 Geology. the slope becomes more gentle. The dip-slope of the more distant ridge is broad and very conspicuous, that of the nearer ridge, though narrower, is remarkable for the singular evenness of its surface. The bed 4 is a somewhat massive and well-jointed rock, and hence the escarpment formed by it is abrupt and craggy. The facts just described are strictly in accordance with our theory. The valleys in the view are longitudinal, for they follow the outcrop of the belts of soft Shale, and they would be found, if we walked down them, to empty themselves into a transverse stream. Let us look a little more closely at the steps by which this very characteristic outline has been produced. Fig. 209 is a section across such a longitudinal valley as we see in figs. 207 and 208. The darker part shows the present surface of the ground and the rocks beneath it : the dotted line at the top is the plain of marine denudation ; the fainter portion between this dotted How Rocks affect the Shape of the Ground. 599 line and the present surface shows what has been carried away by subaerial denudation in the excavation of the valley. Let 1 and 3 be hard rocks, 2 a soft rock. ACB is a section across the little recess with which the excavation l||||||||ifl^^ Jt of the valley begins. The small stream that /,,';/ HlH gathers in this recess and runs through it will be constantly eating its way downwards and so deepening the hollow : rain and other atmos- pheric action will at the same time be constantly washing down the banks and widening it. The hollow will thus grow steadily both deeper and wider ; it will develop into a valley, and its cross- section will assume in succession positions such as DFE, GKH. The gradual growth of the valley will continue to go on in the manner just described till it has cut down to the top of the hard bed 1, a position indicated on the diagram by GKH. This may be called a critical point in the valley's existence ; the circumstances under which its excavation has been so far carried on are now altered, and a correspond- ing modification in the results of the work may be expected. Two courses, so to speak, are now open to the stream : it may go on cutting deeper, but if it does, it will have to work its way through the hard rock 1 ; or it may attack the bank, KH, of soft rock. The latter is so much the easier that it is evident it will be the one adopted ; the direct deepening of the chan- nel will cease, and the running water will expend its energy in undermining the bank on the right hand. Portions of this will thus from time to time be brought down into the brook, where they will be ground fine and swept away. In this way the bank to the right will be continu- ally worked back, and the valley gradually widened, its floor being always formed by the top of the hard bed 1. The action of the stream alone would produce a steep cliff on the side we are considering, but atmospheric wear will come in to modify this result, and by inces- santly breaking down the face, will always keep the slope moderate. Thus one flank of the valley will be continually shifting to the right, assuming in succession positions such as LM, NO, PQ. When the movement has extended up to the hard bed 3, the upper part of the hill- side will, on account of the superior hardness of this rock, stand at a steeper angle than the ground below, and hence an escarpment will be formed at the outcrop of this stratum. We 6po Geology. must further notice that the manner in which the slope is worked back will be somewhat different in those parts which are composed of hard rock and in those where the softer beds occur. The latter are gradually washed away bit by bit in a fine state of division ; from the former large blocks are detached from time to time, which on account of their superior power of resisting the action of the air, are broken up very slowly, and consequently remain in large numbers strewn over the slope. As the soft rock on which they rest is worn away from beneath them, these fallen masses slide down lower and lower, till the whole hillside becomes thickly covered with them. One case of this kind has been already noticed (p. 591), and it is an occurrence very frequently met with on the slope beneath an escarpment of hard rock. While all this has been going on on the right, atmospheric denuda- tion has not been idle on the left-hand side of the valley. The mass GKR of soft rock is gradually washed into the stream, and its ruins carried away. This easily-destroyed portion the subaerial denuding forces clear off without any difficulty, but they can make only very slight impression on the harder bed below ; hence when the bottom of the valley has been brought down to the top of this rock, any further lowering goes on very slowly, and it remains very nearly at this level. And so in the end we have left of the rock-mass we started with only that part which is distinguished by a darker tint on the figure. The valley has now assumed a form exactly corresponding to that depicted on figs. 207 and 208 ; it has a long dip-slope nearly coinciding with the top of the hard bed 1, and a steep face on the other side; the slope of the latter is comparatively gentle where it is formed of the soft stratum 2, and rises in an abrupt scarp where it is capped by the hard rock 3. The reader must not suppose that the explanation just given is pure theory. If he will go into any district where rocks alternately hard and soft rise to the surface, he will not only find numberless instances of escarpments and dip-slopes, but he will see the process to which their formation has been attributed still in action. He will have no diffi- culty in lighting on cases where a brook runs from some distance exactly along the top of a hard rock with a cliff of the overlying softer beds forming one of its banks. If in such a case he mark the raw newly-cut look of the cliff and the heaps of fallen debris at its foot, he cannot fail to conclude that the stream is undermining its bank, and that by this gradual working back of the lower part of the slope, aided by subaerial wear above, the valley is being widened, while its peculiar type is all along preserved. Only suppose that the stream has been doing for a long time back exactly what it is doing now, and we have all the machinery necessary for carving out of a plain of marine denudation just such hills and valleys as the landscape sets before us. In the illustration chosen, the arrangement of the features in escarp- ment and dip-slope is marked with singular distinctness. The reader must not expect to find many instances which conform to the normal type as rigidly as this. The two distinctive features are often masked How Rocks affect the Shape of the Grou to some extent by numerous minor modifications, but they be recognised with more or less of certainty in a country formed of bedded rocks of unequal hardness and inclined at a moderate angle. There is probably no agent so efficient as Ice in obscuring the features produced by the unequal yielding of different kinds of rock to denuda- tion. In some tracts, parts of the Carboniferous districts of the North of England for instance, where there are all the requisites for the formation of escarpments, we find these ridges either conspicuous by their absence, or at best far less strikingly marked than among the corresponding rocks of the centre of England. The explanation probably is, that the more northern region has been swept over by an ice-sheet which planed down all the lesser inequalities of the ground, and there has not been time since the glaciation for subaerial denuda- tion to carve them out afresh. This ice-sheet probably never reached so far south as the centre of our island, and there the results of long ages of uninterrupted subaerial wear are seen in the conspicuous character of its escarpments. The deposits also formed by ice-action frequently prevent our seeing features which actually exist. Some- times large tracts of country are deeply buried in Boulder Clay, and the uneven surface of the stratified rocks is simply smothered ; some- times masses of Boulder Clay are piled up against the steep face of an escarpment, or moundy hills of the same deposit stand on a dip-slope, and in this way the distinguishing characteristics of each feature are destroyed to the eye. If the driftless area of the Carboniferous rocks in South Yorkshire be compared with the corresponding drift-covered area of Lancashire, the contrast between the sharp definition of the features of the one and the indistinctness and faintness of the features of the other is very striking. Valleys determined by Joints. If there be no great in- equality in the rate at which the different rocks yield to denudation, tributary valleys will still be formed, but their position and direction will be determined by some circumstances other than unequal hard- ness. For instance if the rocks are well jointed, master joints will be lines of weakness well calculated to be widened by atmospheric erosion, so as to give a start to a line of valley. Valleys determined by Faults. Faults too in a similar way sometimes determine the lines of valleys. Where hard and soft rocks are brought side by side by a fault, the latter are worn away more largely than the former and a valley results. Such valleys, like the faults which give rise to them, usually run approximately in straight lines. It must be carefully noted that in such cases faults may be said to produce valleys only in so far as they give rise to con- ditions which cause denudation to act unequally : they are only the indirect originators. No case is known where the fissure of a fault is a gaping chasm, such as would form a valley without the aid of denudation. Qualifications. The somewhat hard and rigid classification of valleys which has been just given, and the explanation which has been attempted of the way in which each kind arose, are of course true only in a very general way. It is a broad and, so to speak, diagrammatic 602 Geology. description in which the main characteristics only are retained, while many minor details and divergences from the general scheme are left out. Thus there are valleys which are neither transverse nor longi- tudinal, but have had their directions determined by causes, two of -D which have been just men- ') tioned, other than the lie of f~~ the beds. Other valleys again partake of the character of both kinds, running parallel to the strike for parts of their - course and crossing it in other parts. Such valleys can fre- quently be shown to have arisen in a manner which fig. - 210 will explain. The fine Yl parallel lines represent the c outcrops of the different rocks, Fig. 210. PLAN OF A RIVER- VALLEY, and ABC is a stream, which PARTLY LONGITUDINAL AND PARTLY f rom A to B is longitudinal, but at B turns suddenly and assumes the character of a transverse river. In such a case we usually find an upward continuation, ED, of the valley, BC. This is now so small in comparison with BA, that it is looked upon as a tributary; but it is likely that the valley system at first consisted of a transverse gorge only, of which DB formed the upper and BC the lower part, and that BA was originally a longitudinal tributary subsequently formed. The growth of the .portions AB, BC, has however been more rapid than that of BD, and hence the latter has become insignificant and has sunk to the rank of a tributary, and the original feeder. BA, has on account of its faster growth reached a size which causes it to be regarded as a part of the main stream.* Exceptions to our scheme, such as these, doubtless occur ; but if they are allowed for, it will be found to be in the main a true and useful classification. We must indeed bear in mind that in each individual case peculiar conditions give rise to peculiar modifications of the broad character of both valley systems. And when the geological structure is very compli- cated and has been produced by a long succession of upheavals and denudations, it may be difficult, perhaps impossible, fully to see one's way through the long chain of events that have had a share in the production of the surface ; but in very many cases the description just given will be found to apply fairly well to the present arrangement of hill and valley, and traces of the three different processes that have been mentioned as contributing to its formation may still be detected with more or less of certainty. The facts explained are of great value to the field geologist. In a strongly-featured country he can determine merely by noting dip-slopes * This explanation was given by the late Mr. Jukes to account for the erratic behaviour of some rivers in the south of Ireland. Quart. Journ. Geol. Soc. xviii. 378. How Rocks affect the Shape of the Ground. 603 and escarpments the direction and approximately the amount of the dip. Final Results of Subaerial Denudation. Now that we have formed a notion of the way in which hills arid valleys come into existence, we may carry our inquiries a step further, and ask what is the result of the action of subaerial denudation on the features which it creates. As long as rivers have sufficient fall, they continue to cut down their channels ; and though the hill-tops and plateaus are at the same time being gradually lowered, yet the first process goes on so much faster than the second that the inequalities of the surface continue for a long time to be not only preserved but even increased. But if there be no unequal elevation of the land, there will come at last a time when the continual lowering of the beds of rivers has so far decreased their fall that they are no longer able to deepen their valleys ; and when this has come about, each river will meander from source to mouth over a flat raised but little above the sea-level. Atmospheric wear however will still continue to act on the surrounding hills and sweep the waste of them into the rivers, which bear it away, and so in time they too will be cleared off. Thus the final result of atmospheric denudation is to destroy the features which itself gave rise to, and the end of its action is to plane everything down to a uniform level. It may be that some of the level surfaces, which seem to have preceded the present arrangement of hill and valley, have arisen from the long- continued action of subaerial wear and not from marine denudation. The considerations just stated show that, in the case of an uncon- formity, the denuded surface of the lower group may have been pro- duced in three different ways. Denudation may have been arrested, by the rocks being again submerged as soon as the formation of a plain of marine denudation was complete. Or submergence may have been deferred till subaerial wear had cut out hills and valleys. Or lastly, the older rocks may have remained above water long enough to allow them to be worn down by long-continued subaerial wasting to a flat surface. In the first and last cases the floor on which the newer rocks rest will be fairly level ; in the second it will be strikingly uneven, and the upper group will fill up its depressions and level over its inequalities. Cutting back of the Channels of Rivers. We have hitherto dwelt mainly on the action of rivers in lowering their beds, we have now to look at work they do in a somewhat different direction. In many cases, owing to the unequal hardness of the rocks over which they flow, they are enabled to cut back and lengthen the gorges in which they run. The nature of this action, and the way in which it is carried on, will be understood by a reference to the sketch in fig. 211. We see there a brook flowing in a narrow gorge, which is shut in at its upper end by a cliff, over which the water tumbles in a little fall. The upper part of this cliff is formed of a bed of rock which pro- jects well above the strata below ; the same bed is seen jutting out here and there among the foliage in prominent ledges from the sides of the ravine. This rock evidently stands out because it is harder than those 604 Geology. underneath it, and this is specially the case at the waterfall, because the spray is there always playing on the face of the cliff and aiding other Fig. 211. BKOOK CUTTING BACK A RAVINE. subaerial forces to wear away the soft rocks of its lower part. When- ever the ledge at the top has been in this way sufficiently undermined, a slice breaks off along a joint and an even face is produced. Under- mining then begins again, till another fall of the capping rock results. A pile of freshly-fallen fragments at the foot of the cliff shows that the process is always going on. Thus the waterfall, which has in the lapse of time travelled along the whole of the ravine, is still moving in the same direction, and the gorge is being continually eaten farther and farther back. Instances of this phase of river-action may be seen in every mountain brook that flows over alternations of hard and soft rocks. The grandest case known is that of Niagara, so well described by Lyell.* It is evident that by the action just described very important modi- fications in the surface form and drainage system of a country may be brought about. A trifling rivulet streaming down the face of a ridge may deepen, and at the same time eat back its channel, till a deep valley, cutting completely through the range, is produced. Thus a transverse feeder to a longitudinal valley might spring into existence, from which longitudinal branches would extend themselves ; indeed, while it is likely that the great transverse valleys have been carved in the manner already described out of the original plain of marine denu- dation, we may reasonably refer the lesser valleys of the same class to this cutting-back process. It is not difficult to conceive too how, when a ridge has been cut across by the gradual working back of a ravine, a very trifling amount of unequal upheaval might reverse the direction of the drainage, and turn streams into this new channel which had previously discharged themselves by different outlets. The cutting back of river-channels is. not confined to districts where the rocks are of unequal hardness, though it is in such that it goes on most rapidly ; it happens more or less everywhere, and must take place * Principles of Geology, 10th ed. vol. i. p. 358 ; Travels in North America, vol. i. chap. ii. How Rocks affect the Shape of the Ground. 605 largely in rocks, like Limestone, which are chemically soluble. The brooks for instance that flow down the dip-slope of the Chalk range push their heads yearly higher and higher up the incline, and may in the end give rise to valleys cutting quite through the ridge, which may carry off much of the drainage of the flat country beyond. Alluvial Plains. The work of excavating gorges and carving out strongly-marked features is done by rivers during the heyday of their youth and the full vigour of their manhood, that is to say during those portions of their lifetime or over those parts of their course where their fall is considerable ; but there comes a time of old age to every river, when, over the lower part of its course at least, the surface of the ground has been so far levelled by denudation that the river has not fall enough to allow of its any longer deepening its bed. It then flows placidly in a series of gently-winding curves over a broad flat plain from which hills rise more or less abruptly on either side. The formation of this plain is the work of the river during the latter part of its lifetime, and is effected as follows. Stand opposite one of the bends at the spot marked B on the plan in fig. 212, and look across B Fig. 212. FORMATION OF AN ALLUVIAL PLAIN. the river : a spread of rich flat meadow-land shown by dotting lies behind you : at A the bank rises steeply in a raw newly- cut cliff. The stream sweeps rapidly round the convex part of the bend at A, impinges on the foot of the cliff, undermines it, and from time to time large masses are brought down, ground fine, and swept down stream. The bank at A is being constantly worked back and is always moving from A towards the right. At B the flow is sluggish, and the water is shallow because mud, sand, and gravel are dropped there ; the current has brought them so far, but it runs there too feebly to carry them any farther. On this side then the channel is being filled up. The process results in the lengthening of the sweep which the river makes on the 606 Geology. convex side of the bend and the piling up of mud- or sand-banks on the concave side. The heaping up of these banks is the first step towards the formation of the flat which adjoins the river : the process is completed during floods when sediment is thrown down in the hollows between the banks, the inequalities are levelled up, and an even surface is produced. The sediment thus deposited is called Alluvium, and the flat margins of the stream are known as Alluvial Flats. When two bends approach very near one another, the narrow neck of land between them is sometimes cut through by floods, arid the course is thus straightened. But it is impossible that it can continue straight : if there be the least deviation from a right line, excavation must go on on the convex bank and deposition on the concave bank, and a new series of bends will be gradually established.* Thus rivers in their old age, though they have lost the power of deepening their channels, are incessantly swinging from side to side and do important denuding work in the way of widening and laying down the broad alluvial bottoms of their valleys. SECTION IV. RIVERS AND THEIR WAYS. When the doctrine was first mooted that valleys are simply trenches dug out by the rivers which run or once ran through them, it was easy to discover several very plausible objections to it. It was pointed out that the excavation of transverse valleys for instance could only be managed by rivers running uphill, for this they must obviously do if they are to cut through a ridge which stretches directly athwart their path. Such an objection would have been perfectly sound and good provided the surface of the ground had been the same when the river began to run as it is now, but what has been said on the last few pages completely disposes of it, for it has been there shown that when the river was started, the shape of the surface was widely different from what it is at present. In fact by an application of the principles already laid down the reader would have usually no difficulty in ferreting out for himself the chain of events by which the course of any particular river has been determined ; but the eccentricities which rivers apparently show are so curious and interesting that it will be worth while noticing a few instances ; and there are besides some cases to which the explana- tion given will not apply and which must be accounted for in a different manner. Breaching of Hill-ranges by Rivers. One anomaly to be got over is this. We frequently find a country traversed by a number of hill-ranges running parallel to each other and separated by broad valleys. In such a case we might expect to find the principal streams running along the valleys between each two consecutive ridges. Just the very contrary however is the case ; the main rivers cut in a most * On the Oscillation of Rivers see Ferguson, Quart. Journ. Geol. Soc. xix. (1863) 322. Rivers and their Ways. 607 marked way across ridge after ridge, traversing each in a narrow gorge- like valley, and the waters that drain through the valleys between the ridges empty into these trunk streams. To use a common expression, the great rivers run across the "grain " of the country, and the streams that flow with the "grain" are only tributaries ; in other words, the principle drainage is " transverse," the tributary streams are " longi- tudinal." We need not go far from home to find instances. Two such ridges cross in almost unbroken lines the south-east of England : one, formed of a group of hard Limestones distinguished as the Oolitic formation, extends from Gloucestershire to Lincolnshire ; another, composed of Chalk, stretches from Dorsetshire to the coast of Norfolk. A broad plain spreads out to the north-west of the Oolitic range, and another great flat lies between the two ridges. Each of these ridges too presents a steep face to the north-west, and falls away with a long gentle slope in the opposite direction. Nothing would seem more natural than that the two hill-ranges should act as watersheds that the brooks streaming down the south-easterly slope of the Oolite range for instance should be carried off by a river run- ning at the foot of the Chalk escarpment. Nothing of the sort occurs ; a large portion of the main drainage is carried off by rivers which run directly across one or both of these ranges. Thus the Witham, the Welland, the Nen, and the Great Ouse all rise on the plain to the west of the Oolite range, and each in succession cuts across this ridge and discharges into the Wash. The most marked instance of a trans- verse stream however is furnished by the river which is called the Cherwell above Oxford and the Thames below that city. It springs in the plain to the north-west of the Oolitic escarpment, cuts through that escarpment, continues its course over the flat between the Oolitic and Chalk ranges, and then breaches the latter, cutting across it in a direction almost at right angles to its general trend. A word of explanation as to the case is perhaps required. If we adopt the usual nomenclature of geography, we should say that the Thames rises in the Cotswold Hills, flows in a longitudinal east and west valley to Oxford, then turns suddenly to the south and cuts transversely across the Chalk range. The explanation however of cases like this, illus- trated by fig. 210, will apply here. The transverse gorge of the Thames below Oxford is so clearly a continuation of the valley of the Cherwell, that we must look upon the two as constituting together the original transverse trench with which the drainage system began ; and the portion of the Thames valley above Oxford is as clearly a longi- tudinal feeder excavated subsequently in the manner already described. As one more instance, we may mention the Stour, which rises on the low ground to the north-west of the Chalk range and cuts directly across that ridge to enter the sea in Poole Harbour. Another very striking instance is furnished by the Weald of Kent and Sussex. This is an area surrounded on the north, west, and south by a lofty range of Chalk hills, with their steep sides facing inwards. Starting at Folkestone, we trace this girdling ridge along the North Downs to beyond Guildford ; it then bends south, and after- wards turning east, runs along the line of the South Downs to Beachy 608 Geology. Head. Between the last point and Folkestone the coast is low and flat, and there is no barrier separating the interior from the sea. In a district hemmed in in this way on three sides and open to the sea on the fourth, we might expect to find the drainage passing away in the last direction and escaping by what seems its natural outlet. But just the reverse is the case. The streams that enter the sea between Beachy Head and Folkestone are few, short, and insignificant; the principal rivers rise in a central dome of high ground and flow north or south, escaping through narrow valleys that breach the barriers of the North and South Downs. Here again the trunk streams are "transverse," the feeders "longitudinal." Among the streams that in this way breach the North Downs are the Medway, the Mole, and the Wey ; the Arun, the Adur, and the Ouse in the same way set at nought the barrier of the South Downs. The Isle of Wight again furnishes other remarkable illustrations of the disregard of rivers to the present contour of the ground. It is traversed from east to west by a strongly-defined ridge of high ground formed of Chalk, the country both to the north and south being sensibly lower than the ridge. But the ridge is not, as might be expected, the watershed of the island ; by far the larger portion of the drainage is carried off by three rivers the Brading Brook, the Medina River, and the Yar. All of these, notably the last, take their rise near the southern coast, flow steadily northwards, pass through gaps in the Chalk range, and enter the sea on the north side of the island. Instances might be multiplied without limit. Wherever we study the relation of river-valleys to the present physical geography of the country they traverse, we find them, big and little alike, play- ing the same trick, and forcing their way through hill-ranges every way calculated at first sight to bar their progress. The view in fig. 196 will give an idea of the way in which a range of hills is breached by a river-valley. The stream is flowing from tfie spectator ; it meanders over a broad undulating country till it reaches the line of bold hills in the distance, which rise like a wall from the flatter ground in front ; and then, instead of being turned aside by this barrier, it cuts across it, running on in a narrow gorge. Such are the facts which we have to explain, and the explanation resolves itself into finding out how the gorges which conduct rivers through hill-ranges were formed. The rough-and-ready way out of the difficulty, generally accepted in the early days of Geology, was that they had been torn open by con- vulsions. In no instance could this be proved to have been the case, and in most this explanation could be shown to be directly in the teeth of the facts. The strata on opposite sides of the gap exhibit no signs of violent disturbance, and the river may be in many cases observed to flow over a bed of solid, unruptured rock. Indeed one explanation alone is admissible : the gap, like other valleys, has been cut out by a river flowing through it. If this be granted, it is perfectly clear that the river must have begun to run when the surface configuration of the country was alto- gether different from what it is now. For suppose we endeavour to take Rivers and their Ways. 609 water in an open conduit across the country shown in fig. 196, from a reservoir in the foreground ; on reaching the distant ridge the conduit would have to take a turn and be carried along its foot. A river is only an open conduit, and hence any river that began to run when the surface is such as it is now, must turn aside on reaching the ridge in the same way. But the explanation will be perfectly easy if we suppose the birth of the river dates from a time when the present inequalities of the ground had not yet come into existence. If the reader will recall the account given a little way back of the growth of hills and valleys, he will recollect that there was a time when the surface of the country was a plain as high, or somewhat higher, than the top of the ridge. It was then that the river began to flow, cutting, as has been pointed out, a trench across the plain. In the meanwhile atmospheric denudation was at work, wearing down the country on either side, the stream carrying away the waste as fast as it was washed in. But it must be borne in mind that though the deepening of the channel could not go on faster in the soft than in the hard rocks, the country at large was worn away much more rapidly in the first than in the second. Where the river ran over a tract of easily-denuded rocks, the general level of the surface on either side was lowered nearly as fast as the river-channel was deepened, and the result was a stream flowing through a broad flat raised only slightly above its banks. But where a belt of less destructible rock was crossed, the general degradation of the surface went on much more slowly, and from this two results followed. First, the sides of the channel were but slightly modified, so that the valley retained to some degree the trench-like form with which it started and remained a gorge or narrow- glen. Secondly, in virtue of their superior power of holding out against denudation, these rocks remained standing up in a band of lofty ground above the flat formed by the removal of softer strata. In this way, by the gradual deepening of the channel of the stream, and the unequal lowering at the same time of the surface along dif- ferent parts of its course, the broad flat, the hill-range, and the gorge were produced by a connected and mutually-dependent set of operations. The Green and Yampa Rivers in Utah. Of the many cases in which rivers seem to have deliberately chosen a difficult in preference to an easy path, and to have gone out of their way to cut through ridges of very hard rock, this is the most remarkable that has come under my notice. It is illustrated by the sketch-map and section on figs. 213 and 214. The larger portion of the country consists of gently-rolling plains with a general slope to the south and an average elevation of 7000 feet above the sea. Across this country there runs in an east and west direction the chain of the Uinta Mountains, whose average elevation is some 10,000 feet above the sea. The edge of this range is shown by a border of hill-shading on fig. 213 : all within that line may be looked upon as averaging 10,000 feet above the sea : the 2Q 6io Geology. parts without about 7000. The Green River (AB) rises to the north of the mountains and flows southwards. On reaching the northern Fig. 213. SKETCH-PLAN or A PART OF THE UINTA MOUNTAINS. N S Fig. 214. SECTION or A PART or THE UINTA MOUNTAINS. slope of the Uinta range it continues its southerly course with profound indifference to the barrier which they present, flowing on in a deep narrow canon. Nothing strange in this, for the stream merely conforms to what we have now learned to look upon as a river's habitual practice. But it soon afterwards behaves in a more eccentric way than any- river which has yet come under our notice. It continues to flow south in the canon for a while, then it bends to the east, leaves the mountains, and returns to the plains on the north of them. Soon afterwards it again turns south and makes good its way across the whole range in another canon. The behaviour of the Yampa River (CD) is still more strange. It flows with a westerly course up against the eastern end of the mountain- range ; a very slight bend to the south would have carried it round the end of the ridge, along its southern margin, and so by an easy and what looks like the natural course into the Green River. But it does nothing of the kind ; it too cuts into the hills and flows in a canon to join the Green River in the very centre of the mountains. All this extraordinary conduct is easily accounted for when we look at the geological structure of the country. The Uinta Mountains are formed of an anticlinal of Quartzite and other hard rocks ; the plains are com- posed of soft strata lying nearly horizontal. High up on the mountains are outliers of the soft strata of the plains, one of which is shown at a on fig. 214. This tells us that there was a time when the Uinta ridge was completely buried beneath these soft rocks and when the surface of the ground was a plain whose position is shown by the dotted line at the top of the section in fig. 214. On this plain the rivers began to run, and they meandered over it in winding curves as Rivers and their Ways. 611 is the custom of rivers under such circumstances. One of these curves happened in the case of the Green River to bend in and out over the buried anticlinal : the course of the Yampa River happened to lie almost exactly above the axis of the anticlinal. General subaerial wear lowered the surface ; but it cleared away the soft beds much faster than the hard rocks ; and the underground ridge was thus stripped of its covering and came to stand up in a line of lofty hills. The rivers by this time had become established in their channels, and when they had cut down to the summit of the hard boss, there was nothing for them to do but to go on sawing through it. It is clear that the general lowering of the surface on the north of the hills could not go on faster than the deepening of the canons ; for the surface could not be lowered unless there were channels through which the waste could be carried away, and the only possible channels were the river-valleys. Rivers running away from the Sea. As a rule the source of a river is well inland, and those streams that rise near a coast flow more or less directly into the sea which is closest at hand. But there are rivers which do the very reverse. They spring close to a coast, they do not run into the adjoining part of the sea but flow directly away from it, and discharge themselves at last into some remote part of the sea, or into a distinct marine basin. The Yar in the Isle of Wight is a remarkable instance. It rises actually within six chains of the southern shore of the island ; it does not flow southwards into the English Channel hard by, but goes away due north across the island and empties itself into the. Solent on the north coast. What now intervenes between the source of the Yar and the English Channel is a low narrow barrier of Chalk, but there is every reason to think that at no distant date the waters of the English Channel were separated from the present source of the Yar by a broad strip of land. The Chalk ridge which runs across the Isle of Wight and terminates at the Needles has its exact counterpart in the Chalk ridge to the north of Swanage Bay which now terminates at the Foreland. These two ridges are but the remnants of a long continuous line of Chalk hills that connected the Isle of Wight with the so-called Island of Purbeck, and to the south of this chain there was doubtless a broad spread of land, such as would be enclosed within a line drawn from Durlestone Head to St. Catharine's Point. Under these circumstances the source of the Yar was well inland, and if we had seen the river under these conditions, it would not have struck us that there was anything abnormal about it.* The Yorkshire Derwent behaves in a similar way. It rises close to the coast near Filey, runs straight away from the sea towards the west, and then bending southwards joins the Ouse. Here again marine denudation is working back the coast-line, and when the river began to run there was a broad strip of high ground between its source and the German Ocean : or more likely there was at that time no * For an account of the changes in the physical geography of this district see " On the Superficial Deposits of Hampshire and the Isle of Wight," Quart. Joum. Geol. Soc. xxvi. (1878) 528. 6i2 Geology. German Ocean at all, England was then a part of the continent of Europe, and the Humber was a tributary of a great river which flowed northwards over what is now the bed of the North Sea. What will happen when the coast-line has been cut back up to the source of the river is not very easy to see. Probably what is now the upper portion of the stream will then commence to flow eastwards and empty itself into the German Ocean ; above this there will be a length of low swampy ground forming an ill-defined water-parting, and beyond that the river will keep its present direction of flow. Reversal of the Flow of Rivers. There are cases in which it seems likely that rivers at one time ran in the very opposite direc- tion to that which they now follow. That this has been the case with the Rhine was suggested by Dr. Hibbert,* and the idea has been worked out by Professor Ramsay, f This river after leaving the Lake of Constance cuts through the Jura between Schaffhausen and Basel, then flows north along a broad valley to Mainz, then breaches the great barrier of the Taunus and Hundsriick in a narrow gorge, and finally pursues its course over the great alluvial plain which it has itself formed to its mouths. Now it seems likely that once the broad valley between Mainz and Basel was filled up by a group of soft beds of Miocene age ; and that a river whose course coincided approxi- mately with that of the Rhine rose on the southern flank of the Taunus and ran southwards. The Jura had not then come into existence or was in a very rudimentary stage, and the river crossed the ground where the Jura now stands, and may have bent round the south-western end of the Alps and emptied itself into the Gulf of Lyons. At the end of the Miocene period upheaval of the Alpine range and of the Jura took place ; the southern outlet became blocked, and the country traversed by this river was tilted at its southern end and had a slope to the north given to it. This tilt caused the river to run to the north ; it cut its way down through the soft beds between Basel and Mainz and at the same time sawed a nick through the hard rocks of the Taunus and Hundsriick. Where the stream flowed across the soft beds the valley was widened by subaerial wear ; but in the hard rocks it retained its trench-like character ; and hence the valley is broad between Basel and Mainz and contracts into a gorge to the north of the latter town. Professor Ramsay has traced out the history of some British rivers in papers in the Quart. Journ. Geol. Soc. xxviii. (1872) 148; xxxii. (1876) 219. Breaching of great Mountain-chains by Rivers. Rivers are capable of feats still greater than any we have yet noticed. Even the loftiest mountain-chains are not able to bar their path, they cut across these in some cases with apparently as little effort as they make their way through the insignificant hill-ranges of our own country. The Indus for instance springs north of the Himalayas and forces its way across very nearly the highest part of that formidable chain. The explanation which has served us hitherto cannot be * History of the Extinct Volcanoes of the Basin of the Neuwied. t Quart. Journ. Geol. Soc. xxx. (1874) 81. Features due to the Action of Ice. 613 applied here. The capabilities of denudation are great, but the boldest speculator would hesitate to maintain that the Indus com- menced to run over a plain of marine denudation which was on a level with the summits of the Himalayas, and that these mountains have been carved out by denudation. Besides we shall shortly see that the mountains were certainly not formed in this way. But the geological history of the country supplies us with an explanation which is not open to such objections. The Indus rises in the Karakorum Mountains. This range is of great antiquity ; it was a mountain-chain not only long before the Himalayas had come into being, but even before a great part of the rocks which form the Himalayas had been deposited. At this distant period a stream which coincided pretty much with what is now the upper part of the Indus flowed down the southern slopes of the Karakorum, over a strip of land at their foot, and emptied itself into a sea which then covered the ground on which the Himalayas now stand. In this sea a vast thickness of rock was deposited. Then there came a time when this mass of rock was slowly raised : the bed of the sea became dry land, and as the coast-line moved southwards, the river continued its course over the newly-formed land and became gradually lengthened. The rocks were still squeezed up, and step by step rose into a huge broad-backed ridge that was afterwards to be- come the Himalayas. As the ridge lifted itself gradually into the air, the river sawed a notch across it. The deepening of the notch kept pace with the uplifting of the ridge, and a gorge reaching from the summit nearly to the base of the mountains was in this way cut out. A similar explanation applies to other cases where great mountain- chains are breached by rivers. The upper part of the" river is older than the mountain-chain, and the breach has been formed by the river eating its way down and down through the rocks which form the range while they were being slowly upraised.* SECTION V. FEATURES DUE TO THE ACTION OF ICE. We have already had to do with the denuding action of ice, but only in so far as it furnishes materials for the formation of Derivative rocks ; we will now inquire into its effects on the shape of the surface. We saw that ice-sheets and glaciers are always moving slowly from higher to lower levels, and that as they move they wear away the rocks over which they pass by means of the stones frozen into their under surface. The shape given to the face of a country by this grinding action is totally different from that produced by any other denuding agent, and therefore calls for special notice in the present chapter. But ice-worn surfaces have a further interest for the geologist. Changes in climate may cause ice to disappear, but the markings it impressed on the solid rock survive long after it has passed away; and * See Dr. E. Tietze, Jahrbuch der k. k. Geol. Reichs. xxviii. 581-610 ; and for a parallel case on a smaller scale T. Wilcox, Mountain Drainage of Eastern Tennessee, Proc. Ac. Nat. See. Philadelphia, 1874, p. 165. 6 1 4 Geology. the observer when he has once learned to recognise their distinctive character, infers from these surface-forms the former presence of an ice-sheet and the path it took, as readily and as certainly as the hunter draws his conclusions from the footmarks and trail of an animal. In this way we are enabled to show that countries which now enjoy a temperate climate were once placed under conditions such as now pre- vail only in Arctic regions. General Aspect of Ice-worn Districts. A sheet of ice as it flows over a country wears down all projecting points, and smooths off all rough places, and in this way the hills of an ice- worn tract get a general rounded hummocky outline. No words can give an adequate idea of the appearance of such districts ; but any one who has stood in the middle of a group of lofty hills formed of bare rock where ice- work has gone on on a large scale, never forgets the wonder with which he for the first time gazed on the sameness of the flowing contours which every hill offered to the view. It is as if some giant hand had taken sheet after sheet of emery-paper and rubbed and ground away till every prominent peak and bristling crag had been cleared off, and every valley-side smoothed down. The view in fig. 218 will perhaps give some faint notion of the peculiar aspect of such a landscape ; the hills on the left-hand side of the lake are strongly typical. In some cases this smoothing extends itself over the tops of the highest hills, and furnishes proof that the whole country has been wrapped in one universal mantle of ice, like Greenland and the Antarc- tic continent at the present day. Elsewhere the ice-worn surfaces are confined to the valleys and extend only to a certain height up the hill- sides, in which case the glaciation was less extensive arid the ice took the form of glaciers. The observer must not however jump too hastily to the conclusion that he has determined a line above which the ice did not rise, merely because he finds traces of its action plentiful on the lower part of the slopes and apparently absent on the hill-tops. Ice-tracks of course suffer, and are in time completely removed, by denu- dation ; they will evidently weather away faster on exposed summits than in lower and more sheltered spots, and this may be the explanation of their absence above a certain elevation. But if we find a fairly hard line below which the ground is very generally smoothed, and if above that line it is rugged and bristling and does not yield to the most careful search the faintest remnant or trace of ice- worn patches, then we may be pretty sure that that line marks the upper limit of the ice. Polished Surfaces. The features just described are those which strike the eye on a general view of an ice- worn country. The minor details which closer examination reveals are no less remarkable and no less important geologically. Rock surfaces worn by modern glaciers have frequently a polish as perfect as could be produced in a lathe ; and similar surfaces of very ancient date, when they have been protected from the air by a coating of clay or other impervious material, are found to have lost scarcely any of their original high polish. Scratches. We have already explained that the tools which enable ice to grind away the rocks over which it passes are stones frozen into Features due to the Action of Ice. its under surface. The high polish just mentioned is produced by fine sand ; stones cut grooves or striations according to their size, the small sharp points etching on the rocks scratches as fine as the most delicate work of a steel engraving, the larger blocks ploughing out flutings and coarse ruts. Fig. 215 shows a stone on which both fine and coarse striae have been impressed. These markings are of the utmost value. The scratches on the 616 Geology. stone figured above will be noted to be rudely parallel to one another, and this general direction is evidently that in which the ice flowed. By observing and recording the bearings of similar scratches on rocks in place, we shall be able to lay down the line of the path which the ice that made them took across the country. Also the relation of the scratches to the surface will enable us to form an idea of the extent of the glaciation, and to say whether it amounted to a universal ice-sheet or got no further than valley glaciers. In the first case the enormous accumulation produces a pressure sufficient to drive the ice from the interior to the coast, not exactly in straight lines over hill and across valley, but still with considerable disregard to the inequalities of the surface ; the scratches in consequence radiate in a general way outwards from the centres of maximum elevation, and frequently pass up the sides and across the summits of the hills. In the case of glaciers the motion of the ice and the scratches which it cuts are parallel to the trend of the valleys. Roches Moutonn6es. From the grooves and scratches just described we can lay down the course of the ice's motion, but they do not tell us which way it travelled. We can learn for instance that its path lay north and south, but not whether it came from north to south, or in a contrary direction. This information we gather from the shape of certain rounded hummocks, always found in glaciated districts, and called Roches Moutonnees. Fig. 216 is a sketch of one of these; and it will be noticed that while 216. ROCHE MOUTONNEE. the right-hand side rises from the ground with a gentle smoothly- rounded slope, the front is steep and comparatively unworn. If we examine a valley still occupied by a glacier, but in which the ice formerly extended lower down than now, we shall find that the gentle slopes of all the moutonneed bosses look up the valley, and their steep fronts all face the opposite way ; since the motion of the glacier is down the valley, this amounts to saying that the smooth faces point to the quarter from which the ice comes. Similarly, in a country from which ice has completely disappeared, if the moutonneed surfaces are preserved, we learn from them in what direction to look for the source of the ice. It is easy to see how this peculiar form came about, and why the opposite faces of the hillocks are so different. Let the line ~BS in fig. 217 represent the rugged surface of a rocky country over which a sheet of ice is moving in the direction of the arrow. As the ice is driven up the slope A, it will grind away all the inequalities and work the sur- face down to an even rounded outline. The debris produced will be Lakes. 6 if pushed on over the crest, and fall down in a bank in front of the hillock. The ice as it moves on will ride over the top of the heap, and Fig. 217. DIAGRAM TO ILLUSTRATE THE FORMATION OF ROCHES MOUTONNEES. not touch, or touch to a very small extent, the face I>, which will therefore retain its roughness. Where the ice impinges on the next projection of rock, the face which it meets will be worn smooth, and the opposite side will be protected by debris, and be little affected. In the end, when the ice has gone and the loose debris has been removed by denudation, there will remain only the part tinted dark in the figure, that is to say a couple of hummocks w r ith the outline which has been described as characteristic of roches moutonnees. Moraines. The moraines of vanished glaciers are as important as the traces of ice-action already mentioned, whether we look upon them as features in the scenery or as a means of enabling us to read the history of former glaciations. Longitudinal moraines remain as long lines of hmnmocky mounds or ridges running along the sides of valleys ; terminal moraines have a similar outline, but stretch across the valley in a horse-shoe-shaped curve, with its convexity pointing downwards. The latter have frequently suffered largely from denudation, but where they are well preserved they sometimes furnish an elegant proof that the glacier which gave rise to them dwindled away little by little before it finally disappeared. It is not uncommon to find in the upper recesses of mountain valleys a series of small terminal moraines, one within the other, each one more puny than the one below it. From these we learn that, after the ice had disappeared from the low country, glaciers still held their own in the uplands ; but that, as the climate improved, they shrunk back higher and higher up the valleys. Each moraine marks a line at which the glacier paused for a while in its retreat, and the diminishing size of the rubbish-heaps which it produced during each stationary interval points to a corresponding gradual decrease in the volume of the ice. SECTION VI.-LAKES. We may conveniently consider next the method of the formation of lakes. Dammed-up Lakes. The origin of some lakes is obvious enough. Just as an ornamental sheet of water or reservoir is formed by throwing an embankment across a valley, so the water of some 618 Geology. lakes is held back by natural dams, composed of materials different from the rock that forms the bottom and sides of the basin. An old terminal moraine often acts as a dam, the space above, which was once filled with ice, being now occupied partly by water. In a similar way landslips and streams of lava sometimes block up a valley, and pond back the water of its stream into a lake. Again, hummocks, such as Eskers and Sand-dunes, sometimes enclose lakes, the origin of which will be better understood when the student has read in the next section an account of the method of forma- tion of these moundy elevations. Volcanic craters also are sometimes converted into lakes when the volcanic activity has become extinct. Some deposits, such as that known as Glacial Drift, have been thrown down in an irregular manner, with a rough, uneven surface. Water accumulates in the hollows so formed, and gives rise to little lakes. It seems also not unlikely that great deposits of Glacial Drift may have played an important part in the formation of lakes in this way. Suppose we have a broad river valley with branches and tributaries, and that drift is thrown down along the course of the valleys in a very irregular manner : at some spots great piles of drift fill a valley to its brim, no trace of it can be seen at the surface, and its former existence can be detected only by borings : at other spots little or no drift is formed. The result will be that in place of a continuous line of valley we shall have a chain of deep hollows separated from one another by barriers of drift, and each hollow when filled with water becomes a lake. Dr. J. W. Spencer has tried to show that this is the origin of some at least of the great lakes of North America. The cross-section of Lake Ontario bears a strong resemblance to that of a longitudinal valley excavated by subaerial denudation. The rocks which form its boundaries consist of alternations of hard and soft beds dipping from north to south ; the axis of the lake therefore coincides approximately with the strike. The bed of the lake on the south side is steep, and rises in a series of steps and terraces : on the north side the slope of the bottom is more even and gradual : at the bottom of the steep southern slope is a deep trench. This is exactly the contour of the valley which subaerial denudation would cut out. On the south side the dip is into the hill, the outcrops of the hard beds form escarpments or steps, and between each pair of escarpments there is a terrace : the river flows at the foot of this steep slope, and the trench traversing the lake may well mark its channel : on the north side, where the dip is down into the valley, denudation forms a com- paratively gentle slope corresponding to the easy gradient of the northern part of the bed of the lake. Lakes Erie and Huron are shallow pans traversed by deep channels, and they bear considerable resemblance to broad flat valleys watered by a number of streams. Dr. Spencer has connected the trench along the bottom of Lake Ontario with an old river channel, now almost entirely filled up by drift, which runs into Lake Erie and is continued on to Lake Huron, and this channel he has shown was deep enough to lay dry both these Lakes. 619 lakes : he has not yet detected the course which this channel followed at the eastern end of Lake Ontario. The story therefore is not yet complete, but the evidence as far as it goes makes it not improbable that the basins of these three lakes lie on the line of an old valley of considerable size, that along a large part of its course this valley became choked with drift, but that the filling in was not continuous all along, and that the portions which were not filled up are now occupied by lakes.* Lakes are occasionally formed on the alluvial flats of great rivers by changes in the position of the bed of the stream. When the narrow neck of land between the extremities of one of the great winding sweeps of a river is cut through, the openings of the old channel frequently get choked up and a portion of the former bed is isolated and converted into a crescent-shaped sheet of water. Sometimes it happens that the whole space between the old and new channels is turned into a lake. The alluvial surfaces are not exactly flat, but usually rise towards the banks of the stream, because it is there that sediment is thrown down most largely ; in this way natural embank- ments are formed along the margins of the river, and when the raised edges of the old and new channels coalesce, the sunken space be- tween is filled by rain or high floods, and a closed sheet of water is produced. Lakes formed by these methods are plentiful along the course of the Mississippi, t Rock Basins. The lakes however formed in the ways just described form a very small minority of those which exist. Lakes are most abundant in northern regions, and by far the greater number of these cannot be ranged under any of the above heads. The reader may recollect a picture in Punch where a tourist from a manufacturing district remarks of the Cumberland Lakes, "We call them 'Resevors' in our country." The speech betrayed geological ignorance quite as much as a want of appreciation of the picturesque. These lakes are not bodies of water held back by dams resting on the rock of the country; they lie in hollows which are scooped out in rock itself below the general level of the floor of the valley, and the lip that holds back the water is solid and composed of the same rock as the bottom arid the hills on either side. Basins enclosed in this way by an unbroken rim of solid rock all round are called " rock basins," and it is in depressions of this nature that by far the larger number of lakes, in northern latitudes at least, are found to lie. Fig. 218 is a sketch of a lake which a little examination proves to lie in a rock basin. Along the sides, and especially on the left-hand margin, the ice-worn surfaces of the hills plunge down steeply beneath the water, arid a single glance is enough to assure us that the edges of the hollow are formed of solid rock. The nature of the barrier at the lower end is not so obvious at first, for the ground is much obscured by debris from the surrounding hills; but where the stream issues from the lake it has cleared away the loose matter, and we see clearly enough that the water is flowing over a lip of solid rock, and that * Proceedings American Phil. Soc. xix. (1881) 300. t Lyell, Second Visit to the United States, ii. 185, 203, 233. 620 Geology. it is this and not a dam of transported matter that holds back the water. The reader will notice in the smoothed and rounded outline of the hills and in the moutonn^ed bosses that project from the debris in Lakes. 621 the foreground, signs of former intense glaciation ; the surface of the rock also over which the issuing water flows is smoothed and highly polished. We shall see immediately that in many rock basins similar proofs of glaciation are present, and that the connection is probably not accidental. Rock Basins formed by Subsidence or Upheaval. It is evidently not altogether an easy matter to account for the pre- sence of a rock basin. One possible explanation suggests itself. The basins may have been formed by a sinking of the surface. It is doubtful whether we can point to a single instance in which it can be directly shown that a lake has originated in this manner ; but this is not an objection as weighty as might at first sight appear. We may fancy that it would be easy to submit the theory to the test of observation. If the surface sinks, the rocks underneath must bend with it ; and if we can prove that they have not been bent in the way the theory demands, the explanation falls to the ground. But in the actual cases we have to deal with it is extremely difficult, often quite impossible, to bring absolute proof that the upheavals and depressions required by this theory have or have not taken place. If the rocks of the country in which our lake lies had been but slightly disturbed previous to its formation, then evidence would be obtainable, for the bending down of the surface to form a hollow would give rise to a section like that shown in fig. 219. I believe I may safely say that Fig. 219. SECTION ACROSS A LAKE FORMED BY SUBSIDENCE. no case of this kind has ever been substantiated. It happens that the majority of rock basins occur in countries where the beds are very much disturbed and often violently contorted. Fig. 220 in fact repre- Fig. 220. SECTION ACROSS A ROCK BASIN. sents more nearly than fig. 219 the arrangement of the strata which form the beds of most rock basins. Now it is perfectly possible here that the rising boss on the left which forms the barrier damming up the lake may have been formed by the bending up of the rocks into a low anticlinal, and yet that we may be quite unable to detect proofs of this movement because the rocks were already crumpled and bent when it took place. In a group of rocks excessively folded and twisted it is by no means an easy thing to say whether all the contortions were produced at the same time, or whether after a certain amount of contortion and denudation had taken 622 Geology. place other foldings were added at a later date. Our difficulty is increased by the fact that in the majority of cases the depth of a lake is very small compared with its length, so that the slope of the bottom is very gentle ; the amount of folding required to form the basin would therefore be very slight, so small indeed as to be quite insensible in rocks already in a state of great contortion. In fact if we had a long deep valley cutting through highly-contorted rocks to begin with, and if a number of broad flat anticlinals were formed ranging athwart the valley, rock basins might be produced, and yet the amount of tilt necessary to give rise to basins with shelving bottoms would be so small as to be very difficult of detection in rocks that had been pre- viously disturbed.* This may possibly be the origin of those remarkable lakes in the Jordan valley, which lie far below the present sea-level. There can be no doubt that the long gorge in which the stream flows was cut out by a river, which probably emptied itself into the Gulf of Akabah. Subterranean movements then went on along the basin. The southern end was raised into a barrier, closing the former exit ; higher up, the movement gradually changed into one of depression, arid along a con- siderable part of the valley the ground was sunk deep below the sea- level ; but the depression was greater at some spots than others, and by this unequal bending down profound hollows were formed along the course of the stream, now occupied by the lakes in question. Lakes may also be formed when the bed of a large sea is laid dry by upheaval ; any pits or holes which exist on the bottom will then be converted into lakes. If these have no outlet, they remain salt and their saltness increases for reasons already given (p. 238). The salt lakes on the plateau of Central Asia belong to this class (p. 240). But if a river flows through such lakes, the influx of fresh water gradually sweetens them : even in such a case however we get some- times a hint as to their original condition by finding marine animals still living in them. The great lakes of Central Africa may have been formed in this way : some shells have recently been brought from them which are said to bear a strong resemblance to marine forms and which suggest ' the notion that they are the descendants of salt-water molluscs that have been modified by the changed condition under which they have been placed. Subsidence that may give rise to lakes may be caused by the dis- solving away of beds of soluble materials beneath the surface. This seems to have gone on to a large extent in Cheshire. The New Red Marl, which covers a large part of that county, contains thick lenticu- lar beds of Rock-salt. Percolating water gradually carries these away in solution, and forms great underground cavities into which the over- lying rocks sink down, and so depressions, which are soon filled with water, are formed on the surface. In places where brine is pumped from these beds for the manufacture of salt, the removal goes on more rapidly than under natural conditions and subsidences occur on a large scale. Rock Basins scooped out by Ice. The rock basins which * See Bonney, Quart. Journ. Geol. Soc. xxx. 479. . Lakes. 623 we find ranged along river valleys may possibly have been formed in the manner just described ; but such an explanation can hardly be applied to a vast number of lakes of this class. Over many countries in the northern hemisphere lakes lying in rock basins are scattered by thousands broadcast over the land. The northern part of North America, Norway and Sweden, the hilly parts of Scotland and Ireland, the English Lake country, and North Wales will serve as instances. To maintain that the ground has here sunk down into a shallow pan at every spot where a lake occurs, is to set at defiance the laws which have been found to govern the upheavals and depressions of the earth's crust. These movements do not take place in the haphazard way that such a view would require ; they are produced we have seen by the bending up of the rocks in long parallel arches, and if the basins had been formed by underground movement, they too would have been ranged along a series of parallel lines. A glance at a map of any of the countries just mentioned will show that this is not the case, the lakes are sprinkled here and there in a manner that seems to defy us to group them according to any law. But there is one very suggestive fact which has led Professor Ramsay to a most probable explanation of the origin of these rock basins. All the countries where rock basins are thickly and irregularly sown, have been at a period geologically recent subjected to ice-action. Glaciation and crowds of rock basins always go together ; in these parts of the world on the other hand which show no signs of former glaciation, lakes are comparatively rare, and many of those we meet with do not lie in rock basins. The constant association of ice-action and rock basins suggests the notion that the one may have been the cause of the other, and that the tool which scooped out the basins was a sheet of ice, the process being somewhat as follows. When a sheet of ice descends a slope and impinges on the flatter ground at its foot, the extremity, driven down by the pressure of the mass behind, acts like a great gouging tool and ploughs into the rocks of the plain. The cavity thus commenced is lengthened out as the ice advances, but the force of the thrust will grow less and less as we recede from its source, and also as the glacier moves lower down it melts away, and the thickness, and therefore the pressure due to its weight, gradually decreases. The amount of erosion will thus diminish outwards from the hill-foot, and the hollow formed will gradually shallow in that direction till it comes to nothing. In this way a trough will be worked out with a steep face on the side nearest the source of the ice, and a long slope shelving up gently in the opposite direction. This is found to be very generally the outline of a rock basin. The lake we have already given as an illustration shows it to perfection, as will be seen from fig. 221, which is a section on a true scale across it and a neighbouring lakelet. The slopes above the head of each lake are ice-worn, not unfrequently to such an extent that they are actual inclined planes, so steep and highly polished as to afford a very insecure foothold when clear of debris ; they plunge down at once into the water without the least change of inclination, the submerged portion being a direct continuation of that above the 624 Geology. level of the lake. At the lower ends rock surfaces, equally well smoothed, rise at a low angle from beneath the water and slope up gently till the next abrupt descent begins. There can be no question that the basins have been filled with a mass of moving ice, and we can readily realize how they may have been formed altogether by such an agent. A glacier, cascading as it were down a steep face, was driven forcibly against the flatter ground at the foot, and ate out a hole which was the beginning of the basin. A hollow once started, the constant wearing of the ice-flow would enlarge and deepen it, but it is easy to see that the slope of the bot- tom would be smaller at the lower than at the upper end ; down the one the ice slides with gravity in its favour, while it has to move up the other against the action of gravity; when it enters the hole there- fore its erosive power is greater than when it is leaving, so that in the one case a larger amount of material is removed and a steep face is produced, in the other a more gentle slope is formed. In fact the ice having got into the hole must get out of it, for the pressure from behind will not allow it to stand still. But the only way of getting out is to wear down the rock that stands in its way to a slope gentle enough to allow of the mass sliding up it. Exactly similar results will follow wherever a great sheet of ice flows over an uneven surface. We only want a depression to begin with. Wherever there is a little hollow the ice will go down into it, wear it deeper, and give it the same sort of shape as the basins just described. Not that rock basins will always exactly conform to this pattern. The relative hardness of the rocks of their floor will modify the result, greater erosion and therefore greater depth being produced where the ice crosses beds relatively soft. The thickness of the ice will also have an important effect ; when it has once been started at its work, the thicker the sheet the greater will be the weight driving it down, and the greater the depth to which it will penetrate. Rock basins will therefore be most likely to be formed, all other things being equal, beneath those parts of an ice-sheet where it is thickest. It is important for the full understanding of the theory of the ice origin of rock basins that the student should clearly realize how very shallow in comparison with their length these hollows are. Owing to the very general practice of using a scale for heights and depths larger than that employed for horizontal distances, most illustrations convey a very false idea of the shape of rock basins. The section of Lough Maarn in fig. 221, a section of the Lake of Geneva, given in Professor Ramsay's paper quoted below, or the section of Loch Lomond, facing p. 608 of Dr. James Geikie's " Great Ice Age," all of which are drawn to a true scale and therefore do not exaggerate the slopes, show clearly that these depressions, large as their absolute depth seems, are, when their relative dimensions are taken into account, only shallow pans, and that the inclination of their beds is by no means so great as that of many surfaces up which ice-sheets have certainly flowed. The arguments in favour of the glacial origin of many rock basins are very forcible, even though we may not yet have hit on the exact Surfaces not wholly due to Denudation. 625 nature of the mechanism by which ice has been able to 'scoop out these hollows ; but while we admit this, we must not lose sight of the possi- Lougli Maain. Lough Slievesnaght. Fig. 221. SECTION ALONG LOUGH MAAM AND LOUGH SLIEVESNAGHT, TWO ROCK BASINS, Co. DONEGAL, IRELAND. bility of some rock basins, specially some very large ones, having been formed by subterranean movements in the manner already described.* SECTION VII. SURFACES NOT WHOLLY DUE TO DENUDATION. So far we have been dealing with ordinary hills and valleys what we may call everyday features and we have seen that not only is denudation quite competent to produce these inequalities, but that we know of no other agent among existing forces that could have formed them. Valleys we have learned to look upon as troughs or trenches dug out by denudation, just as much as a ditch is dug out with a spade, while hills are the remnants which denudation has spared. There are however certain reliefs of the earth's surface in the forma- tion of which denudation has played only a subordinate part, and to these we will now turn our attention. The most important of the features that come under this head are Mountain-chains ; next, in order we may put Volcanic Cones ; and then we shall have to notice the minor instances of Eskers, Moraines, Sand-dunes, and Alluvial Flats. In the case of all but the last we shall find that though their main outlines have been determined by some cause other than denudation, they have by no means been unaffected by that all-present agent, and that all the lesser details of their surface-form are due to its action. Mountain-chains. The word mountain in its popular accepta- tion can be scarcely said to carry with it any very definite meaning. It is used vaguely for a very high or otherwise noteworthy hill, but the limit above which a hill must rise before it can be entitled to be called a mountain is purely arbitrary, and depends largely on its sur- roundings. The liigi for instance is so dwarfed by the neighbouring * For details consult Ramsay, Quart. Journ. Geol. Soc. xviii. 189 ; Phil. Mag. October 1864 and April 1865. A. Geikie, Trans. Geol. Soc. Glasgow, i. 86 and iii. 180. Belt, Quart. Journ. Geol. Soc. xx. 463. Haast, ibid. xxi. 130. Gastaldi, ibid. xxix. 396. Ward, ibid. xxv. 96 and xxxi. 152. J. Geikie, The Great Ice Age, chaps, xxiii. and xxiv. ; Quart. Journ. Geol. Soc. xxxiv. 819. Helland, Quart. Journ. Geol. Soc. xxxii. 142. Peach and Home, ibid. xxxv. 778 and xxxvi. 648. 2R 626 Geology. Alpine peaks that it is reckoned no more than a subordinate summit ; if it were transported to the flats of Holland, it would be there looked upon as a conspicuous mountain. But it is possible to frame a definition, though perhaps not a very rigid one, of what is meant by a mountain-chain. The one great lead- ing feature which distinguishes mountain-chains from the hills and ridges we have hitherto been dealing with we shall find in the end to be this. They are not blocks of rock that stick up because the matter that once surrounded them has been removed by denudation ; they owe their superior elevation to the fact that the rocks of which they are composed have been squeezed and ridged up to a greater height than the rocks of the country on either side. But this is not a truth that can be learned by direct observation ; it is rather a conclusion we arrive at only after having gone through a somewhat complex train of reasoning; it is not therefore very well suited to form the basis of a definition. But if all the regions that have undergone this squeezing-up process are found to agree in pos- sessing certain simple and easily-recognisable external characters, and if these characters are not found anywhere except in such regions, we shall have in these peculiarities a means by which the eye alone can decide whether any given tract of lofty ground is, or is not, entitled to be called a mountain-chain. Now there are two distinguishing features which most, if not all, mountain-chains present. 1st. Their breadth is small compared with their length. 2nd. They rise sharply, and are marked off clearly from the country on either side. It is by the first of these tests that we distinguish between a true mountain-range and a mere lofty plateau. The former consists of a long narrow ridge, or a succession of ridges running rudely parallel to each other, along the crests of which projecting peaks are perched in lines approximately rectilinear. A plateau or tableland is a broad expanse of elevated ground of a tolerably uniform height all over, and any points that rise prominently above its level are liable to be dotted about without order or arrangement. This chain-like structure may always be recognised if we take a broad view of any great mountain-range, though. here and there it may be difficult of detection or may be for a while lost altogether. This will be the case at those great knots of mountains which are formed where two or more ranges meet or cross one another ; but such excep- tions are of the nature of local accidents, and do not prevent us from realizing the general character of the ridges as a whole, any more than the fact that a long street opens out every here and there into broad squares prevents our seeing that on the whole it is a street and not a square. The second feature will be found to. be present to a far greater or less extent in all great mountain-ranges. It is true that the main central chain is usually flanked by lower parallel ridges, and that these lessen in some measure the abruptness of the transition from the high lands to the plains, and make it difficult to say exactly where one ends Surfaces not wholly due to Denudation. 627 and the other begins ; but for all this the eye seldom fails to recognise on a general view the existence of a change in feature more or less sudden, even though it may be hardly possible to lay one's finger on the actual spot where it occurs. The reader will perhaps form a good idea of the broad structure and general character of a mountain-chain in this way. Let him take a number of long, squat, triangular prisms, of different sizes, and lay them, with their broad faces downwards, parallel to one another on a table, the highest in the middle, the smallest outside, and the rest ranged between in the order of their size ; then let him cut and hack the upper edges till their outline becomes jagged and serrated. The group will then form a very fair representation of a mountain-chain composed of a number of parallel ridges, increasing in height towards the centre, and with prominent peaks ranged along their crests ; and the way in which the group is clearly marked off from the flat of the table will enable him to realize how a mountain-chain rises boldly and sharply out of the country on either side of it. The definition just given will exclude from the class of mountain- chains many tracts of lofty country usually spoken of as mountainous. For instance it will not allow of the existence of mountain-ranges in the Highlands of Scotland. If we were to look down on that country from a balloon, we should see nothing corresponding to our table and array of prisms. On the contrary, it would appear to be a great table- land, not perfectly flat, but with a surface slightly undulating like that of a sea roughened by the wind ; valleys would be seen to cut across it, but they would look like trenches and \vould scarcely interfere with the apparent general evenness of the surface. And if we checked this first impression by the aid of a raised map of the district, we should find that our eyes had not deceived us. A sheet of paper laid horizon- tally on such a map will touch, or very nearly touch, the tops of almost all the hills ; here and there a hole may have to be made to allow a projecting point to come through, but these are few in number, none of them rise much above the average level of the surrounding summits, and most of them occur at haphazard and with no tendency to a linear arrangement. On the other hand, if we turn to Italy we shall realize the contrast between a lofty tableland, like that of Scotland, and a true mountain- chain ; for the Apennines, in spite of their moderate height, are clearly entitled to that rank. They form a range decidedly long and narrow, and they are flanked on both sides by ground markedly inferior to them in elevation. The two distinguishing features on which we have been commenting, though they are useful as enabling us to recognise mountain-chains, throw no light on the mode of their formation. We now pass on to facts which have a most important bearing in this direction. All hill- ranges which present these features are found to agree in possessing two peculiarities. First the, rock groups of which they are composed are many times thicker in the mountain-chain than anyivhere else. The strata of a mountain-chain can be divided into a number of groups distinguished 628 Geology. from one another by differences in lithological character and still more certainly by the fact that each contains an assemblage of fossils more or less peculiar to it. The strata of the adjoining flats can be divided into a set of corresponding groups, and we know that when a rock group of the mountains and a rock group of the plains contain the same assemblage of fossils, the two were formed at the same time. But where the thickness of a rock group in the plains does not exceed some hundreds of feet, the contemporaneous group of strata in the mountains runs up to thousands. And we can easily see a reason for this. In the plains the several rock groups are separated by uncon- formities ; deposition was interrupted many times over, and during the intervals when deposition was suspended large thicknesses of the rocks last formed were swept away by denudation. But in the mountains group follows group in regular conformable succession. The process of deposition went on without check, and there was no denudation of any one group before the next group was laid down upon it. In the Alps the pile of strata that was thus accumulated reaches a thickness of 50,000 feet, or near upon 10 miles, the maximum thickness of the strata of the same age in England do not exceed 12,000 feet. In the Appalachian chain the sedimentary deposits reach 40,000 and in the Rocky Mountains 60,000 feet. Secondly the rocks of a mountain-chain are violently contorted. The most striking form of distortion is crumpling on an extensive scale, by which the beds have been folded into curves of enormous radius, and puckered up into the most complicated contortions ; in many cases this has gone so far as to bend over the rocks in the manner shown in fig. 222, and give rise to perfect and repeated inversion. Fig. 222. INVERSION OF MOUNTAIN STRATA BY INTENSE FOLDING. Instances of this kind have already been given, and it has been pointed out how completely mountain sections may mislead us as to Surfaces not wholly due to Denudation. 629 the true order of the beds, when parts of the folds have been removed by denudation. Faulting on a large scale is also very frequently met with among the disturbed strata of mountain-chains, and the faults are frequently reversed. The axes of the folds and the faults are in a general way parallel to the trend of the range. Disturbance of a more violent character still is not uncommon, when the rocks have been jammed together and mashed up in a way that defies description. Further the rocks of mountain-chains are very largely cleaved, the planes of cleavage ranging parallel to the general trend of the chain. The fact that the longer axes of the folds and the planes of cleavage are parallel to the general direction of the chain, shows that the rocks of mountain-chains have been powerfully compressed by a force which acted in a direction approximately horizontal and perpendicular to the trend of the chain. Mountain-chains too usually show a central core of Granite or some massive crystalline rock ; this shades off into Gneiss, Mica-schist, or some form of foliated rock ; and from the last a gradual passage can be traced into unaltered derivative rocks. The central core is therefore not intrusive, but is the product of a very advanced stage of metamor- phism. A mountain-chain then is made up of a number of long narrow parallel lofty ridges ; it is sharply marked off from the country on either side ; the rocks of which it is composed are abnormally thick, and have been squeezed forcibly together by a force which acted per- pendicular to the axis of the chain ; there is usually a belt along the centre of intensely-metamorphosed rock. These being the facts, what conclusion do they lead us to as to the formation of mountain-chains 1 The great thickness of the rock groups, and the fact that a large part of them are shallow-water deposits, tell us that the site of a mountain-chain was formerly covered by the sea, and that the bed of that sea continued to sink slowly but steadily till thousands upon thousands of feet of strata had been laid down upon it. To the great trough-shaped mass of rock which was thus accumulated Dana has given the name of a Geosyndinal. At last there came a time when the downward motion was arrested and the rocks of the geosynclinal began to be forced upwards. The uplifting of sedimentary rocks we have already seen has been brought about in every case by a bending of them into troughs arid arches by the action of horizontal pressure. When we come to inquire how this pressure is produced we shall find that the earth's crust is always in a state of compression ; that the force of compression keeps constantly increasing till it grows powerful enough to overcome the resistance offered by the rocks on which it is acting ; and that then the folding of the rocks begins. If the mass of rock to be folded is of no very great thickness, a moderate amount of pressure suffices for the task, the work goes on quietly, broad arches are produced, and there is no very violent contortion. But in the case of the geosynclinal of a mountain-chain the com- pressing force had before it a task of unusual difficulty ; the mass of rock with which it had to deal was of enormous thickness; the 630 Geology. resistance was tremendous, and the force which could cope with it must have been correspondingly great. When at last the pressure gathered head enough to master the stubborn mass, its magnitude led to two results. First the rocks were jammed up into a narrow space, crushed and crumpled into the sharpest of bends, the anticlinals were tilted over and inversion and reversed faults were produced, and all the violent disturbances characteristic of mountain-chains were brought about. Metamorphism would be another result. In the centre of the mass, where the rocks could be no longer jammed any closer, arrested motion would generate heat ; and here granitoid and other forms of non-intrusive metamorphic rocks would be produced. The lower portions of the mass would come within the range of the earth's internal temperature ; by the combined action of heat and moisture they would be softened ; the plastic stuff would be squeezed up through rents ; and the fan-shaped axes of foliated rock which have been already described (p. 424) would be produced. Secondly the close packing of the thick body of rock in a horizontal direction would squeeze it up vertically to a very considerable height, and a long and lofty ridge would be formed rising sharply from the comparatively undisturbed ground on either side. To this ridge Dana has given the name of Geanticlinal. But the squeezing up of the mass of rock into this ridge was only the first step in the formation of a mountain-chain. While this was going on denudation was not idle, and it continued to work when the elevation was completed. As the ridge was raised higher, it became more and more exposed to the action of the elements, and subaerial denuding forces were enabled to act upon it with more and more telling effect. By them the huge uncouth mass was gradually worked into its present shape, and carved out into an assemblage of bristling peaks, craggy precipices, ragged gorges, and open valleys. There is yet another circumstance which tended in some instances at least to intensify the results of the compressing action in the case of mountain-chains, and to confine them to narrow strips of the earth's surface. In some cases the foundation on which a mountain-chain stands was already rent and shattered before the deposition of the rocks of the geosynclinal began ; it was even then a weak belt which would readily give way to any pressure that was brought to bear upon it. After the geosynclinal mass had accumulated the state of things stood thus ; deep down there was a long narrow fissured zone ready to yield to the force of compression, above was a thick and stiff coating which kept down the rocks of this zone and forbade them to give way. This covering may be compared to the scar which forms over a deep wound that is still unhealed below ; and just as the scar prevents the escape of the pus and unhealthy secretions which are forming underneath, so did this scar of rock check for a long time the struggles of the broken rock beneath to relieve itself by giving way to the force of compression. Resistance became at last no longer possible, but the crushing and shattering which ensued when the stubborn mass was forced to yield, show that it succumbed only after a long tussle and only to a force of enormous energy. The results too would be confined to the portion Surfaces not wholly due to Denudation. 63 1 of the earth's crust lying directly over the fissured zone, or would at least be most pronounced within that portion, and this is one reason why the disturbed belt is narrow in comparison with its length. How the preliminary fissures were formed, and what it was that gave rise to the compressing force, are questions that will be discussed in the next chapter. In all cases there have been three steps in the formation of a mountain-chain. First the deposition of the vast thickness of the rocks of the geosynclinal. Secondly the squeezing up of this mass of rock into a geanticlinal, and the production of a long narrow and lofty ridge. Thirdly the carving out of this shapeless mass by denudation into peaks and valleys. The second and third steps of the process went on to a large extent simultaneously. In some cases, perhaps in all, the deposition of the rocks of the geosynclinal was preceded by the production of a long narrow belt of fissured ground in the earth's crust, which became the axis of depression during the formation of the geosynclinal and of elevation during the formation of the geanticlinal. It may be asked whether it is necessary to call in the aid of special machinery for the production of mountain-chains, and why they can- not be looked upon, like other hills, simply as remnants that have escaped denudation. There are two main reasons why we must seek an explanation of the origin of mountain-chains different from that which sufficed for ordinary elevations. In the first place the exceptional amount of contortion, cleavage, and metamorphism present in all mountain-chains shows that mountain- building was effected by a process different from that which formed hills where these peculiarities are not found. Secondly in order to get a mountain-chain by denudation alone, an amount of rock far greater than we have any reason to believe denudation can have removed must have been carried away. It must not be supposed that the distinction between mountain- chains and plateaus can be always rigidly maintained ; there are elevated tracts which, are somewhat intermediate between the two, and about which it is not easy to say to which class they ought to be referred. The strata of tablelands are sometimes folded and con- torted very much in the same manner and nearly to the same extent as those of mountain-chains, and show locally intense . crumpling and inversion. But in the one case the plication has been widespread, and the resulting contortion is consequently on the whole less violent ; in the other it has been localized and concentrated along certain lines, whereby the effects have been rendered more pronounced and confined to a comparatively narrow belt. Volcanic Cones. Volcanic cones, the reader will recollect, are mounds of fragments of rock which were shot out of a hole in the ground and piled up in a heap round it, with layers of lava poured from time to time over the pile in a semifluid state out of the same orifice. Neither denudation nor elevation had anything to do with their original formation, but the former agent of course, as time goes on, modifies their shape ; by the washing down of their friable 63 2 Geology. materials their conical abruptness is diminished, and gullies and gorges are scored down their flanks. Eskers. Among the most remarkable of the minor features of hilly districts in northern latitudes are certain long, winding ridges and hummocky mounds of gravel and sand, which go by the name of Kames in Scotland, and Eskers in Ireland. They rise boldly and sharply with steep slopes, to heights of occasionally as much as 100 feet and sometimes more, from the ground on which they stand, and the singularity of their appearance has attracted the attention of others beside geologists. Fairy legends still hover around them ; they are pointed out as the ropes of sand in the manufacture of which an enchanter strove to keep a restless demon out of mischief ; and they were utilized as natural earthworks in the days of early warfare. To account for the origin of these singular hillocks, numerous theories have been propounded ; there can be little doubt, however, that they have not all been produced in the same way. Some so-called eskers are certainly nothing but mounds which have been carved by denudation out of a thick sheet of gravel ; these present no peculiarity which entitles to notice in the present section. But there are others to which this explanation will not apply, and which undoubtedly owe their shape in a large measure to the manner in which their materials have been heaped up. Several facts lead us to this conclusion. It not unfrequently happens that the long ridges run together and enclose oval-shaped hollows without an outlet, which are sometimes still occupied by tarns and sometimes by peaty or alluvial deposits formed by the silting up of lakes that once lay in them. It is evident that these depressions could not have been cut out of a sheet of gravel by rain or river action, because there is no road by which a stream of water could escape from them ; and the only way we can account for their occur- rence is by supposing that the gravel was piled up in heaps round the central hollow, so as to enclose it completely on all sides. This conclusion is further strengthened by the internal structure of the kames. When cut across, they show a section like that in fig. 223. The gravel is very distinctly though irregularly bedded, and Fig. 223. the beds arch over, so that in a general way the direction and amount of the dip is about the same as the slopes of the surface of the ridge. This is just the structure that would be produced if the materials had been heaped up by currents coming alternately from opposite quarters. Such conditions exist where a river with fall enough to enable it to carry down gravel enters a tidal sea. The greater part of the heavy Surfaces not wholly due to Denudation. 633 material is let fall near the mouth of the river and forms a " bar." At low water there is nothing to check the force of the stream, and it rolls the gravel up the inner face of the bar and arranges it in layers dipping towards the land ; as the tide rises the river is pounded back, and the incoming waves roll pebbles up the outer face of the bar, spreading them out in beds which dip towards the sea. At very many spots where eskers occur, exactly such conditions as these would be produced if the land were submerged. Eskers are extremely common, for instance, where large mountain-valleys open out into flatter country. Supposing the sea to encroach as far as the mouths of the valleys, the load of debris brought down by the moun- tain torrents would be tossed about alternately by the stream and the incoming tide, and arranged in mounds and ridges. An excellent instance of eskers lying in such a situation is found in the lower part of Ennerdale, and is illustrated by fig. 224. The sketch is taken just where the hills of the Lake country begin to rise from the plain of West Cumberland. The long, narrow mountain- valley is seen stretching away in the distance ; the two moundy hills in the foreground with trees on them are eskers planted just where the valley opens out on to the flat country ; they form part of a group which runs across the mouth of the valley and extends far out into the plain. Another favourite locality for eskers is a valley which submergence would convert into a narrow strait connecting opposite seas. Along such a passage tides coming in opposite directions race furiously, arid, where they meet, the materials swept along by the currents are piled up in mounds and ridges having the outline and structure of eskers. Some fine groups of eskers are perched on plateaus ; in such a case we find that a certain submergence would convert the plateau into a low spit of land, over which the tides would wash at high water from opposite quarters. One or other of the explanations just given will account for the formation of a large number of these singular hummocks, but not for all. We occasionally meet with long snake-like ridges, winding over the country with considerable disregard to the inequalities of the sur- face, and it is by no means easy to say exactly how these were formed. In Scandinavia again, long ridges of gravel and sand, known as Asar, are plentiful ; they can sometimes be traced for more than a hundred English miles, and their origin has not yet been satisfactorily made out.* Moraines. Among the minor reliefs not due to denudation we may reckon Glacier Moraines. In outward form they are often very like eskers, and the two have not unfrequently been mistaken for one another. In section however it is always possible to distinguish * See A. Geikie on the Glacial Drift of Scotland, p. 112. J. Geikie, The Great Ice Age, p. 407 ; and Geol. Mag. ix. 307. Some very happy suggestions in a paper of Professor Jamieson's, Quart. Journ. Geol. Soc. xxx. 317. Kinahan, Explanation of Sheets 115 and 116 of the Geol. Survey Map of Ireland, pp. 13 and 30; Dublin Quart. Journ. Science, iv. 109, vi. 249; Dublin Geol. Soc. Journ. x. ; Journ. Royal Geol. Soc. of Ireland, vol. i. pt. 3 ; Geol. Mag. 2nd series, ii. 86. Rev. M. H. Close, Journ. Royal Geol. Soc. of Ireland, vol. .i. pt. 3. 634 Geology. between them. An esker is composed of rounded gravel usually well bedded ; a moraine consists of angular blocks of all sizes and shapes, \ jx; \ ', , ["' I' v '-' jumbled together without order or arrangement and with no regard to size or weight. The moraines of large glaciers form hills of con- siderable size : those of the Dora Baltea, opposite the mouth of the Surfaces not wholly due to Denudation. 635 valley of Aosta, rise from the plains of Piedmont to heights of 1500, and in one place of nearly 2000 feet, and have a frontage of at least fifty miles ; the lateral moraines stretch along the valley in ridges equally conspicuous. Sand-dunes. Somewhat allied to eskers are the mounds of sand swept off the shore by winds and piled up inland in hillocks. They often reach a considerable height and assume wild fantastic forms ; the slope of the inland side is much steeper than that of the side which faces the sea ; in section the successive layers by which they were formed can often be traced. They are never permanent, but shift their position and change their shape with every gale.* Though most commonly found near the shore they are not confined to that locality, but are formed far inland if a supply of fine dry sand is present. Thus the sand furnished by the weathering of the New Red Sandstone of -the centre of England is sometimes heaped up into small dunes. Lakes enclosed by heaped-up Mounds. The different kinds of mound-like elevations just described, which have been formed by the heaping up of their materials, are frequently so arranged as completely to enclose a hollow, and when this becomes filled with water a tarn or lake is produced. Old River-terraces. The formation of the Alluvial Flats which stretch along the lower parts of the course of most rivers has been already described (p. 605). Their flat surface is one of deposi- tion and not of denudation. We also frequently find perched at different heights on the flanks of a valley a succession of terraces with flat surfaces, composed of gravel, sand, or silt, similar to that of the alluvial bottom. These are the remnants of old alluvial flats formed by the river when it flowed at higher levels than now. Fig. 225 is a Fig. 225. SECTION ACROSS A VALLEY WITH OLD RIVER-TERRACES. a, ft. Terraces of old alluvium, c. Present alluvial flat. 1. Level of the river when a was laid down. 2. Level of the river when b was laid down. section across such a valley, showing two such terraces. The dotted line 1 marks what was at one time the bottom of the valley. The river flowed at this level, with a fall not sufficient to enable it to cut down its bed, long enough to enable it to spread out a sheet of alluvium. Afterwards, owing to a change in physical geography, the * For an account of the extensive Sand-dunes of Les Landes, which are among the largest known, see Elie de Beaumont, Le9ons de Geoloique pratique ; Bremontier, Memoire sur les Dunes ; E. Reclus, Le Littoral ae la France ; Delesse, Lithologie du Fond des Mers. 636 Geology. fall or volume of the river increased ; it began to cut down its chan- nel, and the valley was deepened. During this process the whole of the alluvial sheet was carried away except the bit at a. The deepen- ing of the valley went on till it was cut down to the level 2, when the fall was so far decreased that erosion ceased, and a second alluvial flat was produced. Then the deepening process began again, a great part of the second alluvial deposit was swept off, but two patches Ib remain at corresponding levels on either side of the valley to mark its position. When the valley had been eaten out to its present depth, the stream again began to form deposits on each side, and produced the present flat c. Many river-terraces have been formed in the manner just described, but probably not all. For instance a very ingenious explanation of the formation of gravel terraces by the aid of glaciers has been sug- gested by Professor Jamieson in the paper quoted a little way back (Quart. Journ. Geol. Soc. of London, xxx. 333). Sea-beaches. We have seen that the action of the sea tends to wear down whatever stands in its way to a uniform level. By this means, if the land remain long enough at the same level, a notch or shelf is cut around the coast, and upon the terrace so formed the tides spread out sand and shingle. Raised Beaches. These sea-beaches correspond among marine deposits to the alluvial flats of rivers ; and just as a river- valley is sometimes edged with old alluvial terraces, so we occasionally find terraces of sea-sand and shingle, fringing the coast at various heights above the present sea-level, which were formed when the land stood lower than at present. Fig. 226 illustrates such a case. A is the L _ Fig. 226. SECTION or MODERN AND OLD SEA-BEACH. A. Modern beach. B. Modern shingle ridge. a. Ancient ditto. 1). Ancient ditto. L. Present high-tide level. present beach bounded on the landward side by a ridge of shingle thrown up by the waves. Above this there is an old beach a and a shingle ridge b, corresponding in every respect to A and , and evi- dently formed when the land stood so much lower that the tides ran up as far as b. These old marine terraces go by the name of Raised Beaches ; they are frequently bounded towards the land by lines of bluffs, in which it is easy to recognise former sea-cliffs ; the caves worn in them by the action of the waves, and sometimes even the marine shells that lived on their face, often remain long after the sea has retired. Surfaces of Deltas. When a tract of low land has been formed by the accumulation of sediment at the mouth of a river, fresh-water Surfaces not wholly due to Denudation. 637 or marine alluvium is spread over it during floods or high tides, and it acquires an even surface. In this way the whole of the Nether- lands has been formed out of mud brought down by the Rhine. Silted-up Lakes. Where a lake has been filled up by the deposition of sediment, a flat resembling the alluvial plains of rivers is produced. In all these cases of alluvial surfaces their flatness is the result, not of denudation, but of the slow and regular deposition of sediment in horizontal beds. As they are for the most part low-lying, they occupy positions where the action of denudation is feeble, and they therefore retain for a long time their original evenness of surface. Prairies and Deserts. It has been suggested that the wide, rolling, dry prairies of North America have originated in the filling up of a great sheet of water which once extended over parts of Iowa, Illinois, Indiana, and Michigan, and of which the present North American lakes are the dwindled remnants.* It may be also that deserts, such as the Sahara and those in the interior of Australia, are old sea-bottoms but little modified by denudation. Summary. When we come to sum up the results of this chapter, we find that with a very few unimportant exceptions the dry land has everywhere a carved and sculptured surface, and that the tool which gave it its present shape was water, liquid or solid. In the majority of cases the contours and inequalities of the ground are due to this cause alone ; hills exist, not because the materials of which they are composed have been pushed up higher than the sur- rounding country, but because, while denudation carried away some parts, other parts were better able to hold out against its wearing action and were left standing up. Valleys have not been produced by a bending down or fissuring of the earth's crust, but are trenches eaten out by running water or moving ice. The sea and subaerial denuding forces had each a distinct share in the work. As continuous gentle elevation bent up the sea-bottom into the air, the waves pared it down to an even surface, known as a Plain of Marine Denudation, and subaerial agents carved this out into hills and valleys. The action of the one may be compared to the labour of the quarryman, who furnishes a rough-hewn slab ; the work of the others resembles that of a sculptor, who carves out on the surface of the marble a subject in relief. In the case of great mountain-chains however and the broad valleys that lie between them, the elevatory forces have played a more pro- minent part in determining the shape of the surface. A long narrow zone of the earth's crust was ridged up faster than denudation could wear it away, or under circumstances where denudation could not act, and thus the main shape and direction of the range was established. Thus much must be assigned to elevation, but all the lesser details are the work of denudation, which cut out the peaks that crown and the gorges that traverse the ridges. * On the Origin of the Prairies of the Valley of the Mississippi, Professor Alex. Winchell, Silliman's Journ. 2nd series, xxxviii. 332, 444. 638 Geology. In some cases then elevation has had a leading share in determining the reliefs of the earth's surface, and water has given the finishing touches; in the majority of cases the inequalities, great and small alike, have been wholly the result of denudation. The chief exceptions to this sweeping statement are the cones heaped up by volcanic discharges ; the mounds and ridges of sand and gravel piled up by waves and wind ; moraines ; and the flats formed by the deposition of alluvial sediment and by the silting up of lakes. Of the abundant literature on the subject of the present chapter the following may be specially commended to the reader's notice. Hutton's Theory of the Earth, and Play fair's Illustrations of the Huttonian Theory. Scrope. The Geology arid Extinct Volcanoes of Central France, chap. ix. Ramsay. On the Denudation of South Wales and the adjacent Counties of England. Memoirs of the Geological Survey of Great Britain, i. 297. The Physical Geology and Geography of Great Britain. The Old Glaciers of Switzerland and North Wales. J. B. Jukes. On the Mode of Formation of some of the River- val- leys of the South of Ireland. Quart. Journ. Geol. Soc. xviii. 378. A. Geikie. The Scenery of Scotland viewed in connection with its Physical Geology. On the Phenomena of the Glacial Drift of Scot- land. Trans. Geol. Soc. of Glasgow, vol. i. part 2. Earth Sculpture. Nature, ix. 50. Trans. Edinburgh Geol. Soc. ii. 248. W. Whitaker. Subaerial Denudation. Geol. Mag. iv. 327, 447, 483. C. Le Neve Foster and W. Topley. On the Superficial Deposits of the Valley of the Medway, with Remarks on the Denudation of Valleys. Quart. Journ. Geol. Soc. xxi. 443. W. Topley. Notes on the Physical Geography of East Yorkshire. Geol. Mag. iii. 435. J. Geikie. The Great Ice Age, chaps, xxiii., xxiv. Note D. Professor F. V. Hay den. United States Geological Survey of the Territories. Profiles, Sections, and other illustrations designed to accompany the final report of the Chief Geologist of the Survey. New York, Julius Bren, 1872. (Contains admirable instances of escarp- ments, dip-slopes, tabular outliers, and other features resulting from denudation.) Sun Pictures of the Rocky Mountains. The reader will do well to compare with the theory of surface-sculp- ture upheld in the preceding memoirs, chapter xix. of the late Professor Phillips' Geology of the Valley of the Thames. Elegant and ingenious as is the explanation there put forward, there is about it an unsatis- factory vagueness and want of definition, which contrasts strongly with the sharp precision and logical coherence of the views on the subject of which a sketch has been attempted in the preceding pages, and which are steadily gaining ground among modern geologists. CHAPTER XIV. ORIGINAL FLUIDITY AND PRESENT CONDITION OF THE INTERIOR OF THE EARTH. CAUSE OF UPHEAVAL AND CONTORTION. ORIGIN OF THE HEAT REQUIRED FOR VOLCANIC ENERGY AND METAMORPHISM. REMARKS ON SPECULATIVE GEOLOGY. "Sit mihi fas Pandere res alta terra et caligine inersas. " VIRGIL. SECTION I. THE PRESENT PHYSICAL CONDITION OF THE EARTH. IT was pointed out in the opening chapter that the geologist's first business was to make himself acquainted with those portions of the earth which he could actually observe, or the nature of which observa- tions made on the surface would enable him to infer with very trifling risk of error ; and that until he had mastered this branch of the sub- ject he would not be in a position to speculate on the character of the inaccessible interior. The time has now come when we may enter upon this fascinating but, in the present state of our knowledge, some- what unsatisfactory theme. The subject is not one of barren curiosity. Till we do know what is going on far down under our feet, we can only very imperfectly explain several things that are happening or are now visible at the surface. We cannot say, for instance, where lies the source of volcanic energy, or what is the force that has given rise to folding, contortion, and faulting. When we reflect on the great importance of a thorough knowledge of faults to the miner, we see that even the somewhat abstruse speculations in which we are about to indulge are not without a practical bearing. It is evident that we can learn nothing by direct observation about the nature of the earth's interior. As in all those cases where we have to reason about matters which are beyond the grasp of our senses, we must begin with an hypothesis, which may be suggested to us by some facts of observation, or may be purely the outcome of our own mental ingenuity. We then ascertain by deductive reasoning what results ought to follow under certain circumstances if our hypothesis were true. Next we observe what actually does take place under these circumstances. 640 Geology. If the observed facts are just those which must necessarily have been produced provided our hypothesis were true, and if we find this to be the case every time we put our hypothesis to the test, the probability that the hypothesis is true becomes very strong indeed and we then venture to call it a theory. The strongest evidence in favour of the truth of a theory is when it enables us to predict beforehand what will be the result of an experiment which has never been made. Thus the Undulatory Theory of Light explained thoroughly all the facts that were known about the behaviour of light before the theory was framed. But it achieved a far greater triumph than this. It was ascertained by mathematical calculation that if a ray of light passed in a certain direction through a certain crystal, it ought, if the Undu- latory Theory were true, not to emerge as a ray, but to spread out into a hollow cone. The experiment had never been made, indeed no one had ever thought of making it, but when it was made the result came out exactly what theory had predicted. Even the angle of the cone agreed precisely with the angle assigned by calculation. Now in the case before us the main facts we can learn from observa- tion, which are of use in checking and estimating the probability of the truth of any hypothesis that may occur to us, are these that the earth is a spheroid of revolution very nearly ; that its mean density is about double the average density of the surface rocks ; and that, as far as we have been able to penetrate, it grows steadily hotter as we go down, and must therefore be constantly losing heat. Any speculations we may indulge in about the deeply-seated regions of the earth must be consistent with these facts of observation ; but the facts do not of themselves help us much to an hypothesis about the nature of the interior and the process by which its present condi- tion was arrived at. Some such hypothesis we must have, and it appears that we must either trust entirely to our own ingenuity to invent it, or look beyond the earth for the facts that are to suggest it. Now observation of cosmical phenomena has suggested a theory of the development of the solar system known as the Nebular Hypothesis, which if it can be securely established will aid us materially in our present inquiry, for it will tell us what was the state of the earth's interior at a very remote period, and what changes it has been passing through since, and so will enable us to make probable conjectures as to the condition it has by this time arrived at. This hypothesis, \vhich had presented itself to Swedenborg, Thomas Wright, and Kant, was afterwards more fully developed by Laplace. The substance of it is as follows. There are in the heavens faintly luminous cloudy masses known as nebulae ; the spectroscope has lately revealed to us the fact that some of these are bodies of glowing hot gas, and the appearance of some of them is such as would be produced by. rotation round an axis. As the heat escapes from these by radiation into space, they must con- tract ; contraction causes them to rotate more rapidly ; the rotation at last becomes so rapid that the central attraction is no longer able to overcome the tendency of the outside portion to fly off, a ring is sepa- The Present Physical Condition of the Earth. 641 rated which afterwards collects together into a ball. By a continua- tion of this process the nebula is at last broken up into a number of balls, all of which revolve round the centre of the original mass and rotate on their axes in the same direction, and a central globe, which retains its heat after the balls have parted with a large portion of theirs. In a word, the nebula is in this way transformed into a group of planets revolving round a central sun. The theory which supposes the solar system to have originated in the manner just sketched out, accounts so satisfactorily for many of the main characteristics of the planetary system that there is a very strong probability in favour of its being true. But for details on this head the reader must turn to works on astronomy ; we have to do with the theory here only so far as it concerns the earth. According to it, our globe was originally an intensely-heated, rotating mass of gas, and has assumed its present form by gradual cooling. Our task then will be first to lay before the reader all the facts about the constitution of the earth which can be gathered from observations made at or near the surface ; secondly, to see how far these facts fit in with and confirm the hypothesis of the nebular origin of the earth; and thirdly, assuming that hypothesis to be true and that the earth was once fluid, to inquire if we can form any estimate of the state to which the interior must by this time have been brought, whether any portion still remains fluid or whether solidification has extended from surface to centre. We shall further have to ask whether we can form any probable conjecture as to the nature of the materials of which the inside of the earth is composed. Shape of the Earth. In order to get a proper notion of what is meant by the shape of the earth, it is necessary clearly to realize that even the very largest inequalities of its surface, the loftiest mountains and the deepest oceanic depressions, are very small indeed compared with the distance from the centre to the surface, and may be altogether ne- glected when we look at our globe as a whole. So small are they, that if we could take a journey into space and view the earth from a moderate distance, its outline would look as even and regular as that which the moon presents to us. In this broad sense it is well known that the earth may be readily proved to be globular in shape, and that more accurate investigations show it to be not an exact sphere, but to be flattened like an orange. The Fig. 227. question of the determination of the exact figure of the earth has engaged the attention of many mathematicians, and they have shown that the form which agrees best with the observed measurements is 2s 642 Geology. that of a solid generated by the revolution of a half ellipse, AJ3b, fig. 227, about its shortest diameter, Bb. The name given to such a solid is an oblate spheroid ; B and b are the poles, Bb is the polar axis, the circle described by A is the equator ; and if C be the middle point of Bb, AC is the equatorial radius or semi- axis. In the case of the earth AC is a little short of 4000 miles, BC between 13 and 14 miles less.* For an account of the methods used to determine the figure of the earth the reader may refer to Lockyer's Elementary Lessons in Astronomy, chap. viii. ; Airy's Popular Astronomy, Lecture II. ; Encyclopaedia Metropolitana, art. "Figure of the Earth;" Baily, Astronom. Soc. Memoirs, vol. viii. ; Sir H. James, Phil. Trans. 1856 (vol. cxlvi.) p. 607; Comparisons of Standards of Length, Ordnance Survey of Great Britain, Appendix ; Archd. Pratt, a Treatise on the Figure of the Earth, fourth ed. Mean Density of the Earth. The weight of the whole earth has been determined by several physical and astronomical con- siderations, and it has been found that our globe weighs between five and six times as much as an equal bulk of water. We express this by saying that the mean density of the earth is between 5 and 6. The rocks of the crust are on an average about two and a half times as heavy as water, so that the ratio of the mean density of the crust to the mean density of the whole earth lies between 5 to 10 and 5 to 12, or may be put at 5 to 11. It follows from this that the interior of the earth must contain matter far denser than that which forms the crust. We know nothing for certain about the way in which the materials of the earth are arranged, but an expression, due to Laplace, which will be given further on, represents very probably the law of the density of the interior. If we employ this expression to calculate the probable density at different depths, we shall find, taking the density of the surface to be 2 '5 Density at depth of 250 miles 3'1 500 3-8 1000 4-8 1500 6'4 2000 7'5 2500 8-5 3000 9-2 3500 9-6 4000 9-8 The densities of the principal metals are Gold, 19 '3; Lead, 11*3; Silver, 10 '5; Iron, 7 '8; so that the density half-way down is on this * Some geometers have thought that the results of observation can be best reconciled by supposing that the earth is not exactly a solid of revolution, and that the equator is not a circle but an ellipse whose longest diameter is between one and two miles longer than the shortest diameter. For a summary of their views see Nature, x. 160. Archdeacon Pratt has thrown great doubts on the necessity for such a supposition (see Figure of the Earth, fourth ed. 181). Sir W. Thomson supports it, Natural Philosophy, arts. 796, 797. Colonel Clarke's last views on this subject will be found in Phil. Mag. August 1878. The Present Physical Condition of the Earth. 643 supposition about that of Iron, the density at the centre less than that of Silver. Two explanations have been offered to account for the high mean density of the earth. It has been suggested that, far down below the surface, the enormous weight of the overlying rocks would alone suffice to compress the material of the interior and make it as dense as observation shows it to be. It is however an open question how far we can go on increasing the density of bodies by increasing the pressure to which they are subjected. Experiments, as far as they have gone, seem to show that as the pressure is increased the density increases, but at a rate that constantly grows less and less. It is therefore possible that the effect of additional pressure in rendering a body more dense may become less and less till a point of approximate maximum density is reached, and that beyond that no increase in the pressure will add sensibly to the density. Those who hold this view account for the high mean density of the earth by supposing that the interior contains a much larger percentage of heavy metals than the crust. On the subject of the earth's density the reader may consult Lockyer's Elementary Lessons in Astronomy, arts. 634-637 ; Airy's Popular Astronomy, Lecture VI. ; Maskelyne, Phil. Trans. 1775, p. 500; Cavendish, Phil. Trans. 1798, p. 469; Baily, Astronom. Soc. Monthly Notices, iv. 96 ; Phil. Mag. xxi. (1842) 111 ; Sir H. James, Phil. Trans. 1856, p. 591. Internal Temperature of the Earth. The temperature of the surface of the earth varies according to the time of day and the seasons ; as we descend below the surface we find the oscillations due to these causes to grow less and less, and at last we reach a point where they cease to make themselves felt and the temperature of the rock is practically constant. A surface passing through all the points thus determined is called ike stratum of invariable temperature; its depth increases on the whole from the equator to the poles, but many local variations are caused by circumstances such as unequal conduct- ing power of the surface rock, and for this reason the depth of the invariable stratum does not follow any fixed law from place to place. At Greenwich it is found at a depth of 50 feet, and the temperature of the earth is there 49 '5 F., or one degree higher than the mean temperature of the air. When we pass below the stratum of invariable temperature, it has been found, wherever observations have been made, that the deeper we go the hotter does the earth become. The rate of increase deter- mined in various cases varies between very wide limits, perhaps about 1 F. for every 60 feet of descent will be about the average of all the observed rates. The depth of the deepest point whose temperature has been noted falls considerably short of a mile, and observation therefore merely justifies us in saying that, for the moderate depths to which we have been able to penetrate, the temperature increases as we descend. It has been thought that some observations point to the conclusion that the rate of increase in underground temperature decreases with the depth. That is to say, supposing the rise is on an 644 Geology. average 1 F. for every 60 feet near the surface, then at considerable depths we shall have to go down more than 60 feet to find the same rise, and the deeper we are the greater will be the additional depth we shall have to go down .in order to obtain an increase of 1 F. If this were so, we might at last reach a depth below which the temperature was practically constant. Such a result is theoretically not impro- bable, but it is extremely doubtful whether it can be said to have been as yet established by observation. An attempt made by Herr Hein- rich to prove this point from the observations made on the deep bore- hole at Sperenberg (Neues Jahrbuch, 1876, p. 716) has been spoken of as conclusive, but it is altogether based on a fallacy ; see Reports of the British Association Committee on Underground Temperature for 1876 and 1878. The subject is discussed with great acumen by the Rev. 0. Fisher in his " Physics of the Earth's Crust," chap. i. The reader may refer for details to Phillips, Phil. Mag. v. 446 (1834); Forbes, Trans. Royal Soc. of Edinburgh, xvi. 189 (1846); Angstrom, Upsala Nov. Act. Soc. Sci. i. 147 (1851) ; Hopkins, Phil. Trans. 1857, p. 805; Hull, Proceed. Royal Soc. xviii. 175 (1870), Quart. Journ. Sci. v. 14 (1868); Reports of the British Association Committee on Underground Temperature, 1868-1880; Sir W. Thom- son, Trans. Royal Soc. of Edinburgh, xxii. 405 (1860) ; J. D. Everett, ibid. xxii. 429 (1861), xxiii. 21 (1861), Edinburgh New Phil. Journ. xiv. 19 (1861), Reports of British Assoc. 1859, Trans. Sect. 245; Silliman's Journal, xxxv. 17 (1863), Proceed. Belfast Nat. Hist, and Phil. Soc. 1873-1874, p. 41; Greenwich Observations, 1860, p. cxciii ; Dunker, Uber die mb'glichst fehlerfrie Ermittelung der Warme des Innern der Erde, etc., Neues Jahrbuch, 1877, p. 590. An abstract of Bunker's conclusions is given in the British Association Report on Underground Temperature for 1876, and in "Nature," xv. 240. Inferences from the foregoing Facts. Such being the facts we gather from observations at the surface, we have next to see how far they are in accordance with the hypothesis that the earth has assumed its present condition by cooling down from an intensely- heated gaseous or fluid state. It is of course open to any one to maintain that the earth came into being just as it is now, with the exception of those surface modifica- tions which geology shows have been for a long time and are now going on ; but the supporters of such a view, if there be any, will have to get over several very ugly objections. First, with regard to tempera- ture, has it always been the same as now 1 In that case, since heat is constantly passing away by radiation, there must be some means of making good the loss and keeping the interior at a high temperature. No adequate means of bringing about this adjustment has yet been suggested. But if we suppose that the earth was once far more highly heated than now, we can understand that the inside must be hotter than the surface, because the heat passes off from the latter by radia- tion and from the former by conduction through materials of very low conducting power. The only reasonable explanation then which has been offered of the cause of internal heat is, that the earth is, and The Present Physical Condition of the Earth. 645 always lias been, a cooling globe, which is exactly what the nebular hypothesis supposes to be the case. Again, with regard to shape. If any hold that the present figure is original, they are bound to give reasons why it is a spheroid and not a sphere, and why, of the innumerable spheroids possible, a particular one has been chosen rather than any other. No possible reason can be assigned for the preference ; we can see no useful end that was to be served by giving the earth exactly its present ellipticity, or any possible harm that w r ould result from its being more or less elliptical. But if the earth has consolidated from a fluid state we can show that it is mechanically possible first for it to take the shape of a spheroid of small eccentricity, and secondly for that spheroid to have the particular ellipticity which measurement shows the earth to possess. For those who wish to know the grounds on which these statements are based, we offer a short outline of the mathematical treatment of the problem. Investigation of the Figure of the Earth on the Hypothesis of its original Fluidity. For the benefit of non- mathematical readers we give first the following definitions. If the various particles of a rotating mass of fluid do not move about among themselves, that is if the fluid rotates as a whole, it is said to be " in equilibrium." The surface that bounds a rotating mass of fluid which is in equilibrium, is called a "surface of equilibrium." A surface pass- ing through all the points at which the pressure is the same, is called a " surface of equal pressure." In a rotating mass of fluid in equilibrium surfaces of equal pressure are also surfaces of equal density. The problem which we now have to attack is this. Given a body of hetero- geneous fluid, acted on by no forces except the mutual attraction of its particles according to the law of gravitation, which has the same mass and volume as the earth and rotates with uniform angular velocity around a fixed axis in twenty-four hours, to ascertain the shape which it will take. When we apply the ordinary methods to the solution of this problem we see immediately that we must know more than the mass, volume, and velocity of rotation of the fluid before we can arrive at any result. Indeed we do not require mathematics to tell us this ; common-sense shows that much must depend on the way in which the denser and lighter layers are distributed ; and as many such arrange- ments are conceivable, it seems likely that the surface might assume different shapes according as the fluid matter was arranged in different w r ays, in short that the surface of equilibrium for the same fluid mass will vary with the distribution of the matter in the interior. Mathe- matical investigation confirms this view and tells us what it is that we must know in addition to mass, volume, and velocity of rotation before we can decide what particular surface of equilibrium will bound our fluid. And the additional information necessary is this. We must know the density and temperature at every point in the interior, and to get at this we must know what will be the density of the particular fluid we are considering under a given pressure and at a given temperature. If we had this additional information, we might in some cases arrive at a solution ; but even with it, in the majority of cases the mathe- 646 Geology. niatical difficulties that stand in our road are too formidable to allow of our working our way through them to a conclusion. In the case of the earth we have not got even the requisite data ; we do not know the material of which the interior is composed ; and, if we did, we should be no better off, for we could not say what would be the density of that material under the pressures and at the temperatures that exist within. A general solution of the problem is therefore utterly out of the question, but many of the mathematical difficulties can be in a measure evaded by some such method as this. We first ascertain by appro- priate tests whether the body we are dealing with is bounded by a surface of equilibrium. If it is, it is possible that it may have acquired its shape when in a fluid state. Assume that this is so and apply the mathematical formulas to determine as many points as we can about the constitution of the body, its exact shape, and so on. Then see if the numerical values which we obtain by calculation agree with those furnished by observation. If for instance our calculations show that the body would have a certain ellipticity, see if this agrees or nearly agrees with the ellipticity which measurements show the body to possess. We may have to make further assumptions as we go on, but if there is no physical improbability in the assumptions themselves, and if the results they lead us to agree with those of observations, there is a certain amount of evidence in their favour. The fewer the number of assumptions and the larger the number of coincidences between the results of observation and calculation, the stronger will this evidence be, and the greater will be the probability that the body we are con- sidering has obtained its shape by consolidating from a fluid shape. To apply this method to the case of the earth we first ascertain that an oblate spheroid of small eccentricity such as that which bounds the earth is a possible surface of equilibrium. We then assume that the earth in cooling took a shape of this class. In this case we find that the surfaces of equal pressure and density are concentric spheroids of small ellipticity with the axes of rotation for a common axis, and that the ellipticity of the successive surfaces of equal pressure decreases from the surface to the centre. Further that the density increases along any radius from the surface to the centre. This last result is not contradicted by observation, for that shows that the interior of the earth is denser than the surface. Our next step is to ascertain the exact amount of the ellipticity, but before we can do this we must know the rate at which the density increases towards the centre. This we do not know, and a further assumption becomes necessary. The assumption by which Laplace was enabled to calculate a value of the ellipticity was this, that the liquid of which the earth was composed was such that, when the pressure upon it is increased, the increase in pressure varies as the increase in the square of the density. Very little is known about the compressibility of liquids, so that we cannot have much to say either for or against the probability of this assumption being true, but the following points may be urged in its favour. In the case of those liquids whose fluidity is nearly perfect, that is in which there is scarcely The Present Physical Condition of the Earth. 647 any friction between the particles, experiments, as far as they go, seem to show that the increase in density varies as the increase in pressure ; if on increasing the pressure by a certain amount a certain addition is made to the density, then, if the pressure be increased by twice that amount, the addition to the density will be twice as great as before, arid so on. With Laplace's law if we wish to double the addition to the density we must increase the addition to the pressure fourfold. But it is likely that some such law as this is nearer the truth in the case of the material of the earth at the time of consolidation than the relation which holds good in the case of a perfect fluid. That material was probably pasty and viscous, so that there was a good deal of friction between its particles, and therefore the force requisite to bring them closer together would be greater than in the case of a perfect fluid. We must also bear in mind that in the experiments spoken of the pressures were far smaller than those which exist deep down in the earth. It is not unlikely that the increase of pressure necessary to produce a certain increase of density will be far larger when the pressure and therefore the density is great than at low pressures and low densities. Assuming Laplace's law to hold good, the following result is obtained. If we denote by D Y the density at every point on a surface of equal density whose semi-axis is r, then where A and g are constants, that is the same for all surfaces of equal density. Now are there any quantities connected with the earth whose numerical values depend on the way in which the density of the interior is distributed 1 ? If there are, we may obviously check our assumptions by calculating the values of these quantities from the mathematical formulae and seeing how far their calculated and observed values agree. There are four quantities available for this purpose; the Mass, the Ratio of the Surface density to the Mean density, the Ellipticity, and the annual amount of a certain motion of the earth's axis known as Precession, the nature of which will be explained by- and-by (p. 656). We can frame by the aid of Laplace's assumption mathematical expressions for these four quantities, but they will all contain A and g. Before then we can calculate the numerical values of these quantities we must know the numerical values of A and g. "We accordingly proceed thus. We take the expressions for any two of the above-named four quantities, say the mass and density-ratio, and put them equal to the numerical values of the mass and density- ratio which have been obtained by observation. This gives us two equations, and from them we can find the numerical values of A and g. Then we put these values in the place of A and g in the expressions for the ellipticity and precession, and thus obtain the calculated values of the ellipticity and precession. The calculated are then com- pared with the observed values. Or we may use any other pair of the four quantities to determine A and g, and then calculate the values of the remaining pair. The mass of the earth is accurately known and we may therefore 648 Geology. safely take the mass as one of the pair of quantities to be used to find A and g. Suppose for the second we take the ratio of the surface density to the mean density. The values thus obtained for the ellip- ticity and the precession agree very closely with their observed values. A double coincidence like this looks reassuring, but a little reflection will show that such a feeling is premature. Can we be said to know the ratio of the superficial to the mean density with accuracy 1 Hardly. The surface is composed of such a large variety of rocks that it must be very difficult to arrive at a fair average value of its density ; and the values of the mean density, as determined by different methods, lie between 5'25 and 6 '565. In our calculations 2 '75 is taken as a rough estimate of the surface density, and for the mean density the average of its different values is used, excluding the last which is considerably higher than all the rest. Such estimates are too vague to allow of any verification that depends on them being looked upon as of much value. It is to be feared then that we cannot employ the ratio of the superficial to the mean density as the second quantity in determining A and g, and we may now do one of two things. We may use the ellipticity as the second quantity for finding A and g, and then calculate the value of the precession. Doing this, the calculated and observed values of the precession agree satisfactorily. Our other course is to use the precession as the second quantity for determining A and g, and see what this gives for the ellipticity. The ellipticity then comes out very near its observed value. Here is some amount of support for the fluid hypothesis, but what does it really amount to 1 We have made three assumptions : first, that the earth was originally fluid ; second, that on consolidating it assumed one special form out of all the forms of equilibrium possible ; and thirdly, that the fluid of which it was composed obeyed Laplace's law. In support of this triple assumption we cannot be said to have much more than one coincidence between the results of calculation and observation. And this coincidence does not lend so much support to the fluid hypothesis as might at first sight appear, for before we can lay much stress .upon it we ought to show that an equally close coincidence could not be arrived at in any other way. Now coinci- dences equally close can be obtained on the supposition that the interior density follows laws, not in themselves improbable, altogether different from that of Laplace. The general result seems to be that mathematical investigation shows that it is perfectly possible that the earth may have acquired its present shape by consolidation from a fluid state, and furnishes a certain, but not a very large, amount of evidence in favour of this hypothesis. The further question arises, Could the earth have acquired its present shape in any other way than by consolidating from a fluid state ? There is one way in which it seems just possible that it might. The constant action of denudation tends to reduce the material of the surface of the earth to a loose and movable state, and in consequence of rotation this movable stuff would tend to arrange itself in such a The Present Physical Condition of the Earth. 649 way that the surface would be a surface of equilibrium. Whatever then had been the original shape of the earth, the degrading and transporting action of denudation would have a tendency to make the surface assume in the course of time the shape of some surface of equilibrium, and this surface would be protuberant at the equator. But the chances are enormously against a surface formed in this way being of the particular form which the earth possesses. The evidence in favour of the fluid hypothesis cannot be said to be overwhelming ; but it is almost infinitely greater than the evidence in favour of the hypothesis just discussed. We may fairly say then that the hypothesis of the original fluidity of the earth is the only hypothesis yet propounded which furnishes a satisfactory explanation of the origin of the earth's figure, and that, till some better explanation is offered, we may accept it as a probable provisional hypothesis. It is also in its favour that it does not stand on its own legs only, but is really part and parcel of the far wider nebular hypothesis, on behalf of which independent arguments might be urged. The following points also tell to a certain extent in favour of the fluid hypothesis. It may be made to give a reason for the high internal temperature and density ; and we shall see shortly that it leads us up to a reasonable explanation of the way in which rocks have been displaced from the positions in which they were laid down. The elaborate investigations of Mr. Q. H. Darwin have supplied additional evidence in favour of the nebular origin of the solar system. The case specially discussed is that of the earth and the moon, but the method and principles apply equally well to the general problem of the evolution of the solar system from a nebula. The most important addition made to the working out of this hypothesis consists in taking into account the effects of Tidal friction. The mutual attractions of the heavenly bodies raise tides in each other, tides in the ocean if the body has cooled down to a solid state and has water on its surface, tides in the mass of the body itself if it is still fluid or pasty. The friction of the tides slowly diminishes the rate of rotation. This general truth was clearly realized by Kant ; Mr. Darwin has traced its results from the time when the moon \vas shed off from the earth down to the present day, and has arrived at results which agree exactly with the present state of the earth-moon system. His method further yields reasonable explanations of some peculiarities in the case of other planets and their satellites. Such coincidences between the results of calculation and observation furnish evidence in favour of the truth of the hypothesis. One weak point in the original hypothesis has also been removed. There were mechanical difficulties in the notion that the nebula cast off a ring and that the ring collected into a ball. Mr. Darwin has suggested a more likely way in which the separation of the moon from the earth was brought about.* * The original papers are Phil. Trans. 1879, part 1, p. 1 ; 1879, part 2, pp. 447, 539 ; 1880, part 2, p. 713 ; 1881, part 2, p. 491. For a general summary of the results see Dr. Ball, Nature, xxv. (1881) 79, 103. 650 Geology. Mr. Darwin's results bear upon sundry geological questions : they are mentioned here because anything which strengthens the nebular hypothesis generally, gives support to the hypothesis of the earth's original fluidity, which is only a portion of the wider speculation. Present State of the Earth's Interior. Seeing then that it is likely that the earth was once wholly fluid, our next inquiry will be whether any part of it is still in that state, and if so, how much ? Doctrine of a Thin Crust. A very offhand solution of this question was at one time thought sufficient. It had been found that for small distances below the surface the earth grew hotter the deeper we got into it ; and if the heat went on increasing at the same rate, it was easy to see that at points not very remote from the surface a tem- perature must exist which would be quite sufficient at atmospheric pressure to melt the most refractory substances. It was therefore maintained that the interior of the earth must be necessarily in a state of fusion, and that the only supposition reconcilable with the known increase of heat downwards was that there was an outside solid crust not many miles thick, while below that the earth consisted of melted matter down to its centre. To explain this state of things, it was supposed that the solidification of the earth began at the outside, spread slowly downwards, and had not yet extended to any great depth. It was believed that volcanoes drew their lava from the great internal reservoir of molten matter, and that the phenomena of up- heaval and the displacements of the stratified rocks were caused by upswellings of portions of the seething mass.* The doctrine that the earth consists of a thin crust and a molten interior was at one time very generally accepted, and it is by no means certain yet that it is altogether false. It has however been opposed on various grounds, and some of the objections, it must be allowed, have considerable weight. It is easy to see that the arguments by which it was originally supported did not take into account several considera- tions, which might possibly modify its conclusions very seriously ; and other objections have been raised to it on mechanical grounds. We must now consider the arguments of those who oppose this view. Sir W. Thomson's Views on Underground Tempera- ture. In the first place the doctrine of a thin crust involves the assumption that the temperature continues to increase to all depths at the same rate as had been observed near the surface. Sir W. Thomson has shown that it is perfectly possible that this may not be the case. His mathematical investigations lead him to the belief that the tem- perature would increase at the rate of 1 F. for every 51 feet down to a depth of 100,000 feet or so, but that below that depth the rate of increase per foot would begin to diminish sensibly. At 400,000 feet the rate would be 1 for 141 feet; at 800,000 1 for 2550 feet, and so on in a rapidly-diminishing ratio. Such he thinks is the probable distribution of temperature within the earth down to a depth of 100 miles; below that depth the whole mass, whether liquid or * For a summary of these views see Cordier, Edinburgh New Phil. Journ. January 1828, p. 273 ; Paris Mem. Acad. Sci. vii. 473 (1827). The Present Physical Condition of the Earth. 65 1 solid, is probably at, or nearly at, the proper melting-point for the pressure at each depth. Sir W. Thomson has assumed in this investi- gation that no crust would be formed till the whole earth had cooled to a uniform temperature of 7000 F. (which he takes to be about the average melting-point of rock), that is to say, till the whole earth was just on the point of solidification. This and some other assump- tions perhaps detract from the value of the result, still the investi- gation is of great importance, as showing the possibility of a state of things very different from that implied by the doctrine of a thin crust.* Effect of Pressure on the Fusing-point. Another over- sight was committed by the believers in a thin crust in not taking into account the possible effects of pressure. Even supposing the surface rate of increase of temperature to be continued to all depths, yet pres- sure would increase at the same time, and it is perfectly possible that, under great pressure, substances may remain solid at temperatures far higher than would suffice to melt them at the surface. If the power of pressure to keep bodies solid be greater than the power of heat to melt them, the earth might be solid to the core even though the tem- perature continued to increase to all depths as fast as at the surface. This question of the effect of pressure on the fusing-point is one that will come before us again shortly, and we may conveniently discuss it here. We can safely say that if a body expands in melting, it is a priori likely that increase of pressure will raise its fusing-point. For in this case melting means forcing the molecules apart, and the greater the pressure the greater will be the resistance to be overcome, and the larger the amount of energy needed to do this work. Similarly if a body contracts in melting, as ice does, increase of pressure ought to lower the melting-point. Experiment has shown that in the case of ice this is so. Mr. Hopkins made experiments on some bodies which expand on melting with the view of determining whether their melting- points were raised by pressure. t In certain metallic alloys he failed to detect any elevation of the melting-point under increased pressure, but this may well be because the pressures at his command were not great enough to cause a sensible rise in the temperature of fusion. The table given below shows that in other substances a measurable rise was produced. Pressure in Ibs. to the square inch. Fusing-points Spermaceti. Fusing-points of Wax. Fusing-points of Stearine. Fusing-points Sulphur. Atmospheric. 7,790 11,880 124 140 176-5 148-5 166-5 176-5 138 155 165 225 275-5 285 * Trans. Royal Soc. of Edinburgh, xxiii. part i. p. 157 ; Thomson and Tait, Natural Philosophy, p. 689. See also Professor T. Milne, Geol. Mag. [21 vii. (1880) 99. t Keport of British Association, 1854, Trans, of Sections, p. 57. 65 2 Geology. These results may be put under the following form. Between 1 and 530 Atmospheres. Between 530 and 880 Atmospheres. The Fusing-point is raised at the rate of Spermaceti . . Wax .... Stearine . . . Sulphur . . . 030 034 032 095 131 036 036 034 For 1 Atmosphere. >? ') ? ? > i Perhaps the most valuable deduction that can be made from these figures is that the rate at which the fusing-point is raised is slow. But these experiments do not help us much towards determining the probable amount by which the fusing-point of rocks will be raised by pressure : for the rate varies very much ; the substances experimented on are few in number ; and they are very different from those which compose the mass of the earth. And even if from experiments made on rocks themselves we were able to deduce an empirical formula giv- ing the fusing-point in terms of the pressure, and if the constants were determined for different kinds of rocks, still then it would not be safe to apply the formula to pressures and temperatures beyond the limits of the experiments, and it would be the height of rashness to assume that it would hold good for the enormous pressures and probably very high temperatures which exist deep down in the earth. Nor do we find ourselves any better off when we turn from experi- ment to abstract physical reasoning. If we put t = the fusing-point under pressure p expressed in absolute tempera- ture. t -f A< = the fusing-point under pressure p -f A^. v = volume of the unit of weight in the fluid state at temperature t. v - Aw = volume of the unit of weight in the solid state at tempera- ture t. L = the latent heat. Then r- measures the rate at which the fusing-point is raised by increase of pressure. It will increase with an increase in the change of volume on solidification (Aw) ; also it will be the greater the higher the temperature of fusion ; and the less the specific heat. In the case of water Aw is 8 per cent, of the volume in the solid state, and the freezing-point is lowered '013 F. for an additional atmosphere of pressure. In the case of rocks Aw is probably very much less than for water;* the temperature of fusion (t) on the other hand is very much * According to Bischof Granite contracts 25 per cent., and Basalt 10 per cent, per unit of volume in passing from the fluid to the crystallized state, but his method of conducting the experiments from which these results were deduced The Present Physical Condition of the Earth. 653 higher; but till we know something for certain about the value of Ay for rocks, it is impossible to say whether AvJ will be greater for rock than for water or the reverse. The latent heat of rocks is not known ; in all substances for which it has been determined, the latent heat is less than in the case of water, and hence analogy would lead us to infer that the latent heat of rock is less than that of water, but how much less we cannot even conjecture. It is also most probable that the latent heat will not be constant, but will vary with temperature and pressure. The formula shows that the rate of rise of the fusing-point will probably be slower under high than under moderate pressures, for Az; w r ill probably decrease as the density is increased by increasing pres- sure. Mr. Hopkins' experiments show that this is the case with Sulphur : the tendency in Wax and Stearine is slightly, and in Spermaceti very decidedly, in the opposite direction, but this may be because the experiments were not carried far enough to reach the point where a decrease in rate begins. All we can say on the subject amounts to no more than this. At great depths below the surface there is heat tending to produce and pressure tending to prevent fusion ; but our knowledge is not sufficient to enable us to tell which of the two prevails at any given depth. We are equally in the dark as to the effect of high temperature in altering the conducting power and specific heat of rocks. But even in a case of such profound darkness it is not altogether unprofitable to put in a tangible form the results of the most probable hypotheses, and this is attempted in fig. 228. In this figure lines measured along Ox represent depths below the surface on a scale of 40 miles to the inch. The curves OP, OT, AF, show the pressure, temperature, and fusing-points at different depths in the following way. Suppose that ON represents a depth of 80 miles ; draw a line through N perpendicular to Ox cutting the curves in D, C, E ; then ND represents the pressure at a depth of 80 miles on a scale of 40,000 atmospheres to an inch ; NC represents the underground temperature, and NE represents the temperature of fusion at the same depth on a scale of 2000 F. to an inch. AB is a line representing a temperature of 4000 F., which we may take to be the average fusing-point of rock at the surface. EG will evidently represent the amount by which the fusing-point is raised by pressure at a depth of 80 miles. OP has been drawn on the assumption that the underground pressure follows Laplace's law; it bends up with a moderate slope near the surface where the increase in pressure is gradual, but becomes steeper as the depth increases because the pressure increases rapidly with the depth. The curve of fusing-points, AF, will bend up like the curve of does not inspire much confidence in his figures. Some experiments of the late David Forbes, though not themselves free from objection, seem to give from 1'25 to 3 per cent, as much more probable values for the contraction per unit of volume of both acid and basic silicates (Chemical News, October 23, 1868) : this is equivalent to from '5 to 1 "2 per cent, for the unit of weight. 654 Geology. pressures, but far less rapidly because the fusing-point is raised by pressure at a very slow rate. We can for reasons already given make only the roughest guess at the shape of this curve, but here is an attempt to approximate to it. If we take David Forbes' experiments as our guide (note, p. 653), the average value of A^ for rock is -85, or one-tenth its value for water. The absolute freezing-point of water is 460 + 32 = 492 F. ; the absolute melting-point of rock we take to be 4460 F., say ten times the temperature at which water freezes. These estimates will make Av.t about the same for rock as for water. Hence whether r- is greater or less for rock than for water, will Fig. 228. CURVES SHOWING THE POSSIBLE VARIATIONS OF TEMPERATURE, PRESSURE, AND FUSING-POINT WITHIN THE EARTH. depend upon L. L is probably less for rock than for water, hence -r- is probably greater for rock than for water. If therefore we construct a curve AW on the supposition that the fusing-point of rock rises at the same rate as it is lowered in the case of water, AF will probably at first bend up faster than A W. But probably the rate at which the fusing-point is raised decreases as the pressure in- creases, and therefore when we reach a sufficient depth, we must make The Present Physical Condition of the Earth. 65 5 our curve of fusing-points flatten; very likely when we reach great depths the curve will become almost parallel to Ox. We now come to the curve of temperatures. If the temperature continues to increase to all depths at the same rate as at the surface, this would be a straight line OQ. If this cut AF in II and if we draw HM perpendicular to Ox, then below the depth M the under- ground temperature is greater than the temperature of fusion and the larger part of the interior must be in a molten state. But it is far more likely that the rate at which the underground temperature in- creases diminishes with the depth, and that the curve of temperatures has some such shape as AT. The curve in the figure is that given by Sir W. Thomson in his paper on the Secular Cooling of the Earth (p. 650). With such a curve it is possible that AF may be wholly above OT, in which case the whole earth will be solid. Or AF may, say in the neighbourhood of the point (7, dip below OT for some distance and afterwards rise above it, in which case there will be a shell of fused matter between a solid crust where the temperature is too low for fusion and a solid nucleus where the pressure is high enough to pre- vent fusion. Again even if AF lies altogether above OT, they may run near to one another for a certain distance as at (7 : then if anything cause a decrease in the pressure, AF will sink, a portion of it may fall below OT, and the shell corresponding to this depth may pass into a state of fusion. If the pressure afterwards increase, this shell will be solidified. Various Modes of Consolidation possible. If we are to have a thin crust and a molten interior, the solidification of the earth must have begun at the surface ; but we do not know that this was the case, it may have begun at the centre. Whether solidification begins at the surface or at the centre, will depend on that relation between fusing-point and pressure about which we are unluckily ignorant. If during the time when the earth still retained a considerable degree of fluidity portions of the outside became solid, or increased in density owing to loss of heat by radiation, they would sink down into the still fluid mass below. If as they approached the centre the increased pressure had a greater effect in preventing them from being fused than the increased temperature had in promoting their fusion, they would retain their solidity, and thus a solid nucleus would accumulate round the centre. This process would go on till the fluid portion had so far cooled down that it was too pasty to allow of any hardened portions of the surface sinking through it. The external half-fluid shell would then begin to cool by conduc- tion, the superficial part would part with its heat most rapidly, and since none of it could descend, an external crust would be formed. In this way we might 'arrive at an earth with a solid crust and a solid nucleus, and a shell of imperfectly fluid matter between. The gradual loss of heat by conduction might subsequently cause the intermediate shell to solidify, and the earth might thus become solid from surface to centre. If, on -the other hand, the influence of pressure in preventing fusion 656 Geology. were less powerful than that of temperature in promoting it, the por- tions solidified at the surface would be again melted as they sank, and the earth would be kept fluid throughout till it reached the pasty state, when an external crust would begin to be formed. But even in this case we should not be able to say what is the present thickness of the crust unless we knew the original temperature, the time elapsed since it began to be formed, the rate of cooling, and sundry other things, about all of which we are hopelessly in the dark. It is per- fectly possible that the crust may not yet have attained any great thickness, or it may be that solidification has worked its way down to the centre. It appears then that assuming the earth to have come into its present condition by cooling from a melted state, it must have one of the three following constitutions. 1. It may consist of an external solid crust and the interior may be wholly fluid. 2. It may consist of a central solid nucleus and an external solid crust, separated by a shell of imperfectly fluid matter. 3. It may be solid throughout. But we are altogether unable, with our present knowledge, to decide by direct reasoning which of these three states represents most pro- bably the present constitution of the earth, or in the first and second cases to estimate the probable thickness of the crust. We must there- fore see if any light can be thrown on the question by indirect methods. Argument from Precession. Among the attempts made in this direction w r e must notice first the endeavours of the late Mr. W. Hopkins to determine what is the least possible thickness of the earth's crust that is consistent with the phenomena of precession and nutation. The actual calculations are exceedingly refined and intricate, but the following sketch will give an idea of his line of argument. The attractions of the sun and moon on the portions of the earth which bulge out at the equator are always producing slight displace- ments of the earth's axis, and these movements, combined with the earth's rotation, cause the axis to move in the following fashion. Take two straight rods, unite one end of one to one end of the other by a loose joint, and connect the other ends by a bit of string ; then hold one rod perpendicular to the plane of the ecliptic, and move the other round, keeping the string always tight ; the extremity of the second rod will describe a circle in space, and the motion of the rod itself will resemble in everything except speed the precession of the earth's axis. Nutation consists in small deviations first to one side and then to the other from the position which the axis would have if precession alone existed. It may be represented by supposing that the string in our illustration is slightly elastic, and keeps alternately lengthening and shortening itself a little. Under these circumstances the path of the end of the movable rod will be like the edge of a disc with a slightly crimped or wavy outline, and this is the character of the path actually described by the extremity of the earth's axis in space. It is important to note that these movements are due entirely to the spheroidal shape of the earth, and would not exist if it were a true sphere. In fig. 229 let The Present Physical Condition of the Earth. 657 A BCD be a section of the earth through its axis, AECF a circle whose diameter is the polar axis ; then, if we were to take away the pro- tuberant portions of which ABCE, ADCF are sections/ there would be no precession; if we were to take away the sphere AECF the precession would be very much larger than it actually is, because the sphere AECF, being rigidly attached to the protuberances, has to be carried round with them, and acts as a drag, preventing them from moving as fast as they would if they were not thus weighted. ISTow suppose that a portion, GHKL, of the central sphere is replaced by a mass of perfect fluid : the action of the sun and moon will not produce any precessional movement on this fluid, and as there will be no friction between it and the external shell, the latter will slip freely over it. Under these circumstances the amount of precession would be larger than for an earth solid throughout. Now the amount of precession calculated on the hypothesis that the earth is solid agrees very closely with the observed amount, and Mr. Hopkins set himself to work to determine how much of the interior could become fluid without impairing this agreement. On the supposition that the fluidity was perfect, and the change from the fluid to the solid part abrupt, he found that the thickness of the crust could not be less than one-fourth or one-fifth of the radius.* Argument from Rigidity. The subject may also be ap- proached in another way. The attractions of the sun and moon are greater on those parts of the earth that are nearer to them than on those which are farther off ; the solid part of the earth is for this reason subjected to unequal pull, which keeps it in a constant state of * Phil. Trans. 1839, p. 381; 1840, p. 193; 1842, p. 43; Report to British Association on Elevation and Earthquakes, 1847, pp. 45-55 ; Trans, of Cam- bridge Phil. Soc. vol. vi. part i. ; British Association, 1857, Trans, of Sections, p. 70. 658 Geology. strain. The same attractions will also tend to make the internal fluid portion bulge out on the side nearest the attracting body, and exert a pressure on the external shell tending to stretch the latter. We may try to determine what is the least thickness which will enable the crust of the earth to bear this strain and thrust, and prevent its being dragged or forced out of shape. This problem has been attacked by Sir W. Thomson.* He supposes the earth to consist of a spheroidal, homogeneous, slightly elastic shell, filled with incompressible fluid, the transition from the solid to the fluid portions being abrupt ; and on this hypothesis he has calculated to what extent the shell would be pulled out by the disturbing actions of the sun and moon. It is not likely that the amount of distortion would be large enough to be capable of detection by direct measurement, but it might make itself sensible by its effect on the tides. If the crust is drawn up in the same direction as the water that is, if there are tides in the solid part of the earth as well as in the ocean the height of the tide will be less than if the earth were perfectly rigid ; indeed, if the crust were only flexible enough, the surface of the earth and the water might rise together and there might be no tide at all. If then we knew what would be the height of the tide at a given spot, on the sup- position that the earth was perfectly rigid, and if we find the observed height to be less than this, we have a measure of the extent to which the solid part of the earth has been pulled out by the tide-generating influence. Xow Sir W. Thomson showed that, even if the spheroid were solid throughout and as rigid as glass, it would still give way to an extent, which would make the tides only about two-fifths as great as they would be if the earth were perfectly rigid ; if the rigidity were that of steel, the corresponding reduction would be to about two- thirds. He concludes that a thin crust could not possess the requisite amount of rigidity, and puts down the minimum thickness that would suffice to resist the distorting influence at 2000 or 2500 miles. But he goes on to say that the distribution of land and \vater alters the effects of the diurnal and semi-diurnal tides to an extent which no mathematical analysis can estimate, and that we cannot therefore use any deviations which they show from their calculated amount as measures of the earth's want of rigidity ; at the same time he thinks it very unlikely that these terrestrial disturbing causes can reduce these tides to two-fifths or two-thirds of the height they ought to have if the earth were perfectly rigid. He thinks however that the amount of the lunar fortnightly and of the semi-annual tide would not be affected to the same extent by the configuration of land and sea if observations of them were made at suitable points, and that they might be employed for the purpose of comparing the calculated and observed results. Unluckily no sufficient observations of these tides have yet been made. It would seem then that, even if the assump- tions by which Sir W. Thomson was enabled to deduce his results are justifiable, the observations necessary for applying these results to the * Phil. Trans, cliii. (1863) 573 ; Natural Philosophy, sects. 832-834, 847, 848 ; Nature, v. 223, 257. The Present Physical Condition of the Earth. 659 actual case of the earth are just those which have not been made ; and that till this defect is remedied, no conclusions can be arrived at. Sir W. Thomson has also investigated the effect of want of rigidity on the amount of precession and nutation. The observed amount agrees very closely with that obtained by calculation on the hypo- thesis that the earth is perfectly rigid. Any considerable want of rigidity would very materially alter the amount in most cases. But there are three arrangements under which the precession of an earth with a yielding crust would be approximately the same as for perfect rigidity. The first requires a compensating adjustment so very unlikely to be realized that we may dismiss it at once ; the second is incom- patible with a thin crust; the third is that the distortion should be very small in comparison with what it would be if the earth were fluid throughout. ^ow we will attempt to show by-and-by that if the transition from the solid to the fluid part of the earth is gradual and not sudden, this last condition may be satisfied. This argument from precession then is not conclusive. Objections to the preceding Arguments. Masterly as are the investigations just described from a mathematical point of view, it is to be feared that they have not contributed much towards the settlement of the question they were intended to decide. Some of the conditions they start with are so totally different from those of the actual case, that it is very questionable whether they can be fairly applied to the instance of the earth. It cannot be supposed that Mr. Hopkins' fundamental assumption of a solid crust separated by a hard-and-fast boundary from a perfectly fluid interior represents even approximately the internal condition of the earth. The transition from one to the other must be extremely gradual, the interior portion of the crust will grow more and more soft till it passes into a pasty viscous state, and this sticky matter will become more and more fluid as we approach the centre, and may possibly at considerable depths approach perfect fluidity. Professor Hennessy* and M. Delaunayt have both expressed their opinion that the results deduced by Mr. Hopkins are for this reason not applicable to the case of the earth, and the former has raised further objections to his method. Mr. Hopkins himself did not by any means overlook this want of agreement between the actual and assumed conditions, and endeavoured to show that in spite of it his conclusions would hold good. He says that if C be the centre, C/S any radius of the earth, A a point on that radius above which the earth is solid, B a point below which all is fluid, and AB the intermediate transitional portion, then if we take SA to be the thickness of the crust it will give the pre- cession too large, and if SB, too small \ but there will be some intermediate thickness which will give the right amount, and this he calls the Effective Thickness of the crust. It is the depth of this Effective Thickness which he has shown cannot be less than 1000 miles. All then which he has proved amounts to this, a shell of at least 1000 miles thick must participate in the precessional motion. * Phil. Trans, cxli. (1851) 495 ; Geol. Mag. viii. 216 ; Nature, iv. 182. t Comptes Rendus (July 13, 1868), Ixvii. 65 ; Geol. Mag. v. 507. 660 Geology. But this is a very different thing from saying that the whole 1000 miles must be solid : possibly only a very small portion might be in this condition and the rest in a more or less viscous state, and yet the whole be carried round very nearly as if it were all solid, because the friction between the particles of the pasty part prevents them being dragged one over the other. Different parts of the viscous mass would be from time to time compressed and extended, but it seems perfectly conceivable that for all practical purposes the solid and semifluid portions of the crust might hang together as a whole. The conclusions of Sir W. Thomson as to the Rigidity of the Earth have also been attacked by Professor Hennessy.* He has pointed out how seriously the hypothetical differ from the real conditions of the problem, and specially how the neglect of the pasty shell that must exist between the solid crust and the more fluid interior impairs the validity of the results. This pulpy stuff would act as a pad or buffer, and the work done by the disturbing action of the sun and moon on the internal portions of the earth, instead of being transmitted to the surface and altering its shape, would be used up "partly in producing small variations of density among the compressible strata of the nucleus, and partly in changing the shape of the yielding matter of the inner surface of the shell." By this means the deformation of the shell might be very small indeed, and the amount of precession the same as if the earth were solid throughout and perfectly rigid. Practically the observed amount of precession is rather less than it would be if the earth were perfectly rigid ; some small distortion of the crust is probably therefore produced, and this we may take as the measure of that portion of the interior work which has managed to penetrate the buffer and make itself felt in the solid crust. A calculation made by Archdeacon Pratt furnishes an illustration of Professor Hennessy's objection. t Starting with Mr. Hopkins' assumptions about the interior of the earth, he shows that the internal fluid nucleus will be pulled by the attraction of the sun and moon, and will exert a pressure against the crust which tends to increase the precession ; and that the effect of the want of rigidity in the crust will tend to decrease the precession. He then determines to what extent the surface must be elevated in order that these two modifying causes may destroy one another, if the crust be 800 miles thick, and this he finds to be 20 feet. A deformation of one-seventh of this amount would altogether abolish tides in the open ocean, and hence he concludes that the crust must be far thicker than 800 miles. Now if the internal fluid mass, instead of pressing against an unyielding crust, had a soft pad of semi- fluid matter to bury its nose in, we can readily imagine its energy might be consumed in compressing and pushing this aside, and no pressure might be exerted on the crust ; and if the crust, instead of being solid throughout, had a yielding lining, the work due to the external disturbing force might be expended on the lining, and give rise to no change in the shape of the surface. In such a case, as far as the causes mentioned are concerned, the precession would be unaffected. * Nature, v. 288. t Figure of the Earth, fourth edition, p. 135. The Present Physical Condition of the Earth. 66 1 It is further to be feared that the mathematical calculations of Sir W. Thomson's paper on the Rigidity of the Earth do not touch the actual case for this reason. Such experiments as have been made (pp. 511, 512) seem to show that rocks are not even imperfectly elastic but rather what is styled elastico- viscous. If a rock be subjected to a dis- torting force of sufficient magnitude, it will be slowly deformed, and as long as the force keeps acting, the deformation will go on increasing till the rock breaks ; but if the force decrease or be removed alto- gether before the rock has been strained up to the breaking-point, it does not necessarily follow that the rock will tend to return to its original shape. The molecules readjust themselves and the body becomes gradually reconciled to its new shape ; the force necessary to keep it in that shape grows less and less as time goes on, and at last no external constraint becomes necessary to prevent it returning towards its original figure, and it permanently sets into its new shape. A stick of sealing-wax may be made to illustrate this property ; if it be sup- ported at the two ends in a horizontal position with a small weight in the middle, and if the load be cautiously increased by small additions, the stick bends, and when the weights are taken off it retains its bent shape. It is therefore by no means certain that the crust of the earth would oscillate like an imperfectly elastic spheroid under the attractions of the sun and moon. Portions might be deformed, and, as the deforming forces decreased on account of a change in the positions of the attract- ing bodies, might retain the shape into which they had been drawn ; when the deforming forces came to act in the opposite direction, these parts might be pulled back to their original shape and keep that shape till they were again subjected to distorting influences. Instead of oscillation, there might be a slow dragging first in one direction and then in the opposite direction, so that the average shape would be unaltered, and the changes that did take place might be of a totally different character from oscillations. Professor Hennessy's Views. Professor Hennessy has at- tempted, in his paper in the Philosophical Transactions already quoted, a mathematical solution of the question now before us. It is impos- sible to convey any adequate notion of his way of handling the subject without more mathematics than are admissible here, but the following are his chief points. He objects at starting that all previous investigations had tacitly made an assumption that cannot be justified, namely, that the volume of the entire mass and the law of density of the earth have remained the same, or in other words, that the particles of the original fluid mass underwent no change of position during the process of solidifica- tion. He then considers what would be the order of events during solidi- fication. First he thinks there would be much chemical action. When the chemical affinities of the materials had been satisfied, the mass would probably be in a state approaching perfect fluidity, and circulation would go on, the portions that had grown denser by cooling descending and the lighter portions ascending, till the whole had 662 Geology. arranged itself in concentric shells whose density increased from the surface to the centre. When this state had been arrived at he thinks a surface crust would begin to be formed. Of course, when any piece on the outside had become solid, it would on account of its increased density tend to sink, but the three following causes would hinder its descent. 1st. Each stratum into which it descended would be denser than the one above. 2nd. Each stratum would have its density increased by the passage through it of cooler portions from above. 3rd. The descending portions would have their densities diminished by the increase in the temperature downwards.* Under these circumstances he thinks that though circulation would go on, it would be confined to the neighbourhood of the surface, and a crust might be formed ; below the crust would be a shell of imper- fectly fluid matter, and the interior might retain a high degree of fluidity. He is of opinion that it is impossible that consolidation can have begun at the centre, and that even supposing any accumulation of solid matter there ever did take place, it must necessarily be melted again. He believes that, even if from any cause the solid shell and fluid nucleus rotated at any time at different rates, yet that the friction between the two must be great enough to bring the motion of both to the same velocity of rotation. Finally he endeavours to determine the present thickness of the crust thus. He obtains an expression for gravity at the surface in terms of the radius and ellipticity of the fluid nucleus ; and finds that if this expression is to agree with the known law of the variation of gravity over the earth, the crust cannot be less than 18 nor more than 600 miles thick. This result however is at variance with the conclusions of Professor Stokes, who has shown that provided the surface of the earth is a spheroid of equilibrium of small eccentricity the law of the variation of gravity at the surface does not depend on the way in which the matter is distributed in the interior.! Mr. R. Mallet has arrived at conclusions similar to those of Professor Hennessy by a different line of reasoning ; J his views will be given more at length in Section IV. We cannot enter here any further into the question, but for addi- tional discussion of these moot points the reader may consult Professor Hennessy, London, Edinburgh, and Dublin Phil. Mag. third series, xxvii. 376 (1845); Journ. Geol. Soc. of Dublin, 1849; Proc. Royal Irish Acad. iv. 337 ; Archdeacon Pratt, Figure of the Earth, Nature, ii. 264, iv. 28, 141, 344; Geol. Mag. vii. 421; Nature, iv. 45, 182, 383; Professor Haughton, Trans. Irish Acad. xxii. part 1, p. 251; D. Forbes, Geol. Mag. viii. 162; P. Scrope, Geol. Mag. vi. 145; D'Archiac, Histoire des Progres de la Geologic, i. ; Thomson and Tait, Natural Philosophy, Appendix D ; Sir W. Thomson, Address to Mathe- * To these we may add that the escape of gases from the seething mass would blow out any portion that solidified into a spongy vesicular state, and it would therefore float even though its specific gravity were higher than that of the fluid on which it rested. t Cambridge Phil. Trans. 1849. % Phil. Trans, clxiii. (1873) 160. What the Interior of the EartJi is made of. 663 matical Section of British Association, 1876 ; Major-General Barnard, Smithsonian Contributions, vol. xix. No. 240. Summary. The only conclusion, if we can call it a conclusion, that it is safe to come to in the present state of our knowledge is that if we assume that the earth has cooled down from a state of igneous fusion, there is not evidence to decide whether any of it, and if any, how much of it, still retains its original liquidity. Some considerations of a general character may however be urged in favour of a thick crust. Though the mathematical discussion of the question cannot be looked upon as actually conclusive, it certainly tends to impress one very strongly with a belief in the great effective rigidity of the earth, and the more recent investigations of Mr. G. Darwin tend to confirm this view. The assumption by which he has arrived at complete agreement between the results of his calculation and those of observation assigns a degree of viscosity to the earth which is practically a high rigidity. The general stability of the surface would also seem to be incom- patible with a thin crust and a molten interior. Earthquakes and volcanic eruptions, on account of the fearful havoc which they occasion, appeal so forcibly to the imagination that we are apt to forget that they are exceptional phenomena, local and comparatively unfrequent. If there were a surging mass of molten lava everywhere not far beneath our feet, it seems likely that they would be more general and would recur oftener. The balance of evidence is not perhaps very great, but such as it is, it is in favour of a large part of the earth being solid. SECTION II. WHAT THE INTERIOR OF THE EARTH IS MADE OF. AVlien we come to the subject to be handled in this section there is one point and only one on which we can speak with anything approach- ing certainty. The mean density of the earth is about twice as great as the average density of the crust. The matter which composes the interior must therefore be heavier than the generality of the substances that make up the crust. It seems hardly possible that even the enor- mous pressures which exist towards the centre of the earth would be competent to double the density of surface rocks ; and we are therefore driven to the conclusion that inside the earth there is matter which, even at atmospheric pressure would be denser than average surface rock. The question then arises, Is this heavy stuff something different altogether from anything known on the surface ; or is it one of these heavy bodies, a metal or metals for instance, which occur in the crust only in comparatively small quantities 1 All the known evidence is against the first solution. Spectrum analysis, the study of meteorites, every method that enables us to gain an insight into the composition of the various bodies of our system and even of the universe at large, tell the same tale. The chemical elements that have been detected in the sun, stars, and other heavenly bodies, all of them exist on the earth, and nowhere has any one suc- ceeded in establishing the existence of an element which is absent from \\ , 664 Geology. the earth. Everything points to the conclusion that the composition of matter is everywhere the same. Why should the interior of our earth be the only exception to this rule 1 ? We therefore fall back on the conclusion that the heavy metals, or some of them, which form only an insignificant portion of the crust, exist in large quantity in the interior of the earth. The high density of the interior of the earth is an argument in favour of its having condensed from a gaseous state. There may have been a time when all its materials were volatilized by intense heat, and the vapours may have approached perfect gases in their molecular con- stitution. At that time, if there ever was such a time, they would follow the laws of a perfect gas and intermingle freely by diffusion. But as cooling went on and the vapours drew near to their point of condensation, gravity would gradually assert itself and prevail over diffusion ; the vapours of the heavy metals, which would be the first to condense, would be drawn towards the centre, while the lighter and less condensible material would still form a vaporous envelope outside. Oxygen, if there was such a substance in those days, would be among the light bodies that floated outside, and hence the metallic nucleus would be unoxidized. It has been conjectured on some such grounds as these that the earth contains an unoxidized core of heavy metals. It is possible that portions of this core may from time to time work their way up to the surface, and the universal diffusion of iron suggests that this metal predominates in the core ; the presence of large masses of iron in the interior, it has been further conjectured, may explain the earth's magnetism.* The Argument from Meteorites. It has been conjectured that some inkling as to the probable composition of the interior of the earth may be obtained by a study of meteorites. Daubree has classi- fied meteorites under the following heads. 1. Holosiderites, consisting entirely of metallic matter without any stony parts. The metal is Iron alloyed with Nickel, and may be described shortly as Meteoric Iron. Certain minerals are found in these meteorites which are not known on the surface of the earth and which probably could not exist in the presence of Oxygen, such as Troilite (FeS), Schreibersite, a Phosphide of Iron and Nickel contain- ing Magnesium, Daubreelite (FeS,Cr. 2 S 3 ), which differs from Chrome Iron Ore in containing Sulphur in the place of Oxygen. 2. Syssiderites, consisting of stony grains disseminated in a metallic paste. The metal is Meteoric Iron, the stony grains are in a majority of cases Olivine ; in some a Pyroxenic mineral ; in one case a mixture of Olivine, Pyroxene, and Chrome Iron Ore. 3. Sporadosiderites, consisting of metallic grains in a stony paste. * There is a strong temptation to amuse oneself with speculations of this nature, but, if we give way to it, we must carefully bear in mind how very little of solid basis they have to rest upon. The conditions we have to deal with are so totally different from anything that we can imitate experimentally, that any conjectures we may make are little better than fanciful dreams. Those who wish to pursue the subject further may refer to De la Beche, Researches in Theoretical Geology, chap. i. ; S terry Hunt, Quart. Journ. Geol. Soc. xv. 488 ; Geol. Mag. iv. 357, 426, 477, 525 ; v. 49, 106; David Forbes, ibid. v. 92, 106. -Iflfl. THE UNIVERS What the Interior of the Earth is made The grains are Meteoric Iron, Troilite, Chrome Iron Ore, paste contains Olivine, Pyroxene, and a mineral very near to Enstatite. 4. Cryptosiderites, stony meteorites in which there is a very small metallic element of Meteoric Iron. The stony part contains Olivine, Pyroxene, and Anorthite ; Pyrrhotine, Magnetite, Chrome Iron Ore, Apatite, and Sphene have also been detected in small quantities. 5. Aside?*ites, with no metallic element. They contain Magnesian Silicates, Oxides of Nickel, Cobalt, and Chromium, Magnetite and Chrome Iron Ore, and Pyrrhotine; and they are specially distin- guished by the presence of Hydrocarbons, Combined Water, and Soluble and Deliquescent Salts such as Ferrous Chloride and Calcium Chloride. Asiderites are the rarest ; Holosiderites rarer than the stony forms ; and by far the commonest are those Sporadosiderites in which the metallic portion ranges from 8 to 22 per cent. Meteorites are crystalline in texture, and when we compare them with the crystalline rocks of the earth's crust, we see that we must seek for their analogues among the basic and not among the acid members of that body of rocks. The one striking point in the com- position of meteorites is either the total absence of Oxygen, or the low degree of oxidation of their component minerals. No Quartz ; no highly-silicated Felspars, Felspars indeed rare altogether and only basic members of the group, such as Anorthite, when they do occur ; but in their place basic Magnesian Silicates, such as Olivine and Augite. The very general presence of Chromium, Nickel, and Cobalt is also noticeable. Some of the Cryptosiderites approach certain lavas of Etna and Iceland ; but it is among the ultra-basic members of the basic class, the Peridotites, that we find the nearest approach among terrestrial rocks to meteorites ; these rocks bear the closest resemblance to that division of the Sporadosiderites which includes the greater part of the known meteorites. In both Olivine is the most important mineral, the other predominant Silicates are Augite and Enstatite, while Chrome Iron Ore is a constant constituent. Now the Peridotites have the highest specific gravity (3'3 3'5) and the lowest percentage of Oxygen of the crystalline rocks, and on these grounds it has been assumed that they have risen from greater depths than any other lavas, and that they represent the composition of the earth's crust at these depths. Such an assumption is in the highest degree speculative, but, if we allow it, the case stands thus. Certain Cryptosiderites find their parallel in heavy lavas of the basic class that came from a considerable depth ; certain Sporadosiderites, poorer than these in Oxygen, are the analogues of ultra-basic lavas that came from greater depths still. If this be so, it is certainly not unreasonable to infer that the still less oxidized Syssiderites, and the Holosiderites that contain no Oxygen, will find their counterparts in shells of matter that lie still deeper in the earth than the level from which the ultra-basic rocks have risen ; in other words that very deep down unoxidized metallic matter prevails and that Iron is the pre- dominant metal. Some support is afforded to this view by the great masses of native 666 Geology. Iron discovered by Baron Nordenskicild at Ovifak on the southern shore of the island of Disco in Greenland. These masses consisted very largely of Meteoric Iron, and they resembled meteorites so closely in many respects that their discoverer and others maintained at first that they must have had a meteoric origin. But it is now generally admitted that they came from a subterranean source. The district was the scene in Miocene times of great volcanic activity, eruptive masses of Basalt are plentiful in the neighbourhood, and in a dyke hard by the spot where the large Iron masses were found about 100 Ibs. of lenticular-shaped masses of Meteoric Iron, from 3 to 4 inches thick, were extracted from the Basalt. Microscopic examination of the Basalt further shows that it contains a large amount of Iron finely disseminated through its mass. Fragments of a crust of Basalt were also found adhering to the outside of some of the masses. There can be scarcely a doubt that the Iron was brought up from below together with the Basalt. Daubree examined specimens of these masses and found that two might be described as Syssiderites and one as a Sporadosiderite ; they resembled Asiderites in containing Carbon, Ferrous Chloride, and Calcium Chloride. These Iron masses furnish an additional parallel between terrestrial lavas and meteorites, for we can now find a match not only for Crypto- siderites and Sporadosiderites, but for Syssiderites also among the substances which exist within the earth and are brought to the sur- face by volcanic action. With the assumption we have been hitherto going on, we should hold that this Iron came from a depth greater even than that at which the material of the ultra-basic lavas lies.* The so-called "Iron Ore" of Khode Island is a rock which has a bearing on the present subject. It is described by Mr. Wadsworthf as a Peridotite rich in Magnetite, and would seem to be a connecting- link between ordinary Peridolites and the Iron masses of Ovifak. The "Iron Ore" of Taberg in Sweden is a rock of a similar character. J The argument may be put under a form which involves less of speculation than the way in which it has been just stated. Certain of the stony meteorites can hardly be distinguished from some of the terrestrial lavas. Analogy would lead us to infer that the parallel does not stop here, but that the metallic meteorites also have their counterpart somewhere in the earth ; the Iron masses of Ovifak tend very strongly to substantiate this inference. It cannot be at the surface, for Iron cannot continue there in an unoxidized state ; if we are to find anywhere the analogues of the metallic meteorites, it must be at considerable depths. Two quite independent lines of argument then point to the conclu- sion that there are probably large quantities of the heavy metals, very * See Daubree, Geol. Experimental, 2 me partie ; Judd, Volcanoes, p. 312 et seq. t Bulletin of Museum of Comparative Zoology, Geol. Series, i. 183 ; Proceed- ings Boston Society of Natural History, xxi. (1881) 195. J A. Sjogren, Geol. Foren. Stockholm Forh. Bd. iii. 42. Cause of Upheaval and Contortion. 667 likely uncombined with Oxygen, in the earth's interior, arid that Iron forms a large portion of the metallic nucleus. Argument from Metallic Lodes. Ithasbeen further suggested that this hypothesis will explain why the ores in the deeper parts of lodes are scarcely ever compounds containing Oxygen but almost universally Sulphides. This would necessarily be the case if the lodes had been filled from below and had received their contents direct from the unoxidized metallic core. If Holosiderites do represent in any measure the composition of the central portions of the globe, it is somewhat in favour of this view that the minerals which they contain are several of them metallic Sulphides, and it is perfectly possible that some lodes may have been filled in this way. But we have already seen that the prevalence of Sulphides in lodes can be accounted for in another way (Chap. XII. Sec. III.). SECTION III. CAUSE OF UPHEAVAL AND CONTORTION. We have seen reason to believe that the uplifting of rocks laid down beneath water by which sea-bottoms have been converted into dry land, and the tilting, folding, contortion, and faulting which observa- tion shows these rocks to have undergone, are all due to some common cause. We have now to inquire how the force was generated which produced these movements. Reasons have already been given for believing that the disturbing force was, in very many cases at least, of the nature of a horizontal thrust. The phenomena of widespread and excessive contortion can be explained only on this supposition, and the lesser disturbances are so intimately connected with these grand movements, that it seems likely that both must be due to the same cause. Certain however of the observed displacements taken by them- selves are capable of explanation on the supposition that they were caused by a force acting from -within the earth vertically upwards ; and though it is extremely doubtful, for reasons already given, whether this method has been ever employed in nature, it is only fair that the reader should be put in possession of the hypotheses which have been started to show how this vertical up-thrust may be caused. Sense in which Elevation is used. It is somewhat unlucky that the words " upheaval " and " elevation " should have become so thoroughly rooted in geological nomenclature in reference to the move- ments of the earth's surface, that it is now scarcely possible to drop them. The first is decidedly objectionable, for it distinctly implies that the producing cause was a force acting vertically upwards, which it certainly was not in some cases and perhaps was in none. If ever we employ these terms, it must be understood that we use them simply to indicate the carrying up of rocks from a lower to a higher level, and not in the least to suggest the direction in which the force acted by which the movement was effected. General Structure of Mountain-chains. It is in moun- 668 Geology. tain-chains that we find the most marked and conspicuous results of those earth movements whose cause we are now in search of, and before we go any further, it will be desirable that the reader's ideas should be clear on two points : first what the arrangements of the rocks in a mountain-chain* is not like \ and secondly what it is like. In many books he will find the general section of a mountain-chain to be such as is shown in fig. 230. sg^Siig^ 4f&i >. '. i*^!^; 3 ~i C ^. 3 - 1. Granitic Rocks. 2. Foliated Schists. 3. Unaltered Rocks. Fig. 230. WHAT A SECTION ACROSS A MOUNTAIN-CHAIN is NOT LIKE. We see in this section a central nucleus of Granite, and from this the rocks dip away on either side in the same direction as the slope of the ground. The sketch gives the idea that the Granite has been driven up from below and has thrust aside the rocks on either flank. There is no mountain-chain known which has a section at all approach- ing this. The section in fig. 231, though it fails to convey any ade- quate notion of the amount of crumpling, inversion, and smashing that is very frequently met with, gives a much fairer general idea of the disposition of the rocks. The strata have not been simply bent up into a single boss, but have been folded, crumpled, and pushed over along a number of lines ranging roughly parallel to one another; and in many cases so far from the dip being outwards on either side of the range, it is directly in the opposite direction, the rocks plunge at high angles on both flanks into the hill, and a section of them shows something like the plaits of an open fan, the handle of which lies deep down in the centre of the mountain. Granite appears in several belts, but these show no signs of having been thrust up through the surrounding rocks. On the contrary, the rock shades off insensibly into foliated schists, and these melt away in unaltered rocks. The beautiful sec- tions across a very contorted part of the Alps in Heim's " Mechanismus der Gebirgsbildung," which are drawn on a true scale, bring vividly before the eye the enormous extent to which contortion and inversion has been carried in that mountain-chain ; great faults seem to be here altogether absent, and the disturbances have been caused by excessively sharp folding and tilting over of the folds. In the Geological Surveys of Pennsylvania arid Virginia the structure of the Appalachians is described by Professor Rogers ; here we have faults of enormous size in addition to intense folding. Mr. Hopkins' Theory. We have already mentioned Mr. Hopkins as one of the ablest supporters of the vertical up-thrust explanation. The machinery he employed for producing the elevatory force was a body of fluid matter like lava, in a cavity below the * Using the terra in the restricted sense applied to it in the preceding chapter. Cause of Upheaval and Contortion. 669 surface, which under the influence of increasing heat, gave off elastic gas, and the pressure of this gas was supposed to bend up the over- lying rock till it was strained to the breaking-point. There is very little evidence for the existence of the internal lakes of fused, gas-yielding matter required by this theory, but, waiving this objection, we have already pointed out what seems to be its weak points in Chapter XL p. 508. Mr. Hopkins' knowledge of the facts he attempted to explain seems to have been very im- perfect. He speaks for in- stance of anticlinal and syn- clinal lines occurring with alternations of rapid and oppo- site dips at intervals not exceeding a few miles. If he had said every hundred yards or so, it would have conveyed a far more correct notion of the real facts in many cases. His conviction too that all narrow, steep-sided valleys were rents in the crust led him sadly astray. His beauti- ful mathematical investigations have for these reasons not led at present to any useful re- sults ; but they may yet bear good fruit, and geologists will probably some day thank him for the attention he has given to the subject. Theory of Scrope and Babbage. Next comes a very ingenious speculation, originally suggested by Scrope,* and afterwards struck out in- dependently by Babbage. t By the deposition of sediment on sea - bottoms an area, which before the deposition began formed part of the earth's surface and had the temperature of the * Volcanoes, p. 271, note. t Proceedings Geol. Soc. ii. 74 ; Ninth Bridgewater Treatise, Note G, p. 209. 670 Geology. surface, is gradually buried under a cover, constantly increasing in thickness, which checks the escape of the heat from within. The temperature of the rocks of this area will therefore be raised, and the rocks themselves must expand. The resistance to expansion will be less in a vertical than in a horizontal direction, and hence the area will bulge up and cause elevation of the rocks resting upon it. It is evident that this explanation accounts only for vertical elevation, and makes no provision for contortion or any form of compression ; in fact the elevations produced by it would be bosses, the strata of which would be arranged in concentric domes, like that in fig. 230, which, as we have pointed out, is a structure found in no mountain-chain yet examined. Even then supposing some small local elevations may be due to this cause, it is quite inadequate to produce the larger disturb- ances which the crust- exhibits. There is another weak point about this theory. During the accumulation of many thick bodies of sedimentary rocks, the sea- bottom was gradually sinking. It makes no provision for this. It may explain vertical uplifting, but does not allow of subsidence during deposition. Theory of Sir J. Herschel. Sir J. Herschel* threw out the hint that the mere weight of a thick mass of sediment might cause the part of the crust on which it rested to bend down, and the portions on either side to swell up. This would require the crust to be remarkably thin and yielding, and till this is shown to be the case the explanation cannot be admitted. Several geologists have cited the great thickness reached by certain delta deposits, those of the Mississippi and Ganges for instance, as a fact strongly in favour of this hypothesis. It is a fact which proves the existence of an area of subsidence at the mouth of the river, but it is quite another question whether that subsidence was caused by the weight of the deposit. Several facts seem to show that this is not the case ; if it were, the subsidence should be continuous, but in several cases old land surfaces are found in the middle of thick delta deposits and these point to pauses in the sinking : what is more to the point is that the motion was not always downwards, upheaval appears to have alternated with subsidence, t There is another point to be taken into account. The pile of strata must reach a considerable thickness before it becomes heavy enough to bend even a very thin and pliable crust, and subsidence will be required to enable it to reach that thickness. There must have been subsidence to start with, it was at that time not produced by the weight of the deposit, and it seems reasonable to infer that whatever was the cause of it to begin with continued to be the cause of it all along. Intrusion of Granite. Another explanation of the cause of the elevation of mountain-chains, at one time very much in vogue See also Captain Hutton, Geol. Mag. x. 166, xi. 22 ; Nature, ix. 61 ; Rev. 0. Fisher, Geol. Mag. x. 248, xi. 60, 64. * Proceedings Geol. Soc. ii. 548, 596 ; Ninth Bridgewater Treatise, Note I, p. 225. t Lyell, Second Visit to the United States, vol. ii. chap, xxxiv. Cause of UpJieaval and Contortion. 67 1 among geologists, and adopted by the . late Mr. Scrope, is as follows. * In a very large number of cases a mountain-range has a central axis of Granite or some Plutonic rock. This central mass was supposed to have been forcibly intruded in a semifluid state from below, and to have shouldered off on either side the rocks through which it forced its way. The intrusive matter was supposed to have been driven up by local increase of heat, which caused it to expand till the rocks above were no longer able to hold it down. Following the views put forward by himself and Mr. Babbage, Mr. Scrope accounts for the accumulation of heat by the deposition of a thick mass of strata above the spot where it occurred. It is clear that this process would give rise to tilting, and result in the formation of a ridge from which the beds would dip away on either side, as in fig. 230, but it would not directly produce contortion. Mr. Scrope endeavours to show that lateral pressure would result indirectly in two ways : first the molten matter injected into fissures would if urged up w r ith sufficient force press laterally against the walls ; and secondly he thinks it possible that portions of the rocks which were shoved aside by the intruded mass might slide down its inclined flanks, and get crumpled up in the motion. In the first case the cause seems scarcely sufficient to do the work assigned to it.t The sort of motion required in the second case seems hardly likely to occur. This theory too fails to account for the violent contortions which not unfrequently are met with far away from any mass of Plutonic rock. Further, there is good reason to believe that the Granitic axes of mountain-chains are not always intrusive, but are in some cases parts of the very rocks that compose the flanks of the range converted by intense metamorphisrn into Granite. Laccolites. Mr. Gilbert in his "Geology of the Henry Mountains" has described a case in which the intrusion of lava may have caused upheaval of rock and elevation of the surface after a fashion different from that suggested by the preceding hypotheses. The Henry Mountains are a group of detached peaks situated on the northern bank of the Colorado River. In each mountain the strata are arranged in dome-shaped fashion dipping outwards in all directions from a centre. In the case of several of the mountains the heart of the hill is occupied by a core of intrusive Trachyte from which intrusive sheets and dykes are given off; the upper surface of these * Volcanoes, chap. xii. See also Darwin, Trans. Geol. Soc. second series, v. 601. t The Rev. 0. Fisher in his " Physics of the Earth's Crust " has sought in this cause the source of all the contortions which the rocks of the crust have iindergone. It is easy to see that such a cause is altogether inadequate for several reasons. In many cases where contorted rocks are thickly seamed with dykes, the south of Scotland for instance, the intrusion of the lava took place long after the formation of the folds. In other cases where contortion and inversion have been carried to a high pitch, there is a total absence of dykes or intrusive masses. The coast of Pembrokeshire and Glamorganshire and some of the most contorted parts of the Alps furnish examples very much to the point. 672 Geology. cores is arched and seems to run parallel to the bedding of the over- lying rocks ; the under side is a horizontal plane and rests on hori- zontal rocks. Each core in short resembles in its general form a plano-convex lens resting with its flat face on a horizontal table. To these cores Mr. Gilbert gives the name Laccolites, i.e. cisterns full of rock. He supposes the Laccolites to have been formed and the upheaval of the mountains to have been effected in the following way. Lava is pumped up through a chimney or fissure traversing horizontal beds of rock. At a certain point in its upward course it spreads itself out between two adjoining beds in the form of an intrusive sheet. By further additions of lava from below the sheet is thickened; the thickening will necessarily be greatest directly over the orifice and will decrease from that point as a centre all round. Thus the lenticu- lar shape of the cores will be produced. A necessary consequence of this process will be the bending up of the overlying beds into a dome. It can hardly be said that the evidence adduced by Mr. Gilbert in support of his views is absolutely conclusive. Should further research confirm his explanation, mountains of this type will probably turn out to be exceptional ; isolated eminences they must be and not members of a mountain-chain, and they cannot possess the contortion and inversion which are found in all true mountain-chains.* Contraction Theory. With regard to the preceding hypotheses we have certainly no right to say that the causes assigned by them have never acted in producing movements in the earth's crust ; there may be instances where this has been the case. But it is equally certain that the great body of earth movements were not produced by any of these causes for the following reason. If there is one point in connec- tion with the subject now before us on which we can speak with con- fidence, it is this; the movements of the crust have been brought about by a bending of its rocks into folds, and the folding is of such a nature that it could have been caused in no way except- by the rocks having been powerfully squeezed by a horizontal thrust. None of the hypotheses yet noticed provides machinery competent to supply such thrust. We now come to the only method yet suggested by which horizontal pressure is obtainable and which moreover supplies it as a necessary consequence of its mode of action. In this hypothesis we have not to go out of our way to devise far-fetched explanations of how squeezing might be brought about, squeezing is not so much an essential part of the process as the whole of it. The contents of Section I. seem to show that at present we have not knowledge enough to enable us to form a positive opinion on the questions, whether any part of the earth is fluid, and if so, what is the thickness of the solid crust. One point however has been estab- lished the interior of the earth is hotter than the outside ; and since it is very highly probable that the whole earth was once much more intensely heated than now and that the "present internal heat is only what is left of the original temperature, we arrive at a further fact, * Report on the Geology of the Henry Mountains, by G. K. Gilbert, Wash- ington, 1877 ; Nature, xxi. 177 (December 25, 1879). See also Reyer, Jahrbuch der k. k. Geol. Reichs. xxix. (1879) 405. Cause of Upheaval and Contortion. 673 which has an important bearing on the questions to be discussed in this section, namely, that the earth always has been and still is a cooling globe, and if a cooling globe, it must also be a contracting globe, if the materials of its interior are at all analogous to those which form its surface. Further it is almost universally true that the amount of shrinking produced by the loss of a given degree of heat (the coefficient of con- traction) is larger when the body is at a high than when at a low temperature.* From this it follows that since the interior of the earth it matters not whether it be fluid or solid is hotter than the outside, it must shrink faster than the outside. Practically the outside crust has so very little heat to lose that it may be said not to shrink at all, while the hotter nucleus is gradually drawing away from it. There is a ten- dency therefore to leave a hollow space between the crust and the nucleus, and the question arises whether the crust is strong enough to stand by itself without resting on the nucleus. The crust is attracted by the nucleus according to the law of gravitation, and it is easy to calculate the pull to which every point in it is subjected and deter- mine the internal pressures to which these pulls give rise. The reader who is not familiar with mechanics will perhaps' better realize the meaning of this by the aid of the following example. Suppose strings attached to every point of the outside of a hollow indiarubber ball and pulled outwards in such a way that each string would if produced pass through the centre of the ball. The effect will evidently be to stretch the indiarubber and draw its particles away from oiie another or to produce what is called tension ; if we know the force with which each string is pulled, we can calculate the amount of tension produced and determine whether the indiarubber will be torn or not. If now we suppose strings to be fastened on the inside of a hollow ball and pulled inwards towards the centre, the effect of the pulls will be to drag the particles of the shell nearer together and so make each to press against its neighbours; by determining the magnitude of this pressure we can decide whether it is sufficient to crash the shell. This is exactly the problem we have to solve in the case of the earth ; the pulls of the strings correspond to the attraction of the nucleus on every point of the shell, and it is found that the pressure they give rise to may be measured in this way. Suppose the shell anywhere cut in two by a plane passing through the centre, the faces of the two portions into which the shell is divided press against one another, and the pressure each exerts on any part of the other is equal to the weight of a column of rock having that part for a base and 2000 miles high, t Pressure like this is far more than sufficient to smash in the most unyielding materials known, and the crust could not sustain it for an instant. It must therefore follow the nucleus down. The only * Thus a rod of zinc in passing from a temperature of 100 to 99 will con- tract by about 32 million ths of its length at 0, while if it cools from 20 to 19 it will contract only 28 millionths. + Rev. 0. Fisher, Trans. Cambridge Phil. Soc. vol. ix. part iii. Physics of the Earth's Crust, p. 35; R. Mallet, Phil. Trans, clxiii. (1873) 173. 2 u 674 Geology. possible way in which it can do this is by its being crumpled into folds. It is as if we were to make a paper case that would just hold an orange of a certain size and then try to make it fit closely over a smaller orange ; we could evidently succeed only by wrinkling the paper. An old dried apple also furnishes an excellent illustration of the pro- cess. As drying goes on the fruit shrinks, but the skin does not : the latter accordingly, having to accommodate itself to the diminished nucleus, becomes puckered into wrinkles. According to this theory then the normal state of the earth's crust ought to be a crumpled one, and so it is. Where the force acted with concentrated energy along certain lines, or along lines of weakness, portions were ridged up into long narrow prominent protuberances, out of which mountain-chains were afterwards carved by denudation. Here too, owing to the inten- sity of the action, the rocks would be violently contorted. We have seen that excessive contortion is invariably found in lofty mountain- chains. We have certainly then in the unequal shrinking of the cool crust and the hot nucleus of the earth machinery competent to give the horizontal compression to which we feel sure the larger part of the movements of the earth's crust is clue. The further question arises, Will it give enough ; can it have caused the amount of squeezing and elevation which has actually taken place 1 The Rev. O. Fisher claims to have proved in his " Physics of the Earth's Crust " that it cannot, and that the hypothesis is therefore insufficient to explain the facts which it professes to account for. What Mr. Fisher has really done is this. His calculations go far to show that, provided the earth cooled in the way assumed by Sir W. Thomson (see p. 650), contraction would not suffice to produce anything like the amount of compression and elevation that has actually occurred. But this is quite another thing from disproving the Contraction Hypothesis. Mr. Fisher's investiga- tions tend rather to establish a strong probability that the earth did not cool in the way supposed by Sir W. Thomson. The notion that the earth's contraction has been the cause of the displacement of the rocks arid the elevations of the surface seems to have occurred first to Descartes (e"d. franyaise, 1668, p. 322). It was advocated by Constant Prevost* and Elie cle Beaumont,! but the latter unluckily tacked on to it an untenable hypothesis, which served rather to bring it into disrepute. It was the favourite theory of De la Beche,J and is adopted by a large number of the geologists of the day. Professor A. Favre has devised an ingenious way of imitating the manner in which contraction gives rise to contortion. He stretches a sheet of caoutchouc, covers it with a layer of pasty potter's clay, and allows it to contract : the clay is bent into folds which bear the closest * Memoires de la Societe d'Histoire Naturelle, 1829; Bull. Soc. Geol. de France, xi. 183 (1840) ; Comptes Eendus, xxxi. (September 1850) 461. t Notice sur les Systemes de Montagues (Paris, 1852) ; Hopkins, Anniversary Address, Quart. Journ. Geol. Soc. vol. ix. Geological Observer, p. 730 ; Researches in Theoretical Geology, p. 121. Mather, Silliman's Journal, first series, xlix. 284 ; Dana, ibid, second series, ii. 352, iii. 94, 176, 380, iv. 88 ; third series, v. 423, vi. 6, 104, 161. Origin of Volcanic Heat. 675 resemblance to the convolutions and inversions of the rocks of moun- tain-chains.* Layers of coloured cloth held on the caoutchouc by spring clips give equally good results. In conclusion, while we admit that the contraction hypothesis leaves some points to be still cleared up, we are yet justified in looking upon it as the most consistent and satisfactory explanation yet put forward of the cause of the disturbances of the earth's crust. At the same time it would be rash to say that it is the only cause ; some of the other moving forces suggested may have been the means employed in certain cases, t SECTION IV. ORIGIN OF THE HEAT REQUIRED FOR VOLCANIC ACTION AND METAMORPHISM. We have already seen that it is possible that the process which produced metamorphism in rocks, would, if carried to a more advanced stage, result in the formation of lava. If this view be adopted, an hypothesis which explains how the heat necessary for metamorphic action is obtained, will at the same time solve the problem of the origin of lava. And even if we take the opposite view that lava is a portion of the original fluid material of the earth that never has solidified, or if we hold that it is matter underground kept solid by pressure which liquefies when the pressure is decreased, still even then the problems of the cause of metamorphism and of the origin of lava have much in common. Both assume the presence of a high tempera- ture, and the concentration of that temperature in certain areas or along certain lines, and both involve the shifting from time to time of these areas from one spot to another. Without binding ourselves therefore to the assertion that the two problems are essentially one and the same, we may safely say that any- thing which throws light on the one will probably aid in the solution of the other. Explanation on Hypothesis of a Thin Crust. We have already seen how easily the doctrine of a thin crust solved the problem. The melted matter was ready to hand, and the fracturing and sinking down of the crust pumped it up. But this explanation falls to the ground unless the crust is very thin, and, as that is a very doubtful point, it cannot be accepted as satisfactory. There are besides several well-known facts that tell somewhat against it. The lavas of different vents differ considerably from one another, and often seem to be somewhat related in composi- tion to the rocks through which the eruption has burst its way if all lava came from one general reservoir, one would expect it to be more uniform in character. Again, there appears to be often no connection * Nature, xix. 103 (December 5, 1878) ; Comptes Rendus, Ixxxvi. (1878) 1092 ; also Charcontois, ibid. 1091. Daubree has made similar experiments with india- rubber balls, Geol. Experimentale, i. 385. t For an ingenious explanation of the cause of some of the inversions that are found along the flanks of some mountain-chains see Medlicott, Quart. Journ. Geol. Soc. xxiv. 34 ; and Memoirs of the Geol. Survey of India, vol. iii. part ii. 676 Geology. whatever between two vents very close to one another ; the lava standing at different levels in the two, and in one being in gentle and in the other in violent ebullition. Such a fact, though not altogether incompatible with the vents opening into the same great internal tank, does not lend support to such a view. A more satisfactory way of accounting for the rise of the lava was by the supposition that water found its way by percolation down to the molten interior, and being suddenly converted into steam, burst through the overlying rocks, flashed out in explosions, and forced up the lava. Mr. Hopkins' Hypothesis. Independently then of the fact that it is very doubtful whether the crust is as thin as the explanation just described requires, it is unsatisfactory on other grounds. The advocates of a very thick crust propounded several others in its place. Mr. Hopkins supposed the solid part of the earth to contain hollows filled with matter still in a state of fusion, and he supposed these hollows to have arisen in the following way. When an incrustation had begun to form at the surface, the expansive force of the contained gases would every here and there tend to fracture the thin crust ; and when areas of weakness had been once established, the gases would prefer to work their way out by apertures already in existence, even if they had to travel considerable distances to reach them, rather than to break open fresh points of discharge. Thus the crust would be par- celled out into disturbed and undisturbed districts, and solidification would extend downwards faster in the latter than the former. The under side of the crust would thus come to have inverted hollows over its surface, and the edges of these might extend down to the central solid nucleus if there was one, or unite beneath if the interior was wholly fluid, and so portions still liquid would be enclosed in the generally solid crust. Sir W. Thomson believes that the outer portion of the earth has a similar honeycombed structure. Assuming that the method of solidification supported by Mr. Hopkins is probable, his hypothesis still fails to explain some of the leading facts of volcanic action. According to it there seems no reason why the internal fiery lakes, and the volcanoes that they feed, should be arranged according to any law; one would expect to find them dotted about haphazard. But one of the most striking facts about volcanoes is the way in which they are arranged in lines, either coin- cident with or parallel to lines of great elevation. The omission to account for this fact is a very weak point in the explanation. Again the theory does not furnish a very satisfactory explanation of the alternate periods of repose and activity exhibited by volcanoes. One could understand a cavity being pumped out and there being an end of all eruption from it ; but why it should discharge a part of its con- tents, then rest a while, and then begin discharging afresh, it is not easy to see ; and if it was once emptied, what is to fill it again ? Further this hypothesis fails to account for the shifting of the scene of volcanic activity from place to place on the earth's surface. Such objections may not be fatal, but they naturally arise, and till they are met, the hypothesis cannot be said to be satisfactory. Explanations of Mr. Scrope and Rev. O. Fisher. Origin of Volcanic Heat. 677 Other authors will have it that below a certain depth the whole or portions of the crust are just verging on fusion, and that any change in temperature or pressure, or both, may turn the scale and convert solid into melted matter. Mr. Scrope has supposed that sometimes the requisite increase of heat is caused by the accumulation of thick masses of sediment in a way already described ; and that sometimes liquefaction may have been brought about by relaxation of pressure where the overlying rocks have been fissured or uplifted.* Mr. Mallet has gone into calculations which seem to show that the amount of heat generated on the first supposition is not sufficient to produce the amount of volcanic action that is actually going on (note to his paper quoted below). The Rev. 0. Fisher in one of his papers adopted the latter view ; he maintained that whenever the contraction of the crust raises a zone of rock into a mountain-chain the rocks underneath have no longer the weight of that zone pressing on them, and if it was only pressure that kept them solid, they would pass into the liquid state, t This hypothesis accounts for the association of volcanoes with lines of elevation. While speaking of Mr. Scrope's speculations, we must not forget to mention that he did inestimable service by showing that, whatever be the origin of lava, the force that raises it to the surface is undoubtedly the expansive power of steam. J Mr. Clarence King has adopted a somewhat similar line of argument and has made some additional suggestions. The pressure we know increases steadily with the depth : the fusing-point will also increase with the depth but much less rapidly than the pressure : the rate at which the underground temperature increases may diminish very rapidly below a depth of some 20 miles. If this be so, there may be a belt some 50 miles deep where the fusing-point is only slightly above the temperature at that depth, and if the pressure here be diminished, the fusing-point may sink below the temperature which exists at that depth and fusion will take place. We may call the zone where this can happen the zone of potential fusion. In fig. 228 the zone of potential fusion will lie at the depth corresponding to the point (7. Now erosion will take off some of the load and diminish the pres- sure, and hence erosion will tend to produce fusion beneath the eroded surface in the zone of potential fusion. But erosion will also tend to make the surfaces of equal temperature and therefore the zone of potential fusion sink deeper. Whether fusion takes place will depend upon the comparative rates at which erosion goes on and the zone of potential fusion sinks. If erosion goes on the faster, fusion will take place ; and this is not improbable, for the zone of potential fusion will sink very slowly on account of the low conducting power of the rocks above. Further mountain-tops are the places w r here erosion goes on fastest and where also the zone of potential fusion rises highest ; it is therefore beneath mountain-ranges that the underground matter has the best chance of becoming liquefied. Mr. King urges in support of * Volcanoes, pp. 265-275 ; Geol. Mag. v. 537. t Trans. Cambridge Phil. Soc. xi. ; Geol. Mag. v. 493. J Volcanoes, pp. 37, 116 ; Geol. Mag. vi. 196. 678 Geology. this view that in America periods of great volcanic activity seem in some cases to have followed periods of great erosion.* Sterry Hunt's Hypothesis. Dr. Sterry Hunt has put forward the following view of the present state of the earth's interior, and the origin of metamorphic and volcanic action. He admits that the earth was originally a molten mass, but maintains that cooling went on till the globe became solid from surface to centre. He holds that there is a large solid anhydrous nucleus still retaining a high temperature. Around this there is an external crust consisting mainly of sedimen- tary deposits. Between the nucleus and the crust there is a shell soaked with water that has percolated from the surface and has brought down with it in solution or suspension siliceous and aluminous matters, carbonates, sulphates, chlorides, and carbonaceous substances. The heat derived from the nucleus and the pressure due to the depth give to these watery solutions far greater chemical energy than they possess at the surface, and the result is that this intermediate shell is in a plastic or semifluid state. He maintains on the strength of many experimental results that hydrothermal action would manufacture in this shell the various forms of Metamorphic, Plutonic, and A^olcanic rock. He adopts the Scrope-Babbage theory that thick deposits of sedi- ment check the escape of internal heat; and wherever this occurs, the reactions in the intermediate zone are increased in activity, and volcanic eruptions, upheaval, and other allied phenomena result. For details of this elaborate theory the reader must consult the papers quoted below, t The Rev. 0. Fisher in his " Physics of the Earth's Crust " has adopted a somewhat similar view as to the constitution of the interior of the earth. He differs however from Dr. Hunt in considering the shell of fused or semi fused matter to have been formed during the consolidation of the globe in the following way. At the time when the temperature had fallen to the point when the oxygen and hydrogen were able to com- bine, the earth would be surrounded by an atmosphere of the water substance formed by their combination. The pressure exerted by this atmosphere would be very great and the temperature above the critical point. The water substance would have therefore great solvent power and there would be formed on the outside of the earth a shell consisting of a very intimate mixture of water substance, silicates, and other minerals. As the temperature fell, this shell would solidify outside ; but there may yet remain at a certain depth an intensely hot layer of matter in a state of hydrothermal fusion, from which the steam, gases, and other products of volcanic eruptions are derived.:}; It looks as if an hypothesis framed on some such basis as this is the one best calculated to meet all the requirements of the case. The semifluid shell need not be continuous all round ; numerous ribs may * Geological Exploration of the 40th Parallel, i. 701. t Canadian Journ. 1858, p. 203; -Quart. Journ. Geol. Soc. xv. 488; Comptes Rendus, June 9, 1862 ; Dublin Quart. Journ. July 1863 ; Silliman's Journal, second series, xxxvii. 255 ; xxxviii. 182 ; third series, v. 264 ; Geology of Canada, 1863, pp. 643, 669 ; Report, Geology of Canada, 1866, p. 230 ; Geol. Mag. vi. 245 ; Chemical and Geological Essays. Physics of the Earth's Crust, p. 96. Origin of Volcanic Heat. 679 connect the crust with the nucleus and brace the whole together so as to give the rigidity which there are so many reasons for thinking the earth as a whole possesses. The tides generated in the several com- partments would be slowly raised on account of the viscous nature of the material, and would not reach any great height in the short intervals during which the moon was facing any compartment. They might well spend their energy in compressing the semifluid matter and exert little pressure on the crust. We should have the machinery for producing from time to time fresh supplies of lava, and for carrying on continuously the work of metamorphism. Water might find its way down by gravitation through minute fissures into each rib and then exude into the heated compartments, where chemical action was producing a higher temperature than in the adjoining rib, in the fashion exemplified by Daubree's experiment. Whenever the crust adjusted itself to the diminishing nucleus by folding, fissures might be opened and lava squeezed up by the pressure and all the phenomena of volcanic action might be produced. At the same time some of the causes suggested by other hypotheses may have borne a share in the round of changes. Deposition may have caused a rise in the surfaces of equal temperature quite sufficient to produce a sensible effect. The weight of a pile of strata some miles in thickness may well have made itself felt at places where the crust was not more than 50 miles thick. The distribution too of the solid and semifluid parts of the shell would necessarily undergo change, for it would depend upon a very nicely-balanced adjustment of temperature and pressure ; varia- tions in temperature caused say by rise or fall of surfaces of equal temperature, and variations in pressure due to erosion or deposition on the surface, might cause some ribs to pass into a fluid state and new ribs to be formed by solidification elsewhere. We should thus be provided with a means of accounting for the shifting of the scene of volcanic activity. An hypothesis which is fairly satisfactory on all these points has decidedly a claim to attention. Mr. R. Mallet's Hypothesis. There is one very significant fact which all the explanations hitherto glanced at have failed to notice. Metamorphism is always accompanied by contortion, and in many cases, perhaps always, it increases in intensity as the crumpling grows more violent and complicated.* Volcanoes also in many cases range themselves along mountain- chains, that is along belts of excessive contortion. It seems then that not only are Metamorphic, Plutonic, and Volcanic action closely allied, but that contortion belongs to the same family. We have got at a fairly satisfactory explanation of the cause of con- tortion ; will the machinery which produced it also furnish us with the heat whose origin we are in search of? The late Mr. R. Mallet has attempted to show that it will, in a very remarkable memoir, t of which we must attempt to give an abstract. * For one admirable instance see Geikie. Trans. Edinburgh Geol. Soc. ii. 293, 294. t Phil. Trans, clxiii. (1873) 147 ; Proceedings Royal Society, xxii. 328. See also The Eruption of Vesuvius in 1872, by Professor Luigi Palniieri, translated by Robert Mallet ; and Scrope, Geol. Mag. xi. 28. 68o Geology. His ideas about the order of events during the passage of the earth from a gaseous into its present condition are as follows. 1st Stage. The chemical elements of which the earth is made up existed uncombined in a state of gas, and the first step was the union of these into combinations similar to those we find in the globe at present. 2nd Stage. When a state of chemical equilibrium had been estab- lished, the earth would be wholly fluid ; the surface cooling would be more rapid at the poles than over the tropics and in this way currents would be set up, but they would be superficial and would not be sufficient to establish anything like an equality of temperature through the mass, so that the interior would be much hotter than the outside. Cooling would go on by radiation from the surface very rapidly on account of the high temperature, and a solid crust would begin to be formed round the poles, and would spread both ways to the equator. The crust would pass gradually downwards into a shell of viscous matter, and this would graduate into the more fluid interior mass. He sees no reason to doubt the power of the crust to hold together, in spite of its small thickness, on the following grounds. His experi- ments on the cooling of slags show that a crust is formed round the fused mass when the temperature is very little below that of fusion, and the density of the crust differs very slightly from that of fused slag. Hence it is likely that in the case of the earth the density and temperature of the first-formed crust differed so little from that of the still fluid matter below that there wmild be no very strong tendency for the solidified portion to sink, certainly not enough to overcome the resistance to sinking caused by the viscosity of the matter immediately beneath it. As Sir W. Thomson has pointed out, the boiling up of gases would probably cause the crust to be full of cavities, and this might give it buoyancy enough to keep it up. 3rd Stage. After a time, but while the crust was still thin, the fluid nucleus would begin to contract faster than the crust, and the latter would have to accommodate itself by crumpling. He puts it forward as a conjecture, though he thinks a very probable conjecture, that the deformation of the crust at this epoch was effected by broad folds, and that by this means the great leading geographical outlines which the earth possesses at the present day were then impressed upon it ; that the main continental areas and oceanic depressions were then marked out and have remained substantially the same ever since. Where contrary flexures occurred at the junction of continents and sea-basins, lines of fracture were formed, and lines of weakness estab- lished which have continued to be lines of weakness ever since. The surface he believes at that time to have been still too hot to allow of permanent accumulations of water on it, but comparatively cool water may have fallen, to , be driven off as vapour and preci- pitated afresh elsewhere, and so there may have been a constant suc- cession of torrential rains and deluges ; at the same time the rapid transfer of heat from the interior may have suddenly heated the cooled portions of the surface, caused them to fly and crack, and so produced great ruin and shattering of the crust. By such powerful Origin of Volcanic Heat, 68 1 denuding agencies he thinks the so-called " azoic " rocks may have been formed. 4th Stage. The crust so far increased in thickness that there was a very material difference between the rate of its cooling and that of the fluid nucleus, and it became in consequence subjected to enormous tangential pressures as the shrinking away of the nucleus tended to deprive it of support. It was now thick and stiff enough to transmit these compressions within itself, and the consequence was that it became ridged and doubled up into great projecting wrinkles, which were afterwards to be licked into shape by denudation and become mountain-chains. The crust would give way most readily along the lines of weakness already established, and this is the reason why great mountain-ranges are found along sea-coasts. 5th Stage. The crust had become so far thickened and rendered so rigid as not to allow of the wrinkling that characterized the last stage. The shrinking of the hot nucleus still gave rise to tangential compres- sions, and the crust was obliged to yield to these in some other way and relief was now afforded by crushing to powder the rocks of which it is composed. It is the special object of the paper to show that heat is generated by the crushing, and that we have here a source amply sufficient to supply all the heat necessary for the production of the known phenomena of volcanic action. The effect of this crushing will be to produce belts, more or less tending to be vertical in position, of smashed and crushed rock. Down these water finds its way and is absorbed by the pulverized mass, and then, if the temperature generated be high enough, a mixture of fused or partially-fused rock and high-pressure steam is generated, which is forced up by the expansion of the steam, and raised to the surface as lava. For the elaborate experiments, calculations, and estimates by which Mr. Mallet has endeavoured to show that the heat produced by crushing would be sufficient to produce the effects assigned to it, the reader must consult the original papers.* There are many facts about volcanoes that are satisfactorily ex- plained by this theory. The rocks would yield most readily, the crushing would go on to the greatest extent, and the largest amount of heat would be generated along the old-established lines of weak- ness, and hence we get a reason why volcanic vents follow lines of mountain elevation. Crushing would not go on uniformly, but at intervals ; it would only take place when the accumulation of pressure had reached a point where the rock was able to resist no longer. Plence volcanic eruptions will be separated by intervals of rest. The production of an eruption also requires a certain balance between the supply of water and heat. If the former be in excess, the volcano may be either permanently or for a time drowned out or its activity very much reduced. The theory has the following additional recommendations. It has the merit of simplicity ; it calls in no hypothetical agencies the exist- * See also some experiments of Daubree's in the same direction, " Geol. Experimentale," i. 448. 682 Geology. ence of which rests only on supposition, for it merely requires that the interior of the earth should be hotter than the outside, which it almost certainly is, and the co-operation of gravitation. It puts volcanic action in the light of a beautiful compensating arrangement. If the crust were perfectly unyielding, it must relieve itself from the strain set up when the nucleus recedes, by violent disruption. As it is, whenever there is more matter in the interior than there is well room for, the overplus is converted into lava, and periodically transferred to the surface, and the cavities thus produced close in slowly as the crust adjusts itself to the shrinking nucleus. Lastly it accounts in an extremely simple way for the close connection between Metamorphic, Plutonic, Volcanic, and Elevatory action ; for it regards all four, not as isolated phenomena, but as different results of one common cause. On broad general grounds then it seems as if much might be said in favour of Mr. Mallet's views ; and it is really very doubtful whether, in the present state of our knowledge, it is any use trying to do more than reason in a broad general way about this class of questions. The Rev. O. Fisher has however attempted a minute and detailed criticism of the theory, and has on several grounds objected with con- siderable force to its conclusions. Mr. Mallet has replied, and the arguments pro and con will be found in the papers quoted below." 4 In a subsequent paper Mr. Fisher has returned to the attack,! and has tried to strengthen his objections by some elaborate mathematical investigations. It is open to question whether either his figures or those of the original Memoir of Mr. Mallet are really of much value. In a problem -of this kind, where the conditions are so complicated and the circumstances to be taken into account are so numerous, there is always considerable risk, when we attempt to reduce it to numerical calculation, that some point of vital importance has been overlooked. In the present instance it looks as if this risk is so large, that we cannot feel safe in employing numerical results in proof or disproof of the correctness of the hypothesis. It is unfortunate that in this and many other geological questions the data are too scanty to allow of our availing ourselves of the aid of mathematical analysis ; but if this be so it is better frankly to acknowledge the fact, and not to attempt to support or overthrow a theory by a show of numerical accuracy which we cannot feel certain is sound. Till however Mr. Fisher's objections have been satisfactorily disposed of, grave doubts must remain as to the adequacy of Mr. Mallet's hypothesis. But even if the machinery suggested by Mr. Mallet should turn out to be inadequate by itself to the production of volcanic phenomena, it is perfectly possible that it may have borne an important share in bringing them about. If we suppose that the main workshop where lava is manufactured, or that the source from whence it comes, lies in a semifluid shell between a thin outside crust and an interior solid * Quart. Journ. Geol. Soc. xxxi. 469, 511. Professor Hilgard has also criti- cized Mr. Mallet's views, Silliman's Journal, third series, vii. 535 (June 1874), and Phil. Mag. fourth series, xlviii. 41 (July 1874). t Phil. Mag. (October 1875), fourth series, vol. i. p. 302 ; Physics of the Earth's Crust, chap, xviii. Origin of Volcanic Heat. 683 nucleus, still a sufficient amount of heat may be generated in the way suggested by Mr. Mallet to complete fusion which has only begun, to keep lava in a fused state, and otherwise to aid in carrying out the production of volcanic eruption. Indeed it seems scarcely possible to ignore altogether a process which there is every reason to think is going on and which must be attended with the generation of heat to a greater or less degree. But when we come to Mr. Mallet's speculations as to the order of the events that have accompanied the growth of the earth into its present form, we find several of his notions to be directly in the teeth of well- established geological facts. He thinks that during the third stage, when the surface was still too hot to allow water to lie on it, the great mass of the oldest stratified deposits* were accumulated. This cannot have been, for the oldest rocks we know of contain fossils and animal life could not have existed imder the conditions which he supposes to have obtained during that period. Besides there is nothing whatever in the structure of these rocks to indicate so tumultuous and cataclysmal a mode of formation as Mr. Mallet's views would imply. A far more serious error is committed with respect to the fourth and fifth stages. The fourth he looks upon as a period of mountain-build- ing. It had, he admits, manifestations of igneous activity, but they were of a totally different character from those of the present day. Huge flows of lava were forced up fissures by hydrostatic pressure, but there was no action of an explosive nature. He grounds this belief on the startling statement, which will certainly be new to geologists, that the older volcanic rocks consist wholly of lava and " heated dust," and never show accumulations of fragmental matters such as are shot out of modern volcanoes by steam. He thinks this state of things may have lasted down to the date of the formation of the Chalk. It is scarcely necessary to say how completely this statement is opposed to the facts. Not to go far from home, there are in North Wales, the Lake country, and the south of Scotland countless instances of volcanic agglomerates, which are as distinctly the result of explosive action as the product of any modern volcano, and yet date far back beyond the limit fixed by Mr. Mallet. Again, according to him, during the fifth stage, mountain-building had ceased and explosive volcanic action had come in. As a fact, many of the loftiest mountain-ranges received their final uplift during this very period. Explosive volcanic action then can be carried much further back, and mountain-building brought clown much nearer to our day than Mr. Mallet seemed to be aware. He allows that his stages overlap to some extent ; in the case of the fourth and fifth, the over- lapping is so complete that they become practically coincident The crumpling up of the crust sometimes produces only contortion and elevation, sometimes metamorphism and volcanic action accompany these processes ; but the two operations have riot been confined to dis- tinct periods, they have been going on side by side ever since the formation of the oldest stratified rock we know. * I imagine this is what he means by " the assumed azoic and yet more or less stratified rocks. " 684 Geology. Professor J. Le Conte lias suggested that the first formation of mountains begins while the strata are still soft enough to yield to the compressing force. They then give way easily and no heat is pro- duced. Afterwards, when the rocks have become consolidated by compression, further crumpling goes on and then crushing and fusion results. * Professor Le Conte has also proposed the following slight modifica- tion of Mr. Mallet's hypothesis. Instead of supposing that the lines of weakness are fissures established at an early date, he employs the Scrope-Babbage theory to account for their production. Wherever a thick pile of sedimentary rocks accumulates, the escape of heat is checked, the surfaces of equal temperature rise, the rocks under the combined influence of heat and water are softened, and lines of yielding are determined. Then contraction produces horizontal thrust, and its effects are most conspicuous along these lines, t Bearing of Mr. Mallet's Hypothesis on Metamor- phism. Mr. Mallet's views also come in very handy to explain the constant association of metamorphism with intense crumpling and puckering. There is one very weak point about all the explanations which look to the internal heat of the earth as the direct source of the heat required for metamorphism. If this were the real explanation, it would seem that the main thing requisite for metamorphosing a rock was to sink it deep enough ; and that, as a rule, the greater the thick- ness of rock under which any bed had been buried, the more thorough would be its alteration. Of course this would not be true in every case, because other agents besides heat are wanted for metamorphism ; but we should expect very generally to find those rocks most intensely altered which have been sunk deepest. But very often this is not the case. To take one instance given by Professor Geikie : the Carboniferous Limestone of the South Wales Coalfield has at one time been covered by from 10,000 to 12,000 feet of strata, but it shows no traces of metamorphism ; the rocks of the Central Highlands on the other hand are intensely altered, though at the time when their metamorphism took place they cannot have had over them more than 5000 feet of strata. It is clear then that metamorphism does not necessarily depend on the depth to which a rock has descended into the earth ; but it is we have seen allied to contortion, and there does seem to be a probability that the heat developed during contortion may have been sufficient to produce it. We can now understand the structure of the mountain- chain generalized in fig. 231. We have there Granitic axes which shade off into Foliated Schists, and these in turn melt away into unaltered rocks ; this gradual passage leads us to look upon the Granite as simply an excessively metamorphosed rock. Again the whole mass has been subjected to folding on an enormous scale, and crumpling is more and more developed as'we approach the Granite, so that metamorphism and puckering go on increasing together. It seems then that either the one has given rise to the other, or that both are * Silliman's Journal, third series, vii. 167. t Ibid, third series, v. 453. Great Physical Features of the Earth's Surface. 685 the results of a common cause, and the latter is the explanation of the connection upheld by Mr. Mallet.* SECTION V. THE GREAT PHYSICAL FEATURES OF THE EARTH'S SURFACE. Continents. The great masses of continental land differ so widely in their outline, their general trend, and other particulars, that at first sight they would seem to have but few points in common. If however we pass by minor details and fix our attention on some of the larger physical features, we shall see that these are arranged in such a way that it is perfectly possible to conceive a pattern according to which all continents are framed. An analogous case in the animal kingdom will perhaps help us to realize this conception. There is a large body of animals, including quadrupeds, reptiles, birds, and iish, which though they differ as widely as may be in many points, yet all agree in having a backbone, and in which certain other peculiarities of anatomical structure always go along with the possession of a back- bone. All these, whether they walk, crawl, fly, or swim, whether they live on animal or vegetable food or both, and however they may differ in other respects, may be obviously said to be constructed on the same pattern. It is in a somewhat similar sense that we may also use this expression of continents. Indeed it is not an uncommon figure of speech to talk of them as if they were vertebrate in type, and to call mountain-chains their backbones. No metaphor could have been more unfortunate, for mountain-chains do not run along the middle of continents but skirt their borders. The typical conti- nent consists of a low-lying interior portion with mountain-chains running along its margins and separating it from the ocean. North America is the most pronounced example : on the west we have the great belt composed of the Sierra Nevada and the Rocky Mountains, on the east the Appalachians, while between lies a broad undulating spread of low country forming the basin of the Mississippi. And we here find too another feature which seems worthy of notice. The loftiest chain faces the largest ocean ; on the one side we have the broad Pacific bordered by a range which runs up to 18,000 feet in height, on the other the narrower Atlantic and a range whose maximum height does not exceed 6000 feet. No other continent conforms so rigidly to this type as North America, but it does seem possible to recognise it more or less dis- tinctly in all the rest. The slopes of the ocean bed moreover seem in many cases to be related to the gradients of the adjoining land. On that side of the * The notion that the heat developed during contortion has been one of the producing causes of metaniorphisin, has been floating ahout in a vague sort of way in the rninds of geologists for some time. See Orographic Geology, G. L. Vose, 1866 ; Ramsay's Address, British Association, 1866. Dr. S terry Hunt states that Professor W ul 't z enunciated this view in a paper read before the American Association for the Advancement of Science in 1866 and published in the American Journal of Mining, January 1868. 686 Geology. continent which possesses the loftiest mountain-chain the sea-bottom plunges down rapidly and the maximum depth is reached within a moderate distance from the shore. On the further side of this depres- sion the sea-bed rises again at a more gentle angle. In fact the sea- bottom is merely a prolongation of the slope of the adjoining land : where this runs up rapidly into a lofty chain, the steep slope of the land is continued beneath the sea-level and the water deepens very rapidly ; where the bordering chain rises with gentle slopes into moun- tains of moderate elevation, the water also deepens more gradually.* Distribution of Volcanoes. When we cast our eye over a map of the world on which the positions of active and recently-extinct volcanoes are shown, we see at the first glance that volcanic vents are not sprinkled at haphazard over the globe and we quickly recognise a law after which they are distributed. Long trains of volcanoes stretch in slightly sinuous lines for enormous distances, in some cases over nearly one-half the circumference of the globe. In many cases these volcanic trains coincide with or run very near to lines of great moun- tain-chains, and hence they frequently skirt the coast of an ocean. From the long ranges of volcanic vents branches are given off; and these, though they bend to and fro with gentle windings, approximate in their general trend to straight or slightly curved lines. The two longest and most continuous volcanic trains skirt the shores of the Pacific Ocean. The one on the east runs along the western coast of America. It follows in North America the line of the Sierra Nevada and coast ranges ; some of its volcanoes are still active, but many more, which were in full operation during recent geological periods, are either extinct or have sunk into the Solfatara and Geyser stage. Along the continuation of this line through the Isthmus of Panama active volcanoes are thickly ranged, and its further prolongation down the Andes contains the largest and loftiest volcanoes in the world. The great volcanic train on the west of the Pacific ranges through the Aleutian Islands, along the Kamtchatkan peninsula, and on through the Japanese and Philippine Islands. Here it gives off two branches ; the westerly branch ranges through Java, Sumatra, and the Andaman Islands nearly to the mouth of the Ganges ; the easterly branch can be followed through the New Hebrides and the Friendly and Society Islands to the Marquesas Archipelago. The trend of the New Hebrides when produced points to the volcanoes of New Zealand and Mount Erebus and Mount Terror on the Antarctic continent, and we may have here a southerly continuation of the original train. The portion of this train between Kamtchatka and the Philippines does not run along a mountain-range, but it is in a general way parallel to the chains which border Asia on the east. Another great line may be traced down the eastern side of the Atlantic from Greenland and Bear Island through Jan Mayen, Ice- land, the extinct volcanoes of the Faroe Islands, the Western Islands * See Dana, Manual of Geology, pp. 9-38 and 731-737 ; and his papers quoted in the note to p. 674. Lapparent, Traite de Geologic, pp. 67-83, Rev. 0. Fisher, "On the Inequalities of the Earth's Surface viewed in connection with the Secular Cooling," Trans. Cambridge Phil. Soc. xii. part ii. Great Physical Features of the EartJis Surface. 687 of Scotland and Antrim, the Azores, the Canaries, the Cape de Verde Islands, Ascension, St. Helena, Tristan d'Acunha, and other volcanic islands of the South Atlantic. This line contains many volcanoes still active, but a larger portion have become extinct in recent times or are gradually approaching extinction. This train coincides with a sub- marine ridge which runs along the whole length of the Atlantic ; its volcanoes stand either on the ridge itself or on spurs given off from it ; it probably therefore follows a line of folding in the earth's crust. Two branches are given off in an easterly direction from the train. One running through Central Europe contains the volcanoes now extinct of Auvergne and Central Germany ; the volcanic vents of the Caucasus and Demavend, which are in the Solfatara stage, perhaps lie on its easterly prolongation ; and, may be, the volcanoes on the tableland of Central Asia and in Mantchouria also belong to this branch. The second branch follows the extinct volcanoes of Catalonia, the volcanic vents of the Mediterranean and the Grecian Archipelago, and the volcanoes mostly extinct of Asia Minor. These two branches run parallel to the great line of mountain eleva- tion of the Pyrenees, the Alps, the Carpathians, and the Caucasus ; the volcanic outbursts commenced during the period of upheaval of this chain, reached their greatest intensity while upheaval was going on, and have most of them declined in activity since. For further details on this subject the reader may consult Judd's "Volcanoes," chapter viii., and the appendix to Scrope's "Volcanoes." Doctrine of the Permanence of Oceans and Continents. When Geology first taught men that by far the larger part of the dry land was formed of marine strata and that therefore very nearly all, probably the whole, of our present continents had been at some time or other sunk beneath the sea, the notion not unnaturally sprang up that in all probability a large part, perhaps the whole, of the bed of the present oceans had been during past time raised into dry land. Oceans and continents it was supposed had exchanged places and any conceivable redistribution of land and water whatever was thought to be within the bounds of possibility. But many of the most distinguished of modern geologists have been led to think that such a doctrine is not tenable, and have urged very strongly the view that the great depressions in which the oceans lie and the great elevated masses of land which form our continents were marked out at a very early period of the earth's lifetime ; and that, though they have many times over undergone upward and downward movements to a certain extent, these have never sufficed to lay dry the bed of one of the great basins or to submerge to oceanic depths one of the great elevated blocks which form the continents. When we come to inquire into the merits of these opposite hypo- theses, we see easily that the depth of the oceans forms no obstacle to the acceptance of the older view. We know that in the Alps marine strata which had once 50,000 feet of rock piled on the top of them are now found 10.000 feet and more above the sea-level : Nature here had force at her disposal which sufficed to produce an absolute elevation of 10,000 fathoms. And if it be urged that mountain-chains are excep- 688 Geology. tional cases where the elevatory forces were unusually powerful, we can yet find cases in non-mountainous regions, the Carboniferous rocks of South Wales will do for one, where absolute elevation of between 2000 and 3000 fathoms has occurred, an amount quite sufficient to lay dry the larger part of the Atlantic and Pacific. So far the views of the earlier geologists are not inconsistent with what has actually happened, but we meet with a decided hitch when we are asked whether we can point to any case where upheaval to this extent has been carried out over an area equal to that occupied by the large oceans. We know of no such case ; the areas over which great vertical upheaval has taken place may be hundreds, even thousands of miles in length, but they have all a comparatively small breadth. What is more, we cannot point to any group of rocks which extends over an area comparable to those of the great oceans. We speak of all the rocks that were formed at the same time and in the same marine area collectively as a Formation, and in any Formation that has come down to us approximately in its entirety we have Littoral members which run along the old coast-line, Thalassic members which were formed further out at sea, and Oceanic members which were deposited so far from land that they contain little or no mechanically- carried sediment. If any of these formations was spread over the whole of the bed of an ocean such as those of the present day, we ought to find Oceanic deposits covering areas comparable to those of the oceans of to-day. But this is what we do not find. In some cases rocks do maintain an Oceanic character for very long distances in one direction ; but if we follow them along lines at right angles to that direction, we find them putting on Thalassic and Littoral characters within com- paratively small spaces. In other cases the Oceanic character is pre- served over a comparatively small central space, and this is encircled on all sides by belts of Thalassic and Littoral rocks. Rock-formations occur either in long narrow belts, in which case they must have been formed along the margins of large oceans ; or they extend over com- paratively small irregularly-shaped areas, in which case they were formed in land-locked seas of moderate dimensions. No more typical instances of Oceanic rocks can be found than the Chalk and the Carboniferous Limestone, but neither was formed in an ocean ; the Chalk sea may have been somewhat bigger than the Mediterranean, the basins in which the Carboniferous Limestone was formed were none of them anything like so large. It is in this sense and this sense only, that the term Oceanic is applied to rocks. The water in which Oceanic rocks were formed was oceanic in so far that it was clear and free from sediment, but it was not oceanic in its area and extent. These and other arguments have been urged in favour of the modern doctrine of the Permanency of the Great Ocean Basins and Continents, and it must be allowed that they have very considerable weight, but the question is a very obscure one, and we are hardly justified as yet in dogmatizing upon it.* * See Wallace, "Island Life," chap. vi. ; A. Geikie, " Geographical Evolu- tion," Proceedings Royal Soc. Edinburgh, 1879, p. 426. Great Physical Features of the Earth's Surface. 689 If we adopt provisionally this hypothesis, it seems likely that, as Mr. Mallet and others have suggested, far back in the earth's lifetime, when the crust was thin, it bent itself into broad folds as it followed down the contracting interior; that the broad-backed summits of these folds were the nuclei of our existing continents, and that the depressions between the folds were the basins that were afterwards to receive the oceans. Continental areas have been many times depressed and laid under sea-water, but they have always stuck up as a whole above the adjoin- ing ocean- basins; and these basins have always maintained in a general way the character of great depressions, though their beds have under- gone many considerable oscillations of level. Growth, of Continents. Assuming the truth of the doctrine of the Permanency of Oceans and Continents, the type of structure which seems to prevail in a more or less pronounced form in all con- tinents may be supposed to have arisen somewhat in this fashion. Let us start with a continent bordered by mountain-chains which have been recently elevated. Fringing the coast there will generally be a belt of shallow water of variable but comparatively small breadth, and on the bed of this all the mechanically-carried sediment derived from the waste of the land will be deposited, while along its outer margin Oceanic rocks of organic origin may grow up. The bed of this marginal belt we must suppose to be sinking slowly, and the character of the addition which will in the end be made to the extent of dry land will depend upon whether the sinking is frequently interrupted by upward movement or whether it goes on for long periods without check. Sinking we must recollect means that the area which is being depressed is being squeezed down into the form of a trough ; eleva- tion takes place when any portion of the crust is squeezed up into an arch. If sinking goes on only for a short time, and if then the rocks de- posited during the subsidence begin to be elevated, the following results will follow. The formation laid down will be thin ; because it is thin, it will yield readily to the compressing force, its rocks will be bent into broad folds, and they will not be raised to any great height. The broad folding will give rise to gentle dips, and there will be an absence of contortion and violent disturbance. A marginal strip will in this case be added to the continent, which will be low-lying and will consist of a group of rocks of small thickness very slightly disturbed. Subsidence may again follow and submerge the whole or parts of this newly-born land, and a second Formation may be deposited upon the rocks of which it is composed. After a short time elevation may again set in, subsidence may follow it, and sinking and upheaval may for a long period succeed one another at intervals of no very great duration. The final result will be a tract of land of no great elevation, made up of a number of Formations each of no great thickness and separated from one another by unconformities. The rocks will be nowhere sharply 2x 690 Geology. folded but will undulate in a succession of broad arches and troughs ; the surface will be carved by denudation into hills and valleys, but there will be no range which has the distinguishing characters of a mountain-chain. As long as this process goes on the continent may grow, but during the process of its growth it will lose to a certain degree some of the physical features which distinguish the continental type of land. Its mountain-chains will no longer run close to the coast but will be shifted back somewhat inland, and between them and the sea there will be a broad spread of low country. Some such process as this may have been going on in India since the elevation of the Himalayas, and may have produced the plains of Hindostan which now separate that chain from the Indian Ocean. But the order of events may be vastly different from that just imagined. If the sinking goes on for very long periods without check, bed after bed may be laid down till a mass of rock miles in thickness has accumulated. When upward movements commence, the compressing force to which they are due will have to deal with an enormously thick mass of strata. There are two reasons however Avhy it will be easier to squeeze up the newly-deposited beds, in spite of their great bulk, than to produce any effect on the older rocks of the original continent. These have been already crushed together and packed into a firm and solid mass, and they will therefore oppose far more resistance to tangential pressure than the uncompressed strata that have been more recently deposited. Also the lower portions of these newer beds will have been sunk deep enough to be considerably softened by heat and hydrothermal action. It will be the newer strata then that will give way. But though it is here, if anywhere, that yielding must take place, it will be no easy task to overcome the resistance of so vast a mass of rock, and only force of enormous magni- tude will be equal to such a feat ; between these mighty pressures on the one hand and the dead weight of miles upon miles of rock on the other there will be a long and severe struggle, and as the rocks yield they will be enormously compressed and shattered, and they will further be squeezed up to great elevations. The result will be a mountain-chain with great contortion, inversion, faulting, and smashing. Since there was no upheaval and denudation while they were being formed, there will be no unconformity throughout the vast thickness of rocks of which the chain is composed. During all this time the mountain-chain that had previously formed the border of the continent will have been worn down by denudation. It may have been so far lowered as to lose completely its mountainous character as far as elevation goes ; even if it does not suffer to this extent, it may still come down to a height far less than that of its newly-born brother. In either case the continental type will be main- tained ; there will be a great mountain-chain along the coast ; within this the ground may nowhere be mountainous, or this chain may be backed by a much lower range. The Kwemluni range on the north of the Himalayas is an instance of such an interior chain. It is the older of the two; it formed the mountain-border of the continent Great Physical Features of the Earth's Surface. 691 while the rocks of the Himalayas were being deposited ; and it may once have been as high as the line of peaks to the south which now towers over it. It is possible that this may have been the history of continents. They may have grown by successive additions along their borders after the fashion just sketched out ; but even if this hypothesis be sound, it still leaves much to be explained. Why in some cases subsidence goes on for enormous periods without pause, and why in other cases it is frequently interrupted by elevation, we can hardly even guess ; and on other points our explanation is vague and incomplete. But if this view be sound, a very important consequence follows. The amount of continental land must have been constantly on the increase during all past time, and the continents must have been growing gradually more and more compact ; during past portions of the earth's lifetime the channels of communication between different parts of the ocean were more numerous than now, but as time went on land barriers became more continuous and the part of the earth which is covered by water became more thoroughly divided into distinct oceans. This fact, if it be a fact, has a very important bearing on the changes of climate which have occurred in bygone times. It will also serve to explain why the same animals and plants seem at one time to have spread over much larger areas than now, and it would indicate that the establishment of definite provinces each possessing a distinct fauna and flora of its own has been a gradual and compara- tively modern process. Association of Volcanoes with Mountain-chains. It is a fact that some of the long trains of volcanic vents run along moun- tain-chains, and that others run parallel to mountain-chains and at no great distance from them. It is also a well-ascertained fact that in some instances the volcanoes which accompany mountain-chains burst out for the first time and attained their maximum activity during the upheaval of the chain, and that when that was completed their energy began to abate and some of them became extinct. It seems likely then that the association of volcanoes with mountain- chains is not an accident, but that in some cases at least both may be the results of a common cause. We have been able to arrive at a fairly satisfactory explanation of the way in which mountain-chains are raised ; but till we know more about the machinery of volcanic action we can only conjecture why great earth-movements should give rise to volcanic outbursts. Mr. Mallet's views on the subject have been given on p. 679. If we hold that there is at a moderate depth a shell of matter in a state of potential fusion, it may well be that the heaping up of a pile of deposit some miles in thickness so far increased the pressure as to keep the portion of this shell underneath it in a solid state. When however this load began to rise, the pressure may have been relieved far enough to produce fusion ; and when rending took place, the fused matter was squeezed out and driven up through the fissures by the expansion of the contained steam. Any one of moderate ingenuity may frame many such explanations, possible enough 692 Geology. perhaps but little better than guesses, and we will therefore dwell no longer on the subject. It seems highly probable however that we have in these long volcanic trains a magnified representation of what we see on a small scale when we study a volcanic area of moderate size. There rectilinear fissures have been actually observed to open and discharges to take place from vents situated on them ; the separate volcanoes of the group also range themselves along lines approximately straight. We may therefore fairly look upon the longer lines of volcanoes as marking the position of great fissures or systems of fissures, and we may refer the produc- tion of the vents to the powerful earth-movements which have in so many cases taken place along these lines. Other Hypotheses. Professor Staler has made the following suggestion as to the formation of continents and mountain-chains. Continents he believes to be produced by broad foldings affecting the whole thickness of the crust. He accounts for the sharper wrinkles of mountain-chains in the following way. The exterior part of the crust is and has been for a long time at the same temperature as the atmos- phere ; it therefore loses no heat and does not contract at all. But a deeper layer contracts sensibly, and to compensate for this the super- ficial portions of the crust must wrinkle up. According to him it is this crumpling of the outer shell of the crust which is the cause of mountain elevation.* Again attempts have been made to account for the formation of continental and oceanic reliefs by the hypothesis that the conducting power has been greater along some radii than along others. Some portions of the earth would then cool and contract faster than others ; the first would sink down into oceanic depressions, while the second would be left standing up as continental tracts, t Archdeacon Pratt J adopts a similar view, on the ground that the mass of the earth is found in some cases to be denser beneath the ocean than beneath the land, and to be least dense beneath great mountain-chains. It cannot be said that the evidence for the increase of density beneath the oceans is satisfactory, and the hypothesis of unequal con- ducting power is purely gratuitous. SECTION VI. CONCLUDING REMARKS ON SPECULATIVE GEOLOGY. With regard to the questions treated of in this chapter, the con- clusion of the whole matter seems to be that at present we know scarcely anything for certain about them. But such a state of uncer- tainty need not be a source of regret. It would doubtless be pleasant to be able to make up our minds on these fundamental questions, but on the other hand it is anything but disagreeable to reflect what a wide field of inquiry lies as yet all but untouched before the geologist, * Geol. Mag. [1] v. 511. t Professor J-*H>S Le Conte, Silliman's Journal, third series, iv. 345, 460. J Phil. Trans, clxi. (1872) 335; Figure of the Earth, fourth edition, p. 201. Concluding Remarks on Speculative Geology. 693 and it is most encouraging to the inquirer to bear in mind what a host of opportunities are open to him of distinguishing himself. There are, besides those noticed, other problems in the speculative domain of Geology of surpassing interest, but want of space, and still more the very small way that has been made towards their solu- tion, forbid our doing more than glance at some of them here. Geological Time. Under this head we may reckon the attempts that have been made to determine in years the age of the earth, or rather the time which has elapsed since it came into a condition approximately resembling the present ; and also what is the probable expectation of life in the case of our planet and the system of which it forms a part. Sir W. Thomson has tried his hand at these pro- blems, and there has been one speculation thrown out since he wrote which may so seriously modify his conclusions that we shall do well to refer to it. Starting with the Nebular hypothesis as a basis, he has tried to approximate to the date of the time when the sun's heat will be ex- hausted. He has assumed that the sun has been, and will be, cooling all along. Mr. Lockyer has however shown that such may not have been the case, and has suggested a method by which the failing heat may have been replenished, perhaps over and over again.* Adopting views similar to those of Prout and Dumas, he thinks it likely that many of the substances which we believe to be elements, because we have not been able to decompose them, are really compounds ; and that during the early periods of a star's lifetime their components existed in an uncombined state, the dissociation being perhaps due to intense heat ; when the heat was so far reduced that* it was no longer able to keep the elements apart, chemical combination took place. Now when chemical combination takes place, heat is developed. The constituents before they combined possessed potential energy, latent heat as it is styled : when the union takes place this energy becomes active energy or energy of motion and sensible heat is pro- duced. The quantity of heat thus rendered sensible may have been sufficient to raise the general temperature to a large extent. Thus the life of a star may not have been one continuous process of cooling, but it may have every now and then fired up afresh, and the time taken to reduce it to a certain temperature may have been much longer than if it had gone on always steadily losing heat. There has been always a clashing between geologists and physicists on the subject of geological time. The extreme slowness with which geological changes takes place leads the first to demand enormous periods for the production of the results he sees around him ; the speculations of the latter tend to tie down the allowance that can be granted within rather narrow limits. Possibly, if Mr. Lockyer's hypothesis turns out to be well founded, the physicist may be able to be more liberal in his estimates and the want of agreement between him and the geologist may be removed. Since the doctrine of Evolution has obtained a firm hold, attempts * Proceedings Royal Soc. xxi. 513 (November 27, 1873) ; Chemical News, xxviii. 175 (October "3, 1873) ; Nature, xi. 335. 694 Geology. have been made to obtain an approximate value of the time during which life has existed on the earth by estimating the time that would be required for the changes in animal and vegetable life which the fossiliferous rocks show to have taken place. The present rate of change is exceedingly slow, and if we take this as our standard, the whole period necessary will be enormously great. Mr. Wallace has however shown that this may not be a fair way of looking at the question. The modifications which living creatures undergo during the process of evolution are largely dependent on change in surround- ing conditions, and among these changes in climate occupy a leading place. Now one cause, which has probably played an important part in bringing about changes of climate, is, we shall see shortly, a change in the eccentricity of the earth's orbit. Mr. Wallace has pointed out that for a long time back the eccentricity has varied very little ; as far as this cause goes then, we should expect during the present day a comparatively slow rate of evolution ; but the changes in the eccen- tricity have at past periods been large and have followed one another much more closely, and therefore it is not unlikely that evolution may in former days have gone on at a rate correspondingly more rapid.* Former greater Intensity of Geological Action. There is one other point in geological speculation too important to be passed over. The earlier geologists, we have seen, when they were in diffi- culty, did not hesitate to call in to their aid agencies far more powerful than those of the present day, and sometimes altogether different in kind from any we are acquainted with. Their method was profoundly unphilosophical, for they gave no reason why the energies of nature should have been formerly greater than now, or other than those of our own time. This school is often spoken of as the Cataclysmal or Paroxysmal School. The reaction against these false views led to a school which had a tendency to run into the other extreme. Its adherents maintained not only that " the great mutations of the world are acted," but that they were acted long long ago. These geologists hardly went so far as to assert that the condition of the earth from the formation of the oldest rock down to the present time has been all along exactly what it is now; but they looked with suspicion on any proposal to call in agencies different from those of the present time. Their caution, though perhaps sometimes carried too far, was decidedly a step in the right direction. The supporters of this view have been distinguished as Uniformitarians. Their line of argument is, only give time enough, and every change which Geology shows us has taken place on the earth can have been produced by the action of existing causes ; there is therefore no neces- sity for calling in any extraordinary powers, and if there is no neces- sity, it is unphilosophical to do it. ' It is probably true that existing causes are quite sufficient for the production of past geological changes, if they only act long enough. But the time required will be of enormous duration, and the question arises, Can the assumption of an indefinite lapse of time be justified 1 * Island Life, p. 224. Concluding Remarks on Speculative Geology. 695 This question we have just seen is still an open one. There is the further objection that it is not only possible, but even highly probable, that conditions different from those of our day have existed during past epochs. Indeed if the history of the earth's development has been anything like that sketched out in the present chapter, and if there be any truth in the modern doctrines of physics, it is impossible that Uniformitarianisin can be literally true even for a limited period. When the earth was hotter than it is now, all the phenomena which depend directly or indirectly on the internal heat, such as metamor- phism, volcanic energy, and contortion, must have been proportionately more energetic ; and if the sun was at the same time hotter, all the geological operations depending on meteorological conditions, such as denudation, must have gone on faster and on a larger scale than now. As Sir W. Thomson well puts it "A middle path, not generally safest in scientific speculation, seems to be so in this case. It is probable that hypotheses of grand catastrophes, destroying all life from the earth and ruining its whole surface at once, are greatly in error it is impossible that hypotheses assuming an equability of sun and storm for one million years can be wholly true." The investigations of Mr. Gr. Darwin point to a similar conclusion. If, as he has shown good reason for believing, the moon was at one time much nearer the earth, and if the month and day were then shorter than now, the tides would be higher and would recur oftener than at present. Marine denudation would therefore go on more energetically and the rate of deposition might be more rapid. ]STo geologists will accept the somewhat sensational picture which has been drawn by Dr. Ball of the geological operations of that period * and Mr. Darwin has himself pointed out its extravagance,! but it is not unlikely that this cause may have operated to make both denuda- tion and deposition go on faster than in our own epoch. It is however somewhat open to question whether the older rocks have really the enormous thickness which has been assigned to them, and on the strength of which geologists have been led to seek for some cause which would make rock formation during early geological periods a more rapid process than it is at present. These rocks are violently contorted, and we have seen what risk of error there is in measuring the thickness of contorted rocks (p. 485): it is by no means clear moreover that proper care has been taken to carry the sections from which the thicknesses were estimated along the full dip, and here again there is an opening for enormous possible exaggeration of the thickness (p. 466). But though the views held by the school of Uniformity cannot be exactly correct, it may be that, for the period for which they are maintained, they are not far from the truth. It is in the .highest degree improbable that the oldest known rocks are really the first rocks that were ever formed, utterly unlikely that there were none before them ; indeed we may almost say that we know that this is not the case, and that we are certain that the time which has passed by since the deposition of those rocks, enormous as it seems to us, is as nothing * Nature, xxv. (1881) 103. t Ibid. xxv. (1882) 213. 696 Geology. in comparison with the large lapse of ages which have rolled away since the earth became tenanted by life and denudation and deposition began their career. So that though really the earth has been steadily losing energy all along, yet the rate of loss may have been so slow, and the interval between the formation of the oldest-surviving rock and to-day may be in comparison with the whole lifetime of the globe so small, that we may practically look upon the condition of the earth as having been constant during the period with which Stratigraphical Geology deals, and may for so far back be Uniformitarians without sensible error. Of course this view involves a longer lifetime and a slower cooling than physicists have been hitherto disposed to concede ; but their estimates on neither of these points are beyond question ; indeed on the first we have just seen that they may have to be mate- rially extended. On the other hand a modified Uniformitarianism may be the true solution. While we resolutely reject agencies differing in kind from those of the present day, we may yet allow of a difference in degree, and admit the possibility of the rates of deposition and of the change in life having been more rapid in former times than now, and so not exceed the limits in time to which physical speculations seem to tie us down. As far as we can at present judge, it certainly seems likely that one of these two views represents the true state of the case, but our choice does not lie wholly between them. It behoves us to be very careful how we appeal to causes differing in kind from any of which we have had experience, still we must not lose sight of the possibility of there being forces, which are periodic in their action, and yet recur so seldom that the span of human experience has not been long enough to witness even a single instance of their display. And this is not one of those purely gratuitous assumptions, unsupported by analogy or probability, the use of which brought the Paroxysmal School into disrepute. For instance if we adopt the contraction hypothesis of the origin of moun- tain-chains, it is perfectly conceivable that the action of its machinery may be of this nature. The pressure may have to go on accumulating for a very long time before it can give rise to any motion ; and then, when it passes a certain limit, portions of the crust may give way with a start and a very considerable amount of disturbance may be gene- rated suddenly ; after the relief thus afforded, there may come a long interval of comparative rest till a head of pressure has gathered suffi- cient to make fresh disruption necessary. This explanation is perhaps as admissible as the one which supposes mountain-chains to have been raised by a continuous, gentle upridging, prolonged over very long periods. Possibly the best explanation of all would be a combination of both, which imagines slow upheaval to be always going on with fits of more energetic action at intervals. Other instances might be given admitting of similar explanation, but this one must suffice here. The moral of all would be, let us be very careful how we take our own epoch as necessarily the type of all time past and to come. Experience must form the basis of our speculations, but we may fall grievously into error if we make it the limit of them. Concluding Remarks on Speculative Geology. 697 We give in conclusion the titles of a few of the more important memoirs touching on the subjects which have been glanced at in this section. * * Sir W. Thomson, On the Secular Cooling of the Earth, Trans. Royal Soc. of Edinburgh, xxiii. 157, and Natural Philosophy, Appendix D ; On Decrease in the Length of the Day owing to Tidal Friction, Natural Philosophy, arts. 276, 830; On Dates from Terrestrial Temperatures, British Association, 1855, Trans. Sections, p. 18 ; On Geological Time, Trans. Glasgow Geol. Soc. iii. part i. ; On Geological Dynamics, ibid, part ii. ; Professor A. Geikie, On Modern Denudation, ibid. iii. 153 ; Professor Huxley, Anniversary Address, Quart. Journ. Geol. Soc. xxv. ; Professor Ramsay, On Geological Time, Proceedings Royal Soc. xxii. pp. 145, 334 ; Mr. C. Sorby, Nature, ix. 388. CHAPTER XV. ON CHANGES OF CLIMATE, AND HOW THEY HAVE BEEN BROUGHT ABOUT. " These changes in the heavens, though slow, produce Like change on sea and land." MILTON. OF the many remarkable events which the study of geology assures us have taken place during the past history of the earth, none perhaps are more unlocked for, or more startling when the proofs of their occurrence are fairly established, than the changes which the climate of the same spot has undergone. We find for instance in North Greenland, Spitzbergen, and other countries, where now the rigours of an Arctic winter are scarcely relaxed all the year round, and where the presence of a living forest-tree is a sheer impossibility, the fossil remains of an abundant and varied flora, including poplars, willows, beeches, oaks, and other trees which grow only in temperate regions, and some which perhaps indicate even a more genial climate still. And it is likely that this elevation of temperature was not a mere local accident, for it w r as possibly about the same time that trees pointing to a sub-tropical climate abounded in Switzerland, Germany, and Devon- shire. The time during which the Polar regions enjoyed this mild climate is known as the Miocene epoch. At a somewhat later date a change exactly in the contrary direction was brought about, and the severity of Arctic regions was extended down to latitudes which now enjoy temperate conditions. Scotland w r as pretty much in the same condition as Greenland is now, the hill countries of the Lake district of England and North Wales nourished large glaciers, and the ice-flows of the Alps and other mountain-ranges pushed their way far beyond the limits which restrict their puny representatives of the present day. The time during which this happened is called the Glacial epoch. It is possible that there may have been other glacial epochs of older date, and we will distinguish this last as the Great Ice age. When the subject of change of climate first began to attract attention, regard was paid almost exclusively to those cases where it could be shown that the temperature of a country had been formerly higher than now, and it was somewhat hastily assumed that the alteration had been all along in the same direction and had consisted in a gradual lowering of the mean temperature of the globe; and this result was On Changes of Climate. 699 assumed with equal haste to have been brought about by that gradual cooling which the earth, if it had been originally in a fused condition, must of necessity be constantly undergoing. The former Arctic con- dition of Europe was ignored, either because its existence had not been placed beyond question or because it was supposed to be due to some special and exceptional cause. But we now know that such a view is altogether mistaken. The second instance just given of a climate different from that of the pre- sent day, shows that so far from the temperature having steadily declined as time went on, in one case at least the contrary has taken place. Our own country, after having experienced the severity of an Arctic climate, has now returned to more favourable conditions. And as the progress of geological inquiry has gone on it has been found that possibly this may not be an isolated instance. The evidence is far from conclusive, but some facts tend to show that alternations of genial and severe climates may have been repeated over and over again during bygone ages. The grounds for this assertion cannot be given till we come, in the second part of this Manual, to review the course of events, which a study of the rocks of the earth's crust shows to have accompanied their formation ; but the causes which may give rise to oscillations of climate can be fully understood at this point of the reader's studies, and may be conveniently considered here. Of the many solutions which have been offered of the problem, How have past changes in climate been brought about 1 only three seem to have a sufficient show of probability in their favour to call for detailed notice. One of these supposes that a distribution of land and water, differing from that which now exists, caused corresponding differences in the distribution of climate ; the other looks to certain changes which are constantly taking place in the position of the earth's axis and the shape of the earth's orbit, for the producing causes ; a third hypothesis combines these two views and holds that it is by the joint action of astronomical causes and geographical changes that variations in climate are brought about. Effect of Geographical Changes on Climate. That the distribution of sea and land affects to a very important extent the climate of different portions of the earth is beyond question. Turn to a map of what are called isothermal lines, that is lines passing through all the points in each hemisphere which have the same mean tempera- ture. If the temperature at any spot depended only on the amount of heat which that spot received from the sun, these lines would be parallel to the equator. But such is by no means the case; the isothermals are curves of the most complicated character, now stretch- ing away northwards in long loops, and again deflected southwards by broad sweeps, and ever and anon doubling back upon themselves in apparently the most arbitrary manner. But these aberrations are all capable of explanation. Some of the most striking bends are due to the influence of ocean-currents, and no instance of this kind is more marked than where, in the North Atlantic, the isothermals are pulled out in long folds to the north-east, and a most wonderful difference in 700 Geology. climate is produced between the eastern coast of North America and the opposite western shores of Europe. The mean January temperature of New York for instance is 32 ; that of the opposite coast of Portugal about 56. Labrador, in lat. 53, has a winter temperature of zero ; that of the shores of the north- west of Ireland, on the same parallel, is about 46 ; so that while the first is almost permanently cased in ice, water but rarely freezes on the second. And the same difference is maintained as we go northwards ; in fact, on our side of the ocean we must go as far north as Iceland before we meet with a winter temperature as low as that of New York. Now this marvellous contrast is due partly to the fact that a stream of cold water from Arctic seas, the Labrador Current, is always passing down along the eastern coast of North America, and still more to the fact that another current, the Gulf Stream, is always bringing from the tropics an enormous mass of heated water to bathe the western shores of Northern Europe. Now it is perfectly possible to conceive some change in physical geography, such as the upheaval of a barrier of land or the opening of a new passage, which would prevent the Gulf Stream from entering the North Atlantic or would lead it off into another channel. In such a case the western shores of Europe would no longer enjoy their present happy fortune, and our own country would suffer somewhat the same extremities of cold that now prevail in Labrador. Again, the distribution of land and sea affects the temperature inde- pendently of the effect it has in determining the course of currents. In the interior of large masses of land the summers are excessively hot and the winters as abnormally cold ; on sea-coasts and in insular regions there is far less contrast between the seasons ; so that by breaking up a continent into islands, or by allowing arms of the sea to gain access to its interior, we might very materially improve its climate. Land and sea also produce effects on the climate of regions at a distance by means of the influence they bring to bear on the winds which blow over them. For instance we have already mentioned that there was a time when the Alpine glaciers were far larger than at present ; at that time what is now the Sahara was covered by water ; the winds then that reached Switzerland from the south sucked up vapour as they blew over this broad expanse of sea, and came laden with moisture which was precipitated as snow when they came against the cold mountain-sides ; hence the accumulation on the gathering- ground was increased and larger glaciers were needed to relieve it. Now southerly winds blow over a parched desert, and not only bring no moisture with them but by their warmth tend to melt the ice, so that there is a smaller supply of the material for glacier-making and an agency tending to diminish what glaciers there are. Led by considerations such as these, many geologists, specially the late Sir C. Lyell, have believed that even the most extreme revolutions in climate can be accounted for by changes in the distribution of land and sea.* * Principles of Geology, vol. i. chap. xii. ; Hopkins, Quart. Journ. Geol Soc. viii. 56. On Changes of Climate. 70 1 But before we can admit that any actual changes in climate have been caused in the way Sir C. Lyell supposed, we must be satisfied on two points : first that there is evidence that the hypothetical distribu- tion of land and sea invoked to account for them did really exist at the periods in question ; and secondly that if it did, it was competent to produce the effects assigned to it. Let us see how far the hypothesis will stand these tests in the cases of the genial climate of Miocene times and the severity of the Great Ice age. Now according to this explanation the mild period was caused by the land being gathered around the tropics and the Polar regions being largely occupied by sea. This certainly does not seem to have been the arrangement that prevailed during Miocene times ; the European deposits of that date are mainly of lacustrine or shallow- sea origin, and point to the presence, not of large areas of sea, but of extensive tracts of continental land. Again, would an accumulation of land about the equator give rise to a genial climate over the whole globe ? The theory we are considering says it would, and in this way. The land, being highly heated by the tropical sun, would in its turn heat the air, which would rise and flow towards the poles, and thus there would be a constant transfer of heat from the equatorial to the Arctic regions. That this atmospheric cir- culation must always go on, and that it would go on in the supposed case to a larger extent than now, cannot be denied ; but Dr. Croll has shown that it is very doubtful whether these aerial currents would avail anything towards mitigating the severity of the Polar climate. However hot the wind might be when it left the land, it would be liable to rise to heights where the temperature is below the freezing- point ; all its warmth would then be stolen from it long before it reached its journey's end, and it would come down to the earth's surface in northern latitudes as a chilling and not a warming current. The proposed arrangement of land and sea might therefore bring no additional heat to Polar regions ; what is worse, it might hinder the flow of warm ocean-currents from tropical regions towards the poles, and so might put a stop to the working of the machinery by which equatorial warmth is now largely distributed over the globe, and by the agency of which many regions that would otherwise be icy wastes are rendered habitable. For Dr. Croll has shown that it is not currents in the air, but currents in the ocean, that are now performing this beneficent task. Wherever streams of heated w r ater flow northwards from the tropics and spread out as they advance, they diffuse heat from their broad warm surfaces into the air above, and give rise to warm winds the softening influences of which are felt over the adjoining countries. A great belt of equatorial land might materially interfere with these currents which at present all take their rise in the Southern Hemi- sphere, and might cut off the supply of heat they are always bringing to alleviate the rigours of Arctic regions. As far then as accounting for the mildness of Miocene and other genial epochs goes, Sir C. LyelPs arrangement would be very liable to fail. The redistribution of land and water, which he supposed brought about the cold of the 702 Geology. Great Ice age, would have a tendency to produce the effect he supposed, but there are no grounds for believing that it existed at that time. Besides if we accept the doctrine of the Permanency of Continents and Oceans, such a radical change in the distribution of land and sea as Sir C. Lyell's view requires could never have happened. But though the special rearrangements of land and water devised by Sir C. Lyell are in themselves improbable and are not competent to bring about all the results he anticipated, it by no means follows that geographical change has not been an important factor in producing changes in climate. Allowing that continents and oceans have been all along substantially what they are now, it is possible, perhaps even probable, that at former epochs the great masses of land were less compact than at present arid more cut up by arms and inlets from the oceans. Under such conditions warm ocean-currents flowing north- ward would be able to penetrate to a much larger extent than now into Arctic regions and would mitigate very considerably the severity of Polar cold. Close the channels and the rigorous climate returns. And the oscillations of level required for this end would not be more serious than those which we know for certain have frequently happened. In this and in other ways it is perfectly possible for very important modifications of climate to be produced by geographical changes. Astronomical Causes affecting Climate. We will now turn to the second view. This explanation was first suggested by Sir J. Herschel,* but he seems afterwards to have given it up ; it has since been worked out in very full detail by Dr. Croll.t It may save the reader the trouble of reference to a book on astronomy if we recount shortly the astronomical changes which this explanation looks upon as the ultimate causes of change in climate. The path which the earth describes round the sun is a plane curve, called an ellipse, such as ABPD in fig. 232. If drawn truly to scale, the real path would scarcely be distinguished by the eye from a circle, and therefore it is in the figure made much more oval than in nature, lest the reader should suppose it was actually circular, C is the centre, A CP the longest, BCD the shortest diameter. The sun occu- pies a point S on C P, called the focus. P is called the perihelion or point nearest to the sun ; A the aphelion, or point farthest from the sun ; SP the perihelion distance, SA the aphelion distance. Now there are two things we have to note about the path : it is constantly undergoing changes both in shape and position. First with regard to the change in shape ; if the earth and the sun were the only bodies in the universe, the former would always pursue exactly * Proceedings Geol. Soc. i. 244. t Dr. Croll's researches were first published in the fourth series of the Phil. Mag. He has iu Jukes' Manual of Geology suggested the following as the order in which his papers may be most profitably read : On Geological Time, etc. xxxv. 363 (May 1868) ; xxxvi. 141, 362 (August, November 1668) ; On Ocean-Currents, part i. 'xxxix. 81 (February 3870) ; part ii. xxxix. 180 (March 1870) ; part iii. xl. 233 (October 1870), xlii. 241 (October 1871), xlvii. 94, 168 (February and March 1874) ; On supposed greater Loss of Heat by Southern than by Nor- thern Hemisphere, xxxviii. 220 (September 1869). The reader will find the sub- stance of these papers and much additional matter in Dr. Croll's work, Climate and Time in their Geological Relations, a Theory of the Secular Changes of the Earth's Climate. On Changes of Climate. 703 the same path round the latter year after year ; but the attractions of the other planets are always pulling the earth now this way and now that, and in this manner it comes about that the shape of its path is constantly changing at a very slow rate, so that it is at one time more oval than at another. The changes in shape can never go beyond certain fixed limits. For a long series of ages the orbit goes on getting more and more oval or elliptical ; then the ellipticity begins to decrease and the orbit grows more and more nearly circular ; but before it becomes actually a circle the ellipticity begins again to increase, and it keeps increasing for another long epoch, when it again turns back and begins again to grow less. This is the general nature of the change in shape of the earth's path ; but we must yet consider one or two particulars more exactly. The longest diameter, PA, is always the same, and hence we can make the Fig. 232. ORBIT OF THE EARTH, ECCENTRICITY SMALL, WINTER OCCURRING IN PERIHELION. orbit more elliptical only by making EC shorter ; in fact the orbit, while its length remains unaltered, is at some times flatter than others. But the line ES is equal to half the longest diameter, and must therefore always remain the same length whatever change goes on. Now if E comes nearer to (7, ES can keep the same length only by 8 moving towards P. Therefore when the eccentricity is large, the sun is nearer to the perihelion than when it is small. Increase of eccentricity therefore diminishes the perihelion distance, and increases the aphelion distance. If the reader will compare figs. 232 and 233, he will realize the effect of the change ; in both the longest axis of the ellipse is the same, but in the second the curve is more elliptical, the perihelion dis- tance SP is less, and the aphelion distance SA is greater than in the first. He must not forget however that in both figures the ellipticity is far greater than in the actual case. Geology. Secondly, besides a change in shape, the path of the earth is under- going a constant though slow change in position. If at any date the direction of the line PSA be determined, say by noting that it points directly to a particular star, and the observation be repeated after a time, we shall find that the line no longer points to the same star but has moved away in the same direction as the earth revolves. This motion is called the Revolution of the Apsides, and by it the point A is carried round the whole orbit in 109,830 years. Such are the facts we shall have to bear in mind respecting the alteration in shape and the change in position of the earth's orbit. We have now to pass to a further point. A plane through the sun parallel to the plane of the earth's equator is called the celestial equator. If the line of intersection of the celestial equator and the plane of the ecliptic meets the earth's orbit in AE, VE, these points are called the Autumnal and Vernal Equinoxes. If a line through S perpendicular to AE, VE cuts the earth's orbit in WS, SS, these points are called Fig. 233. ORBIT OF THE EARTH, ECCENTRICITY LARGE. the Winter and Summer Solstices. When the earth is at either of the equinoxes, the days and nights are everywhere equal in length ; as the earth moves from the autumnal towards the vernal equinox the pole is turned away from the sun and the nights are always longer than the days, the difference between day and night being greatest at the winter solstice ; as the earth moves from the vernal equinox towards the autumnal equinox the pole is turned towards the sun and the days are longer than the nights, the longest day occurring as she passes through the summer solstice. In other words, the time taken by the earth to travel from AE to VE is the winter portion, and the time from VE to AE is the summer portion of the year. Now it is very easy to see that, as long as the earth's path is not a circle, the summer and winter portions of the year must be of different lengths. Look at fig. 232, which represents pretty nearly the present state of matters for the Northern Hemisphere. The arc AE, P, VE is shorter than the arc VE, A, AE, and, what is more, the earth moves faster over the first arc than over the second, because she On Changes of Climate. 705 moves faster the nearer she is to the sun, so that both these causes now work together to make our summer longer than our winter. Further note that not only is our winter now shorter than our summer, but the earth is nearest to the sun nearly at mid-winter, and the additional amount of heat thus obtained tends to mitigate the severity of the cold season. The Northern Hemisphere now, therefore, is well off as regards climate for two reasons its winter is short, and it is nearest to the sun in winter ; the Southern Hemisphere is badly off, for its winter is long, and it is farthest from the sun in winter. But now comes a point of the utmost importance : it has not always been so. We have already mentioned the motion of the earth's axis known as precession, and explained how that line is constrained to move slowly round, sweeping out a path in space like the surface of an inverted sugar-cone. Now since the plane of the earth's equator is Fig. 234. ORBIT OF THE EARTH, WINTER OCCURRING IN APHELION. perpendicular to the earth's axis, if the axis moves, the terrestrial equator, and therefore the celestial equator too, must move with it ; and a very little reflection will show that in consequence of the revolu- tion of the earth's axis the line AE, VEw\\\ turn slowly round S as a centre. The motion takes place in the direction opposite to that of the earth's revolution, and the line makes a complete circuit in 25,868 years. The line AE, VE is turning then at this rate in one direction, and the line PSA in the opposite direction at a rate which carries it through a whole revolution in 109,830 years ; a short calculation* will * One line moves through ^-g-^sth of an entire revolution in a year, the other moves through -rWrstfth f an entire revolution in a year in the opposite direc- tion. Therefore they separate by ^i^ + iui*TO of an entire revolution in a - TTnrrstf Tmhrr Hence they come together again in 20,937 years. 2 Y 706 Geology. show that if we take any position of these shifting lines, say that in fig. 232, after a lapse of 20,937, or nearly 21,000 years, they will come round to the same position again, and in half that time we shall have a state of things like that shown in fig. 234, where the positions of the equinoxes and solstices are exactly reversed, and where the win- ter in the Northern Hemisphere is longer than the summer. This will be the case with our hemisphere some 10,500 years hence, and we shall then be exactly in the position the Southern Hemisphere is in now. The effect then of precession and the revolution of the apsides is this. Mid-winter will occur at certain periods for each hemisphere when the earth is in perihelion, and the winters will then be short and their severity mitigated by the proximity of the sun ; about 10,500 years after each of these periods, the mid- winter of the same hemisphere will happen when the earth is in aphelion, and the winter will then be long and rendered more severe by the increased distance of the sun ; the summer in the latter case will be short ; and at first sight we might think that it would be also hot because of the near approach to the sun, but we shall see shortly that there are causes which prevent this circumstance from exercising any beneficial effect on the climate. Now as long as the path of the earth deviates at all from a circle, the effects just described must be produced ; even when its eccentricity is small, as it is at present, the hemisphere whose winter occurs at perihelion must have some advantage over the opposite hemisphere ; and the greater severity of the climate of the Antarctic regions at the present day is doubtless partly owing to the winter of the Southern Hemisphere falling now very near aphelion. But the contrast will be evidently immensely greater when the eccentricity is large. Compare figs. 232 and 233. Everything that tends to mitigate the severity of the winter in the first is present in a more pronounced form in the second ; the actual length of the winter is less and the distance from the sun in mid-winter is decreased. To take an instance, our winter is now nearly eight days shorter than the summer ; but if the eccen- tricity had its greatest value and our winter occurred in aphelion, not only would the length of winter exceed that of summer by thirty-six days, but we should be more than eight and a half million miles farther from the sun in winter than we are now. If therefore these celestial changes have anything to do with climate, it will be during periods of high eccentricity that they will produce their most telling effect. At such times the hemisphere whose mid- winter occurs in perihelion will have so short and mild a winter and so long and moderately hot a summer, that its climate will be something like a perpetual spring. The opposite hemisphere will have a long, severe winter and a short summer ; and these conditions will be transferred from one hemisphere to the other every 10,500 years. Some periods of high eccentricity have lasted long enough to allow of such a transfer having taken place several times over. Thus much was pointed out by Sir J. Herschel, in the paper already quoted, in 1830; and he then expressed it as his opinion, that during On Changes of Climate. 707 a period of high eccentricity the effect of these secular changes would be to place each hemisphere alternately in a state approaching per- petual spring, and under a condition of burning summers and rigorous winters. He seems afterwards however to have felt that long periods of severe cold could not have been brought about by these causes, because, however contrasted the seasons might be, the deficiency of heat during a long winter would be made up for by the large amount received during the short but hot summer. In fact the total amount of heat received during a revolution of the earth increases as the smallest diameter of her orbit decreases, and it might therefore seem at first sight as if periods of high eccentricity would give rise to an increase in the general warmth. But Mr. Croll took up the subject, and showed that though these cosmical changes could not directly be the cause of epochs of intense cold, they might produce this result indirectly in the following manner. The dreary winters, which will be the rule whenever the eccentricity is high and the winter comes round when the earth is near aphelion, will be long enough to allow of enormous quantities of snow and ice gathering on land and sea every winter. At the same time, during the summer, the earth, on account of its closer approach to the sun, will receive a larger amount of heat than at present ; but the summers will be so short that, even with this advantage, and supposing there was nothing to prevent the sun from exerting its full power in melting, there will not be time during the lapse of a summer for the whole of the accumulation of the preceding winter to be cleared away. The efforts then made every summer to get rid of the frozen matter will never be able to keep pace with the additions of winter, and at the end of each summer there will always be a balance of unmelted snow and ice to carry forward to the next winter's account, and the piles will grow year by year till broad areas become permanently wrapped in sheets of ice of enormous thickness. This cause alone would favour the accumulation of great masses of ice and snow; but there are other causes which tend in the same direction and prevent the sun from exerting its full effect in the work of melting. The presence of great masses of snow and ice will tend to keep down the summer temperature, or rather they will result in making the existence of anything deserving the name of summer impossible, in spite of the large amount of heat poured on to the earth during the part of the year which corresponds to summer. The power of the sun to heat any substance depends on the amount of sun-heat which that substance can absorb or appropriate to itself. Now air can absorb scarcely any of the direct heat of the sun, and consequently the sun's rays pass through it without raising its temperature to any appre- ciable extent. Many curious and apparently contradictory facts can be explained when this powerlessness of air to absorb sun-heat is taken into account. The pitch on a ship's side off the Greenland coast has been melted by the direct rays of the sun, when the tempera- ture of the air around was far below the freezing-point. The air could not take up any of the heat, but the pitch could. So when the sun's rays fall upon the ground, they meet with a substance that can ;o8 Geology. absorb them ; the earth becomes heated, and in its turn radiates or gives off heat to the cold air above. Now the heat radiated from the ground differs from that which comes direct from the sun in this : it can be absorbed by the aqueous vapour of the atmosphere, and it is taken up greedily, and, raising the temperature of that vapour, produces a generally genial climate. But if a country be cased in snow and ice, there will be no heat absorbed and none given back to raise the temperature of the air : the sun's heat will be all used up in the work of melting, and, as long as the icy coating remains, the tempera- ture of the surface can never be raised above the freezing-point. In such a case the ground, instead of being a source of warmth from which heat is always passing off to warm the air above, is a cold pave- ment, which not only has no heat of its own to give away, but tends to rob the atmosphere of any warmth it may have obtained from other sources. Again, the sun's rays when they fall on the bare ground are very largely absorbed ; but from surfaces of snow or ice a great portion is reflected back and lost to the earth altogether. The beneficial effect which a nearer approach to the sun would tend to produce would be further neutralized in this way. The increased Fig. 235. heut would give rise to abundant evaporation, but the chilling effect of the cold air and icy masses would condense the watery vapour and give rise to dense fogs, which would cut off the sun's rays and prevent any melting of the snow perhaps all the summer long. Here, one would think, we have enough to produce any amount of severity of climate ; but Mr. Croll believes that there is yet another cause that would produce still more important effects. He holds that the great currents of the ocean are due to the pressure of the trade-winds on the surface of the water. These trades are caused by the difference in temperature of the air in polar and equatorial regions, and if the mean temperature of one hemisphere be lower than that of the other, the trades from the first will be stronger than those from the second. Owing to this cause the south-easterly are now more powerful than the north-easterly trades, and in consequence the general set of ocean- currents is towards the Northern Hemisphere. The general tendency is thus for the warm equatorial waters to be carried northwards, and raise the temperature of those northern lands whose shores are washed by them, or across which winds blowing athwart the course of the warm currents are wafted. But when the Northern Hemisphere was under glacial conditions, the Southern Hemisphere would be enjoying a mild climate all the year round, and the present arrangement of On Changes of Climate. 709 currents would be exactly reversed. The warm equatorial water would flow southwards, and our hemisphere would lose all the benefit it now derives from this source. If the explanation just given be cor- rect, alternations of periods of intense cold and of periods when a mild equable temperature prevailed over an entire hemisphere, must have recurred during the past history of the earth over and over again. Further, if we tabulate the values of the eccentricity for past epochs, and note the points at which after increasing for a time it begins to decrease or the contrary, we shall find that its values at these turning-points are by no means all equal, and also that the periods during which the eccentricity keeps at a high or a low figure are in some cases very much longer than in others. Suppose we take a straight line and divide it into a number of equal parts, each of which represents a year, and from each of these points erect per- pendiculars, making the length of each perpendicular proportional to the value of the eccentricity at the date corre- sponding to the point from which it is drawn, and then draw a curve through the extremities of the perpendiculars, the shape of this curve will give us an idea of the nature of the changes in the eccentricity. We shall find that we do not get a series of regular arches each of the same breadth, and each rising to the same height above AB, like the curve in fig. 235, but a curve like that in fig. 236, where the summits of the bends are some much higher than others and the intervals between the bends very un- equal in length. Hence the cold periods will be very unequal in length, and will occur at very unequal intervals. Evidence for and against the preceding Hypothesis. If, as the views just enunciated maintain, astronomical changes have been the only or the main cause of changes of climate, then we ought to find, 1. That periods of extreme cold have recurred many times over. Geology. 2. That the severe climate of these periods extended over the whole of one hemisphere. 3. That periods when there was something like spring all the year round, which may be called "mild interglacial periods," alternated with periods of severe climate. 4. That while one hemisphere was enjoying a genial climate the other was passing through a time of extreme cold. It remains to inquire whether we have evidence that events of this character have actually occurred. We may at once say that we can hardly expect to find proofs that will satisfy us on the fourth head. Cold and warm intervals would according to Dr. Croll's hypothesis follow one another at intervals of 10,500 years. We have no means of fixing the date of any geological occurrence within anything like limits even as wide as this. Two events might be separated by an interval of 10,000 years or more; and yet, as far as our means of determining their date goes, they might be practically contemporaneous. We do for instance find con- clusive proofs that once at least the climate of large portions both of the Northern and Southern Hemispheres was far more severe than it is now, and we can safely say that in both cases this took place at a time geologically recent ; but whether the two cold periods were exactly coincident, or whether they were separated by an interval even as long as 10,000 years, the evidence at our disposal gives us no means of deciding. Passing to the other three tests, we have to note that proof of the prevalence of severe climate may be obtained in two ways. We may find in one group of rocks the remains of animals that live in warm regions, and in the group above these may be replaced by forms which are found only in cold localities. This test we can apply only to formations of so recent a date that their fossils are either identical with recent forms or closely allied to them ; and we can use it with absolute certainty only in the case of perfect specific identity. There will always be some risk of error in assuming that animals and plants allied to recent forms, however close the alliance may be, required the same climate and surroundings as their living relatives. Elephants for instance are now confined to tropical regions, but an extinct elephant, the Mammoth, lived far away north. There has been no mistake in this case, because by a lucky accident the hairy coat of the Mammoth, which enabled it to withstand cold, has been preserved. If only its bones had come down to us, we might have supposed that they indicated a warm climate. The second test is of more general application. Where deposits like Boulder Clay or Till occur, or where the rocks of a country show wide- spread glacial scratching or polishing, there can be no doubt that the climate of tha spot where these are met with was severe enough to give rise to Ice-sheets and Glaciers at the time when they were pro- duced. But the converse proposition is hot necessarily true. The absence of those proofs of glaciation by no means implies the absence of a severe climate. For the accumulation of large masses of ice two things On Changes of Climate. 7 1 1 are necessary besides a low winter temperature, the neighbourhood of a large expanse of water from which moisture may be drawn by evaporation, and large tracts of elevated land to condense that moisture into snow. Unless all three are present, the climate may be bitterly cold and yet neither Ice-sheet nor Glacier may form. In a word, we must have a combination of what may be called geographical conditions with what may be called climatic conditions in order to get glaciation. Let us now work our way backwards through geological time, and see whether we have traces of the occurrence of the events which Dr. Croll's hypothesis demands. We come first to the Great Ice age which belongs to the latest of geological epochs. Here we have undoubted proof both in Western Europe and in Eastern America of glaciation of the severest character, and there is also reason to think that warm intervals alternated with periods of extreme cold. It is true that the signs of glaciation gradually fade away as we go eastwards in Europe and westwards in America, but they reappear on the Pacific coast of the latter continent. Sarcastic objections have been thrown out on this ground against the hypothesis, and attempts have been made to show that the ice-sheets of the Great Ice age were mere local phenomena caused by special geographical circumstances. But we see at once the fallacy of such criticism if we bear in mind that the portions of the Northern Hemisphere which were not covered by ice are just those regions where two of the necessary conditions for the production of ice-sheets, the neighbourhood of the ocean and elevated land, are absent. It is a case in which the absence of glacial deposits and markings proves nothing'either one way or the other. The climatic changes then of the Great Ice age seem to find a satis- factory explanation in Dr. Croll's hypothesis. Among the older formations of the earth's crust we meet with rocks of various ages which can hardly have been formed in any other way than by the agency of ice. These have been cited as indications of a succession of glacial periods, and have therefore been held to tell in favour of Dr. Croll's hypothesis. The evidence is not of a very con- vincing character : these old glacial deposits are local and fragmentary, and it is perfectly possible that instead of being relics of former periods of universal cold they may be the result of special geographical con- ditions which led to local accumulations of ice at the spots where they occur. Either explanation is tenable, and it is quite impossible to say which is correct. Their present restricted range tells neither way : glacial deposits even during a glacial epoch are necessarily restricted to these spots where the geographical conditions permit of the accumulation of ice : glacial deposits are largely terrestrial, and are therefore specially liable to be cut up by denudation ; for these and other reasons we can only expect that fragments of the deposits of any old glacial period will come down to us even though these deposits may have been once as extensive as those of the Great Ice age. The patchy nature then of these old glacial desposits does not neces- sarily disprove Dr. Croll's hypothesis ; at the same time the amount of support they lend to his views cannot be said to be very large. 7 1 2 Geology. Two of these cases call for rather more detailed notice, because they have been held to furnish very damaging evidence against the hypothesis we are considering. During the Miocene period the climate of Europe was sub-tropical and genial conditions prevailed up to the Polar regions ; during the Eocene period, which preceded, the average temperature of Europe was higher still. At one spot near Turin there are deposits, which everybody admits to be glacial, of Miocene age ; and in the Alps and the Carpathians there are conglomerates and breccias of Eocene date which are certainly ice-formed. These cases have been quoted by Dr. Croll's supporters as proof of alternations of glacial and mild inter- glacial periods during Miocene and Eocene times. Further they point out that there are two long periods of high eccentricity, which, assum- ing them to coincide with the Miocene and Eocene epochs respectively, would bring about this rotation of climates. Dr. Croll's opponents retort that these glacial beds are too local to justify us in pointing to them as evidence for a period of universal cold, and that they occur in the neighbourhood of great mountain-chains where glaciers are liable to be met with at all times. In reply it may be urged that localities such as these are the only spots where ice could gather even during a glacial period, and further that the Alps had not come into existence in Eocene times. So far as much may perhaps be said on one side as the other, but further arguments are advanced by the opponents of the hypothesis. If the ice-formed conglomerates of the Alps were accumulated during a portion of the Eocene period when glacial conditions prevailed over the Northern Hemisphere, we ought to find traces of this cold epoch elsewhere in Eocene strata. Now we nowhere find in the Eocene beds of England, nor in those of the more low-lying parts of Europe or of North America, strata which show any indication of ice-action. This negative evidence is for reasons already given of doubtful value ; the absence of glacial deposits may be due to the absence of the necessary geographical conditions. But an objection based upon palaeontological evidence is more serious. Periods of extreme cold, if they leave no other mark of their presence, would yet be indicated by the effect they would produce on the life of the epoch. We see this very clearly in the deposit, known as the Crag, which immediately preceded the Great Ice age. The coming on of the cold is clearly foreshadowed by the gradual disappearance of the shells that live in warm seas and by a corresponding increase in the percentage of Arctic forms ; when the cold reached its height, fossils disappear altogether. But we find nothing corresponding to this throughout the whole thickness of the Eocene strata ; the fossils from top to bottom are such as indi- cate a decidedly warm climate. The sequence of deposits too is singu- larly full and perfect, there is nothing to lead us to suspect unrepre- sented intervals, and the beds are largely charged with fossils through- out. The time represented by this group of strata is very long and alternations of severe and mild climates must have occurred during it several times over according to Dr. Croll's hypothesis. The evidence is of a negative character and it is always dangerous to attach much On Changes of Climate. 7 1 3 value to negative evidence in geological speculation ; another weak point is that the majority of the fossils belong to extinct species ; but allowing all due weight to these considerations, the absence of any trace of serious* changes in climate, if such really occurred, is hard to get over. A similar difficulty meets us when we come to examine into the climate of the Arctic regions during past ages ; the rocks of several periods are there well represented by fossiliferous strata, and in every case the fossils, as far as we can judge from extinct species, indicate a mild and genial climate. If, as is natural, we look upon the present state of the Arctic area as its normal condition, these warm periods require explanation, and Dr. Croll accounts for them by supposing that they were the warm interglacial periods of a time of high eccentricity. If this be so, periods of excessive cold must have intervened, and we are bound to ask whether there are any traces of these. Those best fitted to have an opinion on the point say that they cannot find any ; indeed the testimony of palaeontology is so uniformly in one direction in this matter that some geologists hold that the present state of the Arctic area is exceptional, that during the whole of geological time it has been blessed with a temperate, at times perhaps with a climate even warmer still, and that it is only since the Great Ice age that it has become an icy waste. The geology of the Polar regions is perhaps scarcely known com- pletely enough to justify us in looking upon this objection as fatal, but what we do know is quite enough to make us pause before w r e accept Dr. CrolFs hypothesis as a full and sufficient explanation of all past changes of climate. The Effect on Climate of Geographical and Astrono- mical Causes combined. The means suggested by Sir C. Lyell and Dr. Croll seem neither of them competent to produce alone all the changes of climate which we know have occurred ; possibly their joint action might suffice. It is in this light that the question has been looked at by Mr. Walla ce.t That the astronomical changes relied upon by Dr. Croll would tend to produce the effects he assigns to them, cannot be doubted, but the results might evidently be enhanced very materially if there were at the same time a distribution of land and water tending to produce change of climate in the same direction ; on the other hand it is possible that geographical conditions might be powerful enough to modify or even neutralize altogether the influence of astronomical changes, if the two were opposed to one another. When both pull the same way there can be no doubt about the result ; when they pull opposite ways, it is not so easy to see which will be master. * That there were alternations of climate is highly probable, but they seem none of them to have been sufficient in amount to produce glacial conditions. The astronomical causes which tend to modify climate were doubtless at work, but they were prevented from producing their full effect by geographical causes acting in an opposite direction, as will be explained more fully further on. t Island Life, chaps, viii. and ix. 714 Geology. Mr. Wallace holds by the doctrine of the Permanency of Continents, and according to it maintains that the land surfaces of the globe have always consisted of three great masses in the North Temperate Zone, narrowing southwards and terminating in three comparatively narrow extremities represented now by South America, Africa, and Australia. A large tract of land has always existed around the South Pole washed all round by a broad expanse of sea ; the North Polar regions have been all along mainly occupied by water. This arrangement would favour the permanent accumulation of ice around the South Pole, but it would be unsuitable for the gathering of ice in the North Polar Zone except under such special and excep- tional conditions as exist in Greenland at the present day. The Nor- thern Hemisphere would therefore always tend to be warmer than the Southern, and this would cause a steady set of ocean-currents to the north that would be constantly carrying warmth towards the Polar regions. If further the continental masses of land have been growing more solid and compact as time went on, there must have been in former days more channels of communication between the tropical parts of the ocean and the Polar seas, and the heat-bringing currents must have had far readier means of access to Polar regions than at present. As far then as geographical causes go there is everything to bring about that genial condition of the North Polar Zone which geological evidence seems to show has continued without break down to the commencement of the Great Ice age. The question is, Would astro- nomical causes be able to counteract geographical influences 1 Take first the case when during a period of high eccentricity the northern winter occurred in aphelion. The southern winter would then be short and mild, the summer would be long but it would not be very hot, and it may be questioned whether under these circumstances the ice would be melted off the South Polar land. If it were not, ocean- currents would still set northwards, and as long as they poured into the North Polar basin their warming influence might well keep up the temperature even during the long dreary winter far enough to prevent any large accumulation of ice. If ice did not form to any great extent, the summer though short might be hot enough to melt it all. It is possible then that geographical conditions might completely counteract the tendency of astronomical causes to produce a glacial period in the Arctic Zone, and that as long as the arrangement of land and water which is now contemplated existed, a glacial epoch was an impossi- bility in the Northern Hemisphere. When the northern summer came to occur in aphelion, geographical and astronomical causes pulled together, the climate of Arctic regions became more genial, and the rigours of the Antarctic cold were increased. But if, as the growth of the continents went on, the channels by which warm ocean-currents reached the Polar seas became more and more blocked, the features of the case become very materially changed. The mitigating influence of ocean-currents might be reduced to such an extent as to turn the scale in favour of astronomical causes, and with high eccentricity and a winter in aphelion, a climate might result On Changes of Climate. ^ 1 5 severe enough to produce glaciation in the Northern Hemisphere at spots where the necessary geographical conditions were present. It is impossible in a sketch like this to do justice to the wealth of illustration and argument with which Mr. Wallace has enforced his views, for these the reader must turn to his book, but I trust that the little I have said fairly represents his line of reasoning. He has shown that the geographical conditions which his hypothesis requires did actually exist during Tertiary times ; whether they were sufficient to overpower astronomical influence is perhaps a matter of opinion, but much may be said in favour of such a view. It is also a fact that before the Great Ice age came on the channels by which warm water from the tropics could reach Polar seas were all practically closed but one, and we can well realize that under these conditions astronomi- cal causes might get the upper hand and bring about a glacial epoch. With these facts before us we can see a reason why there may have been no glaciation during Tertiary times in the Northern Hemisphere and yet severe glaciation during the Great Ice age. We can hardly yet map out the distribution of land and water during the older geological periods accurately enough to say whether it has been such as to allow of astronomical causes producing a glacial epoch in the Northern Hemisphere at any previous date. Some day this may be possible, or the day may come when glacial deposits have been detected at so many different spots among the rocks of some epoch or epochs as to justify us in maintaining that a glacial period existed during those epochs, and then it will be probable that there was at that time an arrangement of land and water similar to that of the Great Ice age. The nearest approach that has as yet been made to this result is in the case of the Permian epoch, but even there the evidence is not conclusive. The question is very complicated, and when so many causes are at work, it is difficult to assign to each its due weight ; it is a very happy thing that the problem has been attacked by one who so thoroughly realizes its complexity, and who has reminded us that it is not to any single source that we must look for the cause of changes in climate. Other hypothetical Explanations of Changesof Climate. A change in the position of the earth's axis of rotation has been suggested as a possible cause of changes in climate. Of such changes in position there are two kinds. The earth may be shifted as a whole, and its axis with it, in such a way that what is called the Obliquity, that is the angle between the equator and the ecliptic, becomes altered. Nutation is a change of this kind due to the moon but astronomers apply the term Change in Obliquity to another oscillation in the relative position of these two planes caused by the disturbing action of the planets. In the other case there need be no change in the position of the earth as a whole, but the line about which it revolves shifts its position within the earth itself ; this will be equivalent to an alteration in the position of the poles, and, if it take place, it will be found out from the circumstance that days and nights of more than twenty-four hours long occur at spots where such had not been known before. 7 1 6 Geology. The variations of the obliquity are confined within very narrow limits; in 1801 it was 23 28', the greatest value it can have is according to Laplace 24 35' 58", the least 21 58' 36". But even these small changes produce some effect on climate : increase in the obliquity diminishes slightly the amount of heat received at the equator in a year, but it increases very sensibly the amount received at the poles. Suppose that the obliquity was at its maximum during a time of high eccentricity : other astronomical causes will tend to produce a severe climate in that hemisphere whose winter occurs in aphelion, the high obliquity will abate this tendency, for it will add to the amount of heat received by the Polar regions : at the same time the tendency to produce a mild climate in the other hemisphere will be increased. The opposite results will follow when the obliquity is at its minimum. Here then is a third, factor influencing climate, sometimes in one direction sometimes in the other : it may evidently modify the effects that would be produced by the two causes we have previously considered. That a change in the geographical position of the poles has been a cause of past changes of climate has been a favourite speculation ever since geologists began to exercise their ingenuity on this topic. It has not proved a speculation of much promise, for it is doubtful whether any such change would account for the facts, and it is perhaps more than doubtful whether any such change has happened within geological time. Mr. G. Darwin* has shown that during the consolidation of the earth there was probably great instability in the geographical position of the poles. When the earth had become practically rigid, the posi- tion of the poles would not be liable to be largely shifted, but it would still be possible by making enormous changes in the distribution of land and water to produce slight displacements of the poles. We may be quite certain that no geographical change of anything like this extent has taken place since the beginning of the Tertiary period, and therefore it is quite impossible that the genial climate of the Arctic regions during Miocene times can have been caused in this way. If we allow the Permanency of Oceans and Continents, geographical changes sufficient to shift the poles can never have taken place during geological time. On the supposition that the earth, when its shape is altered by erosion and deposition or by upheaval and subsidence, adjusts itself to a form of equilibrium, some considerable displacement of the poles might result. But this explanation cannot be applied to the Miocene case, because the amount of deformation necessary is far greater than can possibly have taken place. The investigations of the Rev. T. F. Twisden lead to the same con- clusions, t But even supposing a change in the position of the pole to the requisite extent to be admissible, it would not get us out of our diffi- culty. If what are now the Arctic regions are shifted into the North Temperate Zone, some part of that zone will become the new Arctic * Phil. Trans, clxvii. part i. 271. t Quart. Journ. Geol. Soc. xxxiv. (1878) 35. On Changes of Climate. 7 1 7 region. But all the evidence goes to show that the Arctic regions were not shifted, but were improved off the face of the earth altogether during Miocene times, for at that epoch the temperature of the Northern Hemisphere seems to have been everywhere higher than at present. That our sun is a variable star, and that the genial climate of the present Arctic regions during past time was due to an increased supply of sun-heat, has been hinted. In the same way it has been suggested that glacial epochs were caused when the sun's heat was on the wane. Neither explanation can be admitted. A general lowering of the temperature would not suffice to produce great sheets of land-ice ; great evaporation is the first requisite for that end, and a decrease in the sun's heat would check instead of promoting evaporation. Again if the improvement in the climate of the Arctic regions were caused solely by increased heating power in the sun, we should require a sun so hot that the greater part of the world would be uninhabitable. We may also mention that an attempt has been made to account for variations in climate by the hypothesis that our system in its travels through space has passed through regions which were at one time colder and at another warmer than that in which it is now situated. This is a purely arbitrary hypothesis, altogether unsupported by experience, and it is most improbable for the following reason. Increased heat or increased cold could have been caused only by the neighbourhood of a hot or cold body ; and a body large enough to produce any sensible effect in either direction would have disturbed by its attraction the stability of the planetary system, and would have introduced perturbations in the motions of the planets which would still be recognisable. But independently of its intrinsic improbability, the hypothesis would not account for the facts, for cold alone will not bring about glaciation. INDEX. ABRAUM salts, 241 Absence of pebbles on sea-bed, 576 Acanthite, 537 Acicular, 85 Acid crystalline rocks, 312 Actinolite, 107 Adamantoid, 48 Adularia, 97 Adur, R., 608 ^olian denudation, 202 Affleurement, 550 Agate, 91, 282 Agglomerate, volcanic, 359 Agh Sibyr, 381 Air-saddle, 477 Alabandite, 156 Alabaster, 140 Alaunstein, 159 Albite, 97 Alderley Edge, cupriferous sandstone of, 566 Allanite, 128 Allophane, 134 Allotropism, 85 Alluvial plains, 605 Alstonite, 145 Alston Moor, Wallace on lodes of, 559 Altenberg, tin-bearing Zwitter of, 562 Alteration of lava, 342, 343 Alteration of rocks by lava, 371 Alteration products, 120 Alumina, 13 ; blowpipe test for, 158 Aluminite, 159 Alum shale, 177 Alumstone, 159 Alunite, 159 Amas entrelace, 555 Amethyst. 158 Amianthus, 107 Amorphous metamorphic rocks, 427 Amorphous states, 82 Amphigene, 102 Amygdaloidal structure, 311 Analcime, 129 Anarnesite, 332 Anatase, 546 Anchor ice, 201 Andalucia, iron pyrites of, 564 ; old plain of marine denuda- tion in, 578 Andalusite, 124 Andesite, 97, 326 Anglesite, 532 Anhydrite, 14, 141, 186 ; changed into gypsum, 412 ; consequent increase in bulk, 413 ; formation cf, 248 Animal-tracks, 217 Anisotropic, 72 Ankerite, 138 Anorthite, 97 Anthracite, 181, 185 Anticlinal, 476 Antimonglanz, 542 Antimonite, 542 Antimonsilberblende, 537 Antimony, tests for, 542 Aosta, moraines in valley of, 634 Apatite, 142 Aphanitic matter, 328 Aphelion, 702 Aplite, 316; of Schemnitz, 447 Appalachians, folds in, 504, 668 ; section across, 480 Apsides, Revolution of, 704 Aragonite, 137 ; in shells, 230 Aralo-Caspian area, lakes of, 240 Arctic regions, past genial climate of, 698, 713 Argentite, 537 Arksutite, 161 Arsenic, tests for, 542 Arsenical iron pyrites, 151 Arsenikkies, 151 Arsennickelglanz, 539 Arsenopyrite, 151 Arsensilberblende, 538 Arthur's Seat, 364 Arun, R., 608 Asar, 633 Asbestus, 107 Asbolite, 541 Asiderites, 665 Asphalt, 381, note Association of contortion and induration, 269 Atacamite, 531 Atlantic ooze, 221 Atoll, 225 Atrio, 350 Atrio de Cavallo, 350 Augite, 108, 109 Augite-andesite, 326 Augite-porphyry, 330 Augite-syenite, 324 Augitic granite of Vosges, 316 Australia, deserts of, 637 Autunite, 547 Auvergne, extinct volcanoes of, 363 ; lacustrine rocks of, 297 ; Scrope on subaerial de- nudation in, 586 Axial planes, 21 Axinite, 130 Axiolites, 307 Axis of earth, possible change in position of, 715 Axmouth, landslip at, 584 Ayrshire, volcanic necks of, 373 Azurite, 531 BABBAGE, his theory of up- heaval, 669 Baku, mud volcanoes of, 382 Ball, Dr., on former greater intensity of denudation, 695 Banatite, 329 Barium, detection of, 143 Barrier reef, 224 Barytes, 144 Barytocalcite, 145 Basalt, 332 Basic crystalline rocks, 312 Basin, 476 Baslow, valley of, 591 Bass, 178 Batt, 178 Baveno twin, 78, 95 Beaches, raised, 686 Bearing grit of Grassington, 554 Beaumont, Elie de, on the contraction of the earth, 674 ; on dolomitization, 409 Bedding, 165; irregular, 168; lenticular, 168 ; regular, 168 Beds, 168 Bell-metal ore, 536 Belonites, 305 Bending of solid rocks, 512 Bergseife, 133 Bermudas, formed of coral powder, 229 Beryl, 126 Bimstein, 322 Bind, 177 Biotite, 116 Bischof, his experiments on formation of dolomite, 245 ; his formation of galena, 558 ; on contraction of rocks in cooling, 652, note; on dolo- mitization, 408 Bismite, 544 Bismuth, tests for, 543 Bismuth glance, 544 Bismuthinite, 544 Bismuth ochre, 544 Bitter spar, 13, 138 Bitterspath, 138 Bitume liquide, 382 Bitumen, 381, note Bixaxal crystals, 71 Blackband ironstone, 138 Black Jack, 534 " Blacklead," 160, 186 Black tin, 536 Blast-furnace slag, spheroidal structure in, 307 720 Index. Blast ore, 566 Bleaching action of decaying organic matter, 248 Bleiberg, plurabiferous sand- stone of, 567 Bleiglanz, 532 Bleilasur, 531 Bleispath, 533 Bleivitriol, 532 Blende equivalent to sulphide, 156, note Blister copper, 529 Blotches in red rocks, 248 Blown sand, 256 Blowpipe reactions, 15 Blue John, 142 Bog-iron ore, 569 Bole, 134 Bombs, volcanic, 359 Boracite, 162 Boricky, his method of distin- guishing the felspars, 99 Bornite, 529 Botryoidal, 85 Boulder clays, 266 : masking of features by, 601 Boulders in lodes, 550, 561 Bourbonne-les-Bains, forma- tion of metallic ores at, 557 Brachy-diagonal, 16 Brachydomes, 65, 70 Brachypinacoids, 65, 70 Brading Brook, 608 Branch coal, 185 Braunbleierz, 534 Brauneisenstein, 149 Braunite, 156 Breccia, 172 ; volcanic, 344, 360 Brecciated lodes, 550 Brick-clay, 175 Brick-earth, 175 Bright white cobalt, 541 Brittany, bedded granites of, 432 Brittleness, 86 Brittle silver ore, 537 Brockram, 250 Bromlite, 145 Bronzite, 111, 112 Brookite, 546 Brown coal, 181, 185 Brown haematite, 149 Brown spar, 138 Buntkupfererz, 529 Burdie House limestone, 243 CAKING coal, 185 Calamine, 535 Calcareous mica-schist, 415 Calcite, 13, 135 ; cleavage of, 33 Calc-spar, 135 Canada, granite veins in, 394 ; iron ores of, 569 Caticrinite, 325 Cank, 173 Cannel coal, 185, 262 Canon of Colorado R., 580 Cape of Good Hope, Darwin on foliated rocks of, 419 Capillary, 85 Carbon, 13 Carbonates of lime, iron, and magnesia, means of distin- guishing them. 139 Carbon dioxide in fluid- cavities, 310 Carboniferous limestone, ba- sins in which it was formed, 688 Carclazyte, 317 Carlingforcl Mountains, meta- morphism by granite of, 395 Carlsbad, sprudelstein, 282 Carlsbad, twin, 76, 94 Carnallite, 161 Carrick, metamorphic rocks of, 430 Cassiterite, 536; formation of, 562 Cataclysmal school, 694 Cavities,fluid-,glass-,stone-,310 Cawk, 144 Cayton Bay, fault in, 500 Celestine, 145 ; formation of, 247 Ceneri, 359 Centroclinal dip, 476 Cerargyrite, 538 Cerussite, 533 Cessation of volcanic erup- tions, 346 Chabazite, 129 Chalcedony, 91 Chalcocite^ 529 Chalcolite, 547 Chalcopyrite, 529 Chalk, 178 ; and Atlantic ooze compared, 221 ; basin in which it was formed, 222, 688; its power of resisting denudation, 589; of Isle of Wight, 589 Chalk marl, 178 ; resembles Atlantic ooze, 222 Chalk sea, depth of, 222 Challengerida, 231 Chalybite, 138 Chapeau de Fer, 550 Charcontois, artificial pro- duction of contortion, 675, note Chemically-forined rocks, vol- canic source of materials of, 363 Cherry coal. 185 Chert, 92 ; Hardman and Hull on, 233 ; in limestone, 179 Cherwell, R., 607 Chessylite, 531 Chiastolite, 124 Chiastolite schist, 415 Chili, claystone of, 429 China clay, 133, 174 China-clay rock, 317 Chlorite, 123 Chlorite schist, 415 Chlorsilber, 538 Chrome Hill, 587 Chromeisenstein, 152 Chrome ochre, 545 Chrome spinel, 334 Chromic iron, 152 Chromite, 152 ; in microscopic sections, 155 Chromium, tests for, 545 Chromocker, 545 Chrysocolla, 531 Chrysolite, 113 Chrysotile, 121 Cindery base of lava-flow, 356 Cindery texture, 356 Cinnabar, 544 Clastic rocks, 189 Clay, 174 ; analyses of, 177 Clay ironstone, 138, 569 ; in shale, 178 Clay-slate, 406 Clay with flints, 191 Cleat of coal, 276 Cleavage, crystalline, 15, 33, 72, 86; slaty, 270, 272 Cleveland ironstone, 569 ; Sorby on, 412 Climates, continental and in- sular, 700 Clinkstone, 326 Clinoaxis, 66 Clinochlore, 123 Clinodiagonal, 16 Clinodomes, 68 Clinopinacoids, 68 Closely-grained rocks, 172 Clyde, analysis of water of, 220 Coal, 14, 181; cannel, 262; dicey, 276 ; end and bord of, 276 ; formation of, 257 ; lycopodiaceous spores in, 183 ; subaqueous, 261 Coal seams, deterioration, of, 261 ; partings in, 260 Coarsely-grained rocks, 172 Coast ice, 201 Cobalt, tests for, 538 Cobalt-glance, 541 Cobalt glass, use of to detect potash, 101, note Cobaltite, 541 Cobalt ochre, 541 Cobalt pyrites, 540 Coccoliths, 221 Colloidal states, 83 Colorado, canon of, 580 Colorados, 550 Colour of minerals, 86 Columnar structure, 308 Combinations in crystals, 33 Comby lodes, 549, 561 Compacting of sediment into rock, 267 Compact rocks, 172 Conchoidal fracture, 86 Concretionary action, 2SO Concretions, 279 Conformity, deceptive, 521 Conglomerates. 172, 173 Conjugate systems of faults, 493 Consolidation of the earth, 655 Contact deposits, 555 Contemporaneous erosion, 217 ; and filling up, 516 Continents, growth of, 689; permanence of, 688 ; typical structure of, 685 Contortion, 473 Contour lines, 468 Contraction theory of up- heaval, 672 Co-ordinate axes, 21 Copper, blister, 529; native of Lake Superior, 567 ; tests for, 528 Copper-glance, 529 Copper-nickel, 538 Copper pyrites, 529 Copper uranite, 547 Coral, 223 Coral islands, gypsum and dolomite in, 291 Coral limestone, 228 Corallines, 229 Coralloid, 85 Coral mud, 229 Cornstone, 179 Cornwall, granite of, 393 Corundum, 158 Cotopaxi, 350 ; mud streams of, 360 Crab rock, 256 Index. 721 Crag, indications of cold in the, 712 Crete, 550 Crich Hill, 486 Crocidolite, 152 Crocoisite, 534 Croll on changes in climate, 702, 707 ; on climatic effect of ocean-currents, 701, 708 Crumbly rocks, 172 Crust of the earth, denser be- neath oceansthan continents, 692 Cryolite, 161 Cryophyllite, 116 Crypto-crystalline rocks, 304 ; texture, 82 Crypto-siderites, 665 Crystalline form, 15; do. and chemical composition, 79 ; cleavage, 15, 33, 72 Crystalline rocks, 163; subdi- visions of, 311 Crystalline schists, early theo- ries about, 426 Crystalline state, 82 Crystallites, 305 Crystallography, works on, 162 Crystals, biaxal, 71 ; circum- stances under which they were formed, 339 ; classifica- tion of, 30; combinations in, 16, 33, 73, definition of, 15 ; doubly refracting, 71 ; for- mation of, in lava, 357 ; ideal, 19; model, 19; physi- cal constitution of, 70 ; pseu- do-hexagonal, 73 ; simple forms of, 16, 33; symbols for faces of, 21, 23 ; symmetry of, 25 ; systems of, 30 ; twin, 75 ; uniaxal, 71 Cube, four-faced, 42 Cuivre Panache, 529 Cumberland, haematite de- posits of, 565 Cupriferous iron pyrites, 529 Cuprite, 530 Current-bedding, 215 Cyanite, 124 DACHSCHIEFKR, 406 Dacite, 327 Dales of Derbyshire, 196 Damourite, 116 Dana on coral reefs, 227, note ; on mountain-chains, 629 Danaite, 151 Danemora, magnetite of, 570 Dark -red silver ore, 537 Dartmoor, granite of, 392 Darwin, C., on claystoneofChili, 429 ; on coral reefs, 226 ; on foliated rocks of Cape of Good Hope, 419; of South America, 422 ; of Terra del Fr.ego, 418 Darwin, G., on change in posi- tion of the poles, 716; on former greater intensity of denudation, 695; on origin of" earth, 649 Daubree on access of water to volcanic foci, 345; artificial production of contortion, 675, note ; folds in rocks, 504, 505 ; formation of amygda- loids, 311 ; of cassiterite, 562 ; of joints, 278 ; grinding action of running water, 194 ; metallic ores formed at Bour- bonne and Plombieres, 557 ; metamorphism, 438; meteor- ites, 664 ; rounded grains in blown sand, 256 ; sulphur in Paris rubbish, 251 ; structure en eventail, 425 Dawson on discolouration of red mud, 240 Day-stones, 591 Dead Sea, 239 Decaying organic matter, its bleaching action, 248 Deceptive conformity, 521 De la Beche on contortion, 511 Delaunay on the interior of the earth, 659 Deltas, 292, 636 ; sinking dur- ing deposition of, 670 Deltohedron, 18, 47; twelve- faced, 45 Dendrites, 14, 157 Dendritic, 85 Density, increase of, produced by given increase in pres- sure, 643, 647 Denudation, 189; formergreater intensity of, 695 Denuding agents, 189 Derivative rocks, 189; classifica- tion of, 284 Derwent river, gorge of, near Matlock, 579 ; of Yorkshire, 611 Derwentwater, Floating Island of, 262 Descartes on contraction of the earth, 674 Des Cloiseaux, his method of distinguishing the felspars, 99 Descriptive geology, summary of, 169 Deserts, 637 Desmine, 129 Devitrification, 84 Devon, granite of, 393 Diabase, 329, 330 Diabase porphyrite, 329, 330 Diakis dodecahedron, 49 Diallage, 107, 108 Diallage-peridotite, 334 Dialogite, 157 Dialysis, 83 Dialytic rocks, 189 Diamond, 160 Dichroism, 86 Dihexagonal pyramid, 57 Dimetric prisms, 63 Dimetric system, 31, 60 Dimorphism, 79 Dinas brick, 177 Diopside, 107 Dioptase, 531 Diorite, 328 Diorite porphyrite, 328, 329 Dip, 460 ; its effect on shape of surface, 591 Diploid, 49, 62 Dip-slope, 597 Dipyr-schist, 416 Directions of extinction, 99 ; in the felspars, 100 Dirt-bed of Isle of Portland, 252 Disthene, 124 Ditetragonal pyramid, 61 Ditroite, 325 Dodecahedrons, 18 ; deltoidal 45; diakis, 49; pentagonal, 43 ; rhombic, 39, 41 2 z Dogtooth spar, 136 Dolerite, 332 Dolomite, 138, 179, 180, 408; formation of, 244; in coral islands, 291 ; of Co. Cork, 409 Dolomitic conglomerate, 255 Dolomitization, 408 Dome, 476 Domes, 65 Domite, 326 Donegal, metamorphic rocks of, 401 ; plutonic rocks of, 392 Dora Baltea, moraines of, 634 Dorsetshire coast, landslips on, 584 Double refraction, 71 Draughton, contorted beds near, 475 Drift-bedding, 215 Druses, 549 Duckstein, 361 Dunbar, sandstone blocks in lava at, 371 Dunes, 635 Dunite, 335 Dunklesrotligiiltigerz, 537 Durham, concretionary mag- nesian limestone of, 281 Durocher on crystalline rocks, 313 ; on dolomitization, 408 Dykes, 344, 351, 370, 372 ; ab- ruptly terminated above, 521, note EARTH, chemical composition of crust of, 13; consolidation of, 655 ; crust of, 7 ; density of, 642 ; greater beneath oceans than continents, 692 ; internal temperature of, 643; ratio of superficial to mean density of, 648 ; shape of, 641; not due to denudation, 648 ; variations of pressure, temperature, and, fusing- point within, 654 Earthquakes, 383 ; accompany volcanic eruptions, 344 Earthy cobalt, 541 Earthy fracture, 86 Eccentricity, variations in, 709 "Eddy rock," 216 Eden valley, Till of, 265 Eifel, extinct volcanoes of, 363 Eigg, Scur of, 573; tachylite of, 334 Eisener hut, 550 Eisenglanz, 147 Eisenkies, 149 Eisennickelkies, 539 Eisenspath, 138 Eisensteinmark, 134 Elasticity, 86 Elastico-viscosity of rocks 512, 661 Elevation, meaning of in geology, 667 Eloeolite, 102 Eloeolite-syenite, 325 Elton, Lake, 241 Elvanite, 317 Emerald, 126 ; Oriental, 158 Emerald copper, 531 Emerald nickel, 539 Emery, 158 ; separation from magnetite, 158 Encrinites, 229 Engheim, sulphurous waters of, 251, note 722 Index. Ennerdale, eskers at mouth of, 633 Enstatite, 111, 113 Enstatite-peridotite, 334 Eocene glacial deposits, 712 Epidote, 127 Equilibrium, surface of, 645 Erdkobalt, 541 Erdol, 382, note Erratics, 266; in oceanic de- posits, 290 Erubescite, 529 Eruptive volcanic rocks, 370 Erythrite, 541 Escarpment, 597 Eskers, 632 Estuarine rocks, 292; fossils of, 293 Ethane in cannel coal, 186 Etna, 351 Eurite, 310 Eurit-porphyr, 318 Evolution a measure of geo- logical time, 693 Extraordinary ray, 71 FACE of coal, 276 Fahlbands, 554 Fahlerz, 529 Falkland Island, peat of, 257 False betiding, 216 False veins of sandstone in lava, 374 Fanlike structure, 424 Faults, 488 ; change in size of, 493 ; formation of, 507-509 ; overlap, 504 ; reversed, 492. 504, 509 ; valleys determined by, 601 Favre, Professor A., his arti- ficial production of contor- tion, 674 Fayalite, 114 Feel of minerals, 88 Felsite, 320 Felsitic matter, 309 Felsit-pitchstone, 322 Felsit-porphyr, 318 Felspar group, 94 Felspars, chemical composition of, 96; distinctions between, 98; twinning in, 94 Felspathic mica-schist, 416 Felspathic mud, 133 Felstone, 319 Fer chrome, 152 Ferguson on oscillation of rivers, 606 Fer oligiste, 147 Fer oxydule, 146 Ferric oxide, 146 Ferroso-ferric oxide, 146 Ferrous oxide, 146 Feuerstein, 91 Fibrolite, 125 Filiform, 85 Findlinge, 266, note Finely-grained rocks, 172 Fire-clay, 175 Firestone of Ventnor, con- tains glauconite grains, 233 ; resembles Atlantic ooze, 222 Fisher, Rev O., his objections to Mallet's views, 682; his theory of volcanic action, 676, 678 ; on access of water to volcanic foci, 345 ; on cause of contortion, 671, note ; on pressure produced by earth's contraction, 673, note, 674 Fissile rocks, 168 Fissile structure, 308 Fissure eruptions, 352 Flagstone, 168 Flats, 555 Flattening of pebbles in con- glomerates, 271 Fleckschiefer, 418 Flemingites, 183 Flexibility of minerals, 86 Flexures of strata, their effect on shape of surface, 597 Flint, 91 ; in limestone, 179, 231 ; Professor Sollas on for- mation of, 232 Florida, coral reefs of, 228 Flour of rock, 200, 265 Fluidal structure, 305, 306 Fluid-cavities, 89, 310 Fluidity of lava, 353 Fluoride, 42 Fluor-spar, 14, 142 ; cleavage of, 35 Flusspath, 142 Folding of rocks, by horizontal pressure, 504 ; by vertical npthrust, 503 Foliated rocks, 163, 403; crumpled laminae of, 423 Foliation, 414; artificial pro- duction of, 422 ; cause of, 418 ; parallel to bedding and cleavage, 421 Fontainebleau, sandstone of, 277 Foot-wall, 550 Foraminifera, 220 Forbes, D., his artificial pro- duction of foliation, 422 ; on contraction of rocks in cool- ing, 653, note "Form" in crystallography, 32 Formazione zolfifera, 250 Fossils, 167; as indications of change in climate, 710 Fox on galvanic currents in lodes, 564 Foyaite, 325 Fracture of minerals, 86 Fragments included in lava, 371 Franklinite, 153 Freieslebenite, 537 French chalk, 120 Freshwater Bay, cleavage of chalk in, 274 ; horizontal dis- placement by fault in, 498 Friability, 86 Friable rocks, 172 Fringing reef, 223 Frost, denudfng work of, 196 Fruchtschiefer, 418 Fuchs on foliated rocks of Pyrenees, 419 Fuller's earth, 134 Fuschite, 116 Fusing-point, effect of pressure on, 651 Fusiyama, 350 GABBRO, 331, 333 Galena, 532 Galenoid, 45 Galliard, 173 Galmei, 535 Galvanic currents in lodes, 564 Ganges, delta of, 292 Ganggestein, 548 Gangue, 548 Garbenschiefer, 418 Garnets, 126 Garnet-olivine rock, 334 Garnierite, 540 Geanticlinal, 630 Geikie, A., on connection be- tween contortion and meta- morphism, 685 ; on sun- cracked sandstone in Caith- ness, 218 Geikie, J., on granites of south of Scotland, 395 Gelatinization, 92 Gelatinous state, 83 Geneva, Lake of, 624 Geological time, 693 Geology, descriptive and his- torical, 5, 8 ; a history, 286 Geosynclinal, 629 Gersdoffite, 539 Giant's Causeway, columnar structure of, 277 Giant granite, 316 Gibraltar current, 237 Gilbert on laccolites of Henry Mountains, 671 Gilbertite, 116 Glacial epochs, 711 Glacial mud, 265 Glaciers, 196 Glanzkobaltkies, 541 Glass-cavities, 310 Glassy rocks (acid), 321 Glassy state, 84 Glauconite, 134 Glauconite grains, 233 Glen Eisdale, vesicles in intru- sive lava of, 375, note Glimmerschiefer, 415 Globigerina ooze, 221 Globular, 85 Globulites, 305 Gneiss, 416 Gold, tests for, 544 Gossan, 550 Gothite, 149 Goyt Trough, 479 Grammatite, 107 Grange Irish, metamorphism by granite of, 395 Granite, 315 ; Banffshire, 447 ; Brittany, 432, 448 ; Dundalk, 447 ; lithological examination of, 9; order of solidification of minerals of, 397; Priestlaw, 430 ; south-west of Scotland, 431 ; surface weathering of, 204 ; Western Territories, 433 Granitell, 316 Granite veins, 393 Granitic gneiss, 417 Granitoid rocks, 428 Granit-porphyr, 317 Granulite, 417 Graphic granite, 316 Graphite, 160, 186; origin of, 413 Graphitic gneiss, 416 Grassington, lodes of, 554 Grauwacke, 405 Great flat lode of Redruth, 555 Great ice age, the, 698, 711 Green earth, 134 Greenland, ice-sheet of, 200 Greenovite, 131 Green river, 609 Green sands, 233 Index. 723 Greisen, 316; tin-bearing of the Zinnwald, 562 Grey antimony, 542 Grey cobalt ore, 540 Grey copper ore, 529 Grey ooze of Atlantic, 233 Grit, 173 Gritstone, formation of out of granite, 187 Ground ice, 201 Griinbleierz, 533 Grundmorane, 201, 264 Griinererde, 134 Gulf Stream, 700 Gypsum, 14, 140, 186 ; Cana- dian, Sterry Hunt on, 412 ; formation of, 247, 248, 412; in coral islands, 291 ; joints in, 277 ; metamorphic, 412 HAARKIES, 539 Hackly fracture, 86 Hade, 492, 550 Haematite, 147 ; deposits of in Cumberland, 565 Halb-granit, 316 Half-opal, 91 Halite, 161 Hall, Sir J., his forn ation of marble, 407 ; on contortion, 512 Halleflinta, 429 Hanging wall, 550 Hannay on gases above critical point, 345 Haplite, 316 Hard coal, 185 Hardening of sediment by compression, 268 Hardman and Hull on chert, 233 Hardness, scale of, 87 Harkness on dolomites of Co. Cork, 409 Haughton, Prof., analyses of felstone, 318 ; on metamor- phism in Carlingford Moun- tains, 395 Hausmannite, 157 Hauyne, 104 Hauyninite, 333 Hauynophyr, 333' Heave, 497 Heavy spar, 14, 73, 144 ; for- mation of, 247 Heddle, Prof., on chlorite, 123 Heim, " Mechanismus der Ge- birgsbildung," 668; on con- torted strata, 507 Heinrich on internal tempera- ture, 644 Hemihedral, 33 Hemioctahedron, 67 Hemiprisms, 70 Henessey on the interior of the earth, 659, 661 Henry Mountains, 671 ; dykes of, 521, note Hepatic cinnabar, 544 Heron Island, coral limestone of, 228 Herschel, Sir J., his theory of upheaval, 670; on changes in climate, 702, 706 Hexagonal prisms, 59 Hexagonal pyramid, 53 ; of second order, 54 Hexagonal system, 30, 50 Hexakis octahedron, 48 Hexakis tetrahedron, 49 Hibbert on reversal of the Rhine, 612 Hills and valleys formed by denudation, 571 Himalayas breached by Indus, 612 Hohlspath, 1-24 Holohedral, 32 Holosiderites, 664 Holy well, U>5 Holzzinn, 536 Hopkins, W., determination of thickness of earth's crust from precession, 656 ; his ex- periments on the raising of the melting-point by pres- sure, 651 ; his theory of up- heaval, 507, 668 ; his theory of volcanic action, 676 Hornblende, 107, 108 Hornblende-andesite, 326 Hornblende rock, 416 Hornblendic gneiss, 416 Hornblenclic granite, 316 Horn silver, 538 Homstein, 92 Hornstein-porphyr, 318 Hornstone, 310 Horse-flesh ore, 529 Horses, 550 Household coal, 181, 185 Hugh Miller on Sutherland- shire mountains, 573 Humboldt Mountains, meta- morphic rocks of, 402 Humus, 181 Hunt, R., on galvanic currents in lodes, 564 Hunt, Sterry, on Canadian gypsum, 412 ; on volcanic action, 678 Hutton,3; on denudation, 207; on formation of surface by denudation, 573; on subae'rial denudation, 585 ; on uncon- formity, 517, note Hyacinth, 133 Hyalite, 91 Hyalomelane, 334 Hydraulic limestone, 178 Hydrohamatite, 149 Hydrothermal action, 341 Hydrozincite, 535 Hypersthene, 111, 112 Hypersthenite, 333 ICE, conditions necessary for accumulation of, 711 ; denud- ing work of, 196 Icebergs, 200 Ice-foot, 201 Ice-formed deposits, 262 Iceland spar, 137 Ice-polishing, 614 Ice-scratches, 263, 614 Ice-sheets, 200 Ice-worn districts, 614 Idocrase, 358 Ikositetrahedron, 47 Ilfracombe, cleaved limestone near, 271 Ilmenite, 14, 148, in micro- scopic sections, 155 Ilvaite, 151 Imitative shapes of minerals, 85 Incretionary nodules, 282 Indianite, 97 Indices, 23 ; law of rationality of, 25 Indus, breaching of Himalayas by, 612 Inlier, 485 Inostranzeff on crystalline limestone, 408 Intercepts, 21 Intrusive lava-sheets, circum- stances under which they were formed, 377 Inversion, 482 Ireland, landslips in north- east of, 584 Irish Sea, analysis of water of, 2*0 Iron as a colouring matter, 154; compounds of, 146; occurrence in rocks, 154; oxides of, 146; tests for, 146 Iron hat, 550 Iron ores, bedded, 569 ; forma- tion of, 560 Iron pyrites, 149; arsenical, 151 ; cupriferous, 529 ; in microscopical sections, 155 ; in shale, 178 ; of Andalucia, 564 Irtish river, and defiles of Altai, 574 Iserine, 148 Isle of Wight, drainage of, 608; landslips at Undercliff in, 584 ; marine and subae'rial denudation in, 589 ; Needles of, 591 Isomerism, 85 Isomorphism, 80 Isothermal lines, 699 Isotropic, 72 Isthmus of Suez, Bitter Lakes of, 240 JACOBSITE, 147 Jade, 107 Jamieson on river-terraces, 636 Jargon, 133 Jarvis Island, gypsum and dolomite of, 291 Jasper, 91 Jheels, 259 Joints, 274; distinction be- tween and crystalline cleav- age, 276 ; their effect on for- mation of valleys, 601; on shape of surface, 591 Jordan, elementary crystallo- graphy, 38 Jordan valley, 622 Jura, inversion in the, 482 KALKSPATH, 135 Kalkuranglimmer, 547 Kames, 632 Kaolin, 133, 174 ; formation of, 192 Kara-bogas, Bay of, 241 Karakorum Mountains, 613 Kentish Town, well at, 527 Kersantite, 327 Kersanton, 327 Kidney ore, 147 Kies, equivalent to sulphide 149, note Kieselkupfer, 531 Kieselzinkerz, 535 Kilauea, glassy basic rocks of, 334 King, Clarence, on volcanic action, 677 724 Index. Knistersalz, 186 Knotenschiefer, 418 Kobaltbliithe, 541 Kobaltglauz, 541 Kobaltkies, 540 Kobaltnickelkies, 540 Kongsberg, Fahlbands of, 554 Korniger kalkstein, 406 Kossen beds, 302 Krolith, 161 Kupferglanz, 529 Kupferkies, 529 Kupferlasur, 531 Kupfernickel, 539 Kupferschiefer, 566 Kupferschwarze, 530 Kupfertiranite, 547 Kuproicl, 47 Kwemlmn Mountains, 690 LABRADOR Current, 700 Labradorite, 97 Laccolites, 671 Lacustrine rocks, 297 Lakes, 617 ; enclosed by eskers 632, 635; ice- worn, 622; in rock -basins, 619 ; marine animals in fresh-water, 298, 622; of Central Asia, 240, 622; of North America, 618 Lake Superior, native copper of, 567 Lamellar cleavage, 86 Lamina, 168 Laminated structure in crystal- line rocks, 307 Landslips, 583 Lapilli, 359 Laplace's law of internal den- sity, 646 Laterite of Madeira, 254 Lava, 344; fluidity of, 353; near Keswick, 340 ; relative age of acid and basic, 379 ; sheets of, 370, 374; water in, 345, 354, 677, 679 Lave di fango, 361 Lave di fuooo, 361 Lead, tests for, 531 Le Conte, Prof. J., on moun- tain building, 684 Lepidodendron, 184 Lepidolite, 116 Lepidomelane, 116 Lepidostrobus, 183 Leucite, 102 Leucite-basalt, 333 Leucitite, 333 Leucitophyr, 333 Leucopyrite, 151 Leucoxene, 155 Level Mne, 460 Levy arid Fouque on directions of extinction, 99 Lherzolite, 334 Lichtesrothgultigerz, 538 Lievrite, 151 Light-red silver ore, 538 Lignite, 181, 185 Limburgite, 335 Limestone, 172, 178 ; crystal- line 406 ; dissolution of by carbonated water, 190 ; pin- nacles of, 591 ; place on sea- bed of, 230; saccharoidal, 407 ; tests for, 180 Lime uranite*547 Limonite, 14, 149 Linarite, 531 Linnaeite, 540 Liparite, 320 Lithia mica, 116 Lithionglimmer, 116 Lithionite, 116 Lithium, tests for, 145 Lithology, 8 Lithomarge, 134 Littoral rocks, 288 Liver ore, 544 Llanberis, rocks of, 434 Loam, 177 Loch Lomond, 624 Lockyer on stages in a star's cooling, 693 Lodes, 548; back of, 550; boulders in, 550, 561 ; brecci- ated, 550 ; comby, 549, 561 ; dimensions of, 552; direction of, 552 ; Great Flat of Red- ruth, 555; heaving of, 551; outgoing of, 550 ; of Corn- wall, 554; of Derbyshire, 554 ; of Grassington, 554 ; of north of England, 554; relation between their con- tents and adjoining rocks, 553 Lollingite, 151 Longitudinal valleys, 596 Longnor, physical features of country near, 587 Lough Maam, 634 Lough Slievesnaght, 620 Lower Greensand, 295 Lultsattel, 477 Lustre of minerals, 87 Luxilianite, 317 Lydian stone, 405 Lyell, Sir C., his explanation of changes in climate, 701 MACALUBA, 381, 382 Macro-crystalline rocks, 304 Macro-diagonal, 16 Macrodomes, 65, 70 Macropinacoids, 65, 70 Madeira, laterite of, 254 Magma-basalt, 335 Magnesian carbonate in organic structures, 411 Magnesian limestone, 179, 180 ; of England, 299; Sorby on, 411 Magnesioferrite, 147 Magnesite, 138 Magneteisenerz, 146 Magnetic pyrites, 150 Magnetite, 14, 146; in micro- scopic sections, 155 Magnetkies, 150 Main river, dissolved matter in, 194 Malachite, 530 Malacolite, 107 Malleability, 86 Mallet on columnar structure, 308, note ; on pressure pro- duced by earth's contraction, 673, note ; on volcanic action, 679 Maltha, 382, note Mammillary, 85 Manganblende, 156 Manganese, compounds of, 155; tests for, 140, 146, 155 Manganglanz, 156 Manganite, 157 Manganspath, 157 Marble, 179; formed artificially by Sir J. Hall, 407 Marcasite, 150 Margarodite, 116, 120 Marine animals in fresh-water lakes, 298, 622 Marine denudation, 206 ; plain of, 577 ; surface formed by, 576 Marl, marl slate, 178 Martite, 148 Massive minerals, 82 Master joints, 275 Matea, magnesian limestone of, 291 Matlock, gorge of Derwent near, 579 ; petrifying springs of, 242 Mechanical deposits, arrange- ment on sea-bed, 210 Medina river, 608 Mediterranean, saltness of eastern part of, 237 Medlieott on inversion in the Himalayas, 675, note Medway river, 608 Meerschaum, 120 Melaconite, 530 Melanchroite, 534 Melanglanz, 537 Melanite, 127 Melaphyr, 330 Menaccanite, 148 Mennige, 534 Meres in Cheshire, 196, 622 Mesitite, 138 Metals occurring principally as sulphides, 557 Metamorphic rocks, 399; ir- ruptive behaviour of, 447 ; of Donegal, 401; of the Sierra Nevada, 402 Metamorphism, causes of, 435 ; not a function of depth only, 684 ; selective, 429 Metamorphisme de juxtaposi- tion, 436 ; regional, 436 Meteorites, 664 Miall, his experiments on con- tortion, 511 Miascite, 325 Mica-basalt, 332 Micaceous cleavage, 86 ; sand- stones, 213; shales, 213 Miea-diorite, 328 Mica group, 114, 117 Mica-quartz-diorite, 329 Mica-schist, 415; ripple-drift in, 400 Mica-syenite, 327 Mica-traps, 327 Microcline, 97, 101 Micro-crystalline rocks, 304 Microliths, 305 Milford Haven, inversion near, 482 Milk-white quartz, 89 Miller, his crystallographic no- tation, 36, 53 Millerite, 539 Mimetite, 534 Mineral charcoal, 184 Mineralogy, works on, 162 Minerals, accessory, 12 ; defi- nition of, 10; feel of, 88; fracture of, 86 ; means of recognising, 14 ; physical characters of, 10 ; rock-form- ing, 11; smell of, 88; specific gravity of, 88 ; streak of, 86 ; taste, 88 ; tenacity cf, 86 Minette, 327 Index. 7 2 5 Minium, 534 Miocene glacial deposits, 712 Mispickel, 151 Mississippi, delta of, 292; sediment carried by, 193 Model illustrating uncon- formity, 519 Molasse, 299 Mole river, 608 Molecules, 72 Molybdanglanz, 545 Molybdanocker, 545 Molybdenite, 545 ; distin- guished from graphite, 1(50 Molybdenum, tests for, 545 Molybdic ochre, 545 Molybdite, 545 Monoclinic system, 31, 66 Monometric system, 30, 36 Moutera de hierro, 565 Monte Somma, 350 Monzonite, 324 Moraine profonde, 201, 264 Moraines, 198, 265, 617, 633 Morenosite, 540 Mother of coal, 184 Mountain-chains, contortions in, 628; definition of, 626; edge continents, 685 ; thick- ness of strata in, 628 ; struc- ture of, 667 Mountain leather, 107 Moya, 361 Mudstone, 175 Mud streams, volcanic, 360 Mud volcanoes, 381 Mulberry Mine, 555 Mulleer river, great flood in, 193 Mundic, 149 Murray, Mr. John, on forma- tion of coral reefs, 227 ; on red clay of Atlantic, 235 Muschelkalk, 302 Muscovite, 115 NADELEISENERZ, 149 Nailhead spar, 136 Naphtha, 381, note Natrolite, 129 Naumann, his crystallographic notation, 36, 53, 66, 69 Nebengestein, 550 Nebular hypothesis, 640 Necks, of volcanic agglomer- ate, 372; volcanic, 370, 373 Needles, the, 591 Negrillos, 550 Nen river, 607 Nepheline, 101 Nepheline-basalt, 333 Nephelinite, 333 Nephrite, 107 Nevadite, 320 Neve, 198 New Zealand, extinct volcanoes of, 363 Niagara, 604 Niccolite, 539 Nickel, tests for, 538 Nickelarsenkies, 539 Nickel-glance, 539 Nickelkies, 539 Nickelsmaragd, 539 Nickelvitriol, 540 Nile, dissolved matter in, 194 Nodular minerals, 85 Nodular schist, 417 Nodules, secretionary or in- cretionary, 282 Nodules, siliceous in limestone, 231 Non-crystalline rocks, 164 ; subdivisions of, 172 Nordenskiold on iron of Ovi- fak, 666 Norite, 333 Normanton, section near, 213 Northamptonshire ironstone, 569 North Berwick, ejected blocks in volcanic ash near, 359 ; siliceous deposits near, 243 North Berwick Law, 373 North Wales, old volcanoes of, 368 Nosean, 104, 105 Nummulites, 223 OBLIQUE sections, 466 Obliquity of the ecliptic, 715 Obsidian, 322 Occlusion of steam in lava, 345 Ocean-currents, their effect on climate, 701, 708 Oceanic chemical deposits, 291; rocks, 290 Oceans, greater density of crust beneath, 692; permanence of, 687 Octahedrite, 546 Octahedrons, 17; hemihedral form of, 37 ; hexakis, 48 ; monoclinic, 67 ; regular, 37 ; rhombic, 63; triakis, 45; triclinic, 69 ; twinned, 76 ; unequally-developed, 20 Oil shale, 178 Old glacial deposits, 267, 711, 715 Old Red Sandstone on borders of Lake country, 525 Oligoclase, 97 Olivine, 113; altered into ser- pentine, 122 Olivine-augite-peridotite, 335 Olivine-diabase, 330 Olivine -diallage-enstatite-peri- dotite, 334 Olivine-diallage rock, 334 Olivine-gabbro, 331 Oolitic structure, 281 Opal, 84, 91 Ophicalcite, 408 Optic axes, 71 Ordinary ray, 71 Organic denuding agents, 203 Orpiment, 543 Orthoaxis, 66 Orthoclase, 96 : Baveno twin of, 78 ; Carlsbad twin of, 76 ; crystalline form of, 74 Ortho-diagonal, 16 Orthodomes, 68 Orthopinacoids, 68 Ottrelite schist, 416 Ouse river, 608 Ouse, Great river, 607 Outcrop, 468 ; breadth of 472 ; effect of faults on, 496 Outgoing of lodes, 550 Outliers, 485 ; basin-shaped lie of beds in, 585 Overlap, 523 Overlap fault, 504 Ovifak, native iron of, 666 PACOS, 550 Palestine, flinty soil of, 192 Papandayang, 350 Paper shales, 168 Parameters, 23 Parasite, 107 Park Hill, 587 Paroxysmal school, 694 Partings in coal seams, 260 Passage of the earth through cold regions of space, 717 Patagonia, vegetable accumu- lation on, 260 Paulite, 112 Pearl spar, 138 Peastones, 282 Peat, 181, 257 Pebbles in lodes, 550, 561 Pechstein-porphyr, 322 Pectous state, 83 Pegmatite, 316 Penarth beds, 301 Penninite, 123 Pentagonal dodecahedron, 43 Pentlandite, 539 Peperino, 361 Perched blocks, 266 Peridote, 113, 334 Perihelion, 702 Perikline type of twinning, 95 Perlite, 322 Perlitic structure, 305 Permanence of oceans and con- tinents, 687 Permian glacial deposits, 715; rocksofnorth-eastofEngland, 299 ; near "Wliitehaven, 410 Petrography, 8 Petroleum, 382, note Petrology, 9 Petrosilex, 310 Phlogopite, 116 Phoenicocroite, 534 Phonolite, 326 Phosphate of lime in dolomite, 412 Phosphorite, 143 Phthanite, Prof. Renardon, 233 Phyllite, 406, 415 Piano, 350 Picotite, 334 Pictou, grey mud in harbour of, 249 Piedmontite, 128 Pikrite, 335 Pikrit-porphyr, 335 Pinacoids, 65, 68 Pipe-clay, 175 Pipe veins, 555 Pisolitic structure, 282 Pistazite, 127 Pistomesitite, 138 Pitchblende, 547 Pitchstone, 321 Placers, 570 i Plain of marine denudation, 577 Plate, 177 Platinum, tests for, 544 Playfair, 7; on formation of sur- face by denudation, 573 Plombieres, formation of metal- lic ores at, 557 Plumbago, 160, 186 Plutonic rocks, 390; litho- logical varieties of, 396 Pole, change in position of, 716 Polianite, 156 Pollen, showers of, 184 Polycistinse, 231 Polymorphism, 79 Poly synthetic twinning, 95 Pontefract, section near, 213 7 26 Index. Porcellanite, 406 Porodinous rocks, 165 Porphyrite, 328 Porphyritic pitchstone, 322 Porphyritic rocks, 304 Porphyroids, 434 Portland, dirt-bed of Isle of. 252 Potato-stones, 410 Pot- clay, 175 Potomac, breaching of Alle- ghanies by, 574 Poudingue a pate euritique. 435 Prairies, 637 Pratt, Archdeacon, on greater density of earth beneath oceans than continents, 692 Precession of the equinoxes, 656 Precipitation, conditions neces- sary for, 237; how brought about, 236 Pressure, effect of on metamor- phism, 438; on the fusing- point, 651 Prevost, Constant, on the con- traction of the earth, 674 Priestlaw, granite of, 430 Prisms, 16 ; dimetric, 63 ; hex- agonal, 59; monoclinic, 67; rhombic, 64; triclinic, 70; trimetric, 64 Propylite, 380 Protogine, 416 Proustite, 538 Przibramite, 535 Pseudo-fluxion structure, 306 Pseudo-hexagonal crystals, 73 Pseudomorphism, 81 Pseudomorphs of salt in red rocks, 249 Psilornelane, 14, 16 Puddingstone, 172 Puddling ore, 566 Pumice, 322 Punfleld beds, 295 Purple copper ore, 529 Puys, 349, 351 Puzzolana, 359 Pyramids, 17 ; dihexagonal, 57; ditetragonal, 61 ; hexa- gonal, 53 ; do. of second order, 54 Pyrargyrite, 537 Pyrenees, foliated rocks of, 419; included blocks in granite of, 395 Pyrites, 14, 149; capillary, 539 ; copper, 529 ; iron, 149 ; magnetic, 150; microscopic section, 155 Pyroclastic rocks, 360 Pyrolusite, 14, 156 Pyromorphite, 533 Pyrrhotine, 150 QUAQUAVERSAL dip, 476 Quarter-point veins, 552 Quartz, crystalline form of, 18, 59 ; description of, 88 ; fused by oxyhydrogen blowpipe, 397 Quartz-andesite, 327 Quartz crystals on grains in grit, 405 Quartz-diorite, 328 Quartz-felsite, 317 Quartz-freier pbrphyr, 325 Quartzite, 404 Quartz-porphyr, 318 Quartz rock, 404 Quartz-schist, 416 Quartz-trachyte, 320 Quecksilberlebererz, 544 RABDOLITHS, 221 Radiolaria, 231 Rain, denuding action of, 190 Raindrops, 217 Rain-wash, 204 Raised beaches, 636 ; of Great Britain, 458 j Ramsay on amount of denuda- tion, 574; excavation of rock- basins by ice, 623 ; foliated rocks of Anglesea, 421, 423 ; relation of pressure and metamorphism, 685, note ; reversal of river Rhine, 612 Rautenspath, 138 Realgar, 543 Rearranged glacial beds, 266 Red clay of Atlantic, 233 Red cobalt, 541 Red copper ore, 530 Reddle, 147 Red-lead, 534 Red ochre, 147 Red rocks, absence of fossils Jrom, 248; associated with dolomite, 244, 248 Redrnthite, 529 Red soils of Antrim and West- ern Islands, 254 Red zinc, 535 Reich on galvanic currents in lodes, 564 Reniform, 85 Reversed faults, 492; produc- tion of, 504 Rhine, reversal of flow of, 612 Rhode Island, ' ' iron ore " of, 666 Rhodocroisite, 157 Rhodonite, 158 Rhombic octahedron, 63 Rhombohedron, 54 Rhone, dissolved matter in, 194; sediment carried by, 193 Rhyolite, 321 Ribboned structure, 306 Richthofen, his succession of volcanic rocks, 379 Riders, 550 Rigidity of the earth, 657 Rio Tinto, 5(54 Ripidolite, 123 Ripple-drift, 216; in mica- schist, 400 Ripple-mark, 217 Rivers, alluvial plains of, 605 ; carrying power of, 193; cut- ting back of channels of, 603 ; denuding action of, 192, 578 ; trenches cut by, 579; ways of, 607 River-terraces, 635 River-valleys, alternations of gorge and flat in, 579 ; partly transverse, partly longitu- dinal, 602 Roches Moutonnees, 616 . Rocks, acid crystalline, 312, 323; amorphous metamor- phic, 427; amygdaloidal, 311; arenaceous, 172, 173 ; argil- laceous, 172, 174 ; basic crys- talline, 312; calcareous, 172; carbonaceous, 172 ; clayey, 172, 174 ; classification of derivative, 284 ; clastic, 189 ; crypto-crystalline, 304 ; crys- talline 163, 169; subdivi- sions of, 311 ; how formed, 340 ; definition of, 11 ; deriva- tive, 189; determination of origin of, 171 ; dialytic, 189 ; estuarine, 292; iissile, 168; foliated, 163,403; fossilit'erous, 167; lacustrine, 297; litlio- logical classification of, 163 ; littoral, 288; macro-crystal- line, 304 ; metamorphic, 399; micro-crystalline, 304 ; non- crystalline, 164, 169; oceanic, 290, 688 ; petrological classi- fication of, 168; plutonic, 390, 396; porodinous, 165; porphyritic,304; pyro-clastic, 360 ; sandy, 172, 173 ; schis- tose, 163, 169, 403 ; stratified, 167, 169; texture, of, 172; Thalassic, 289 ; Triassic, 301 ; vegetable origin of, 257 ; vesicular, 311 ; volcanic, de- finition of, 342; eruptive, 370; intrusive and contem- poraneous, 369 ; irruptive, 370 ; lithological varieties of, 378; resemblance to lava, 343, 378 Rock-basins, 619 Rock bind, 178 Rock-salt, 161, 186 ; formation of, 238 Rock-sand, 173 Rock-soap, 134 Ropy structure, 342, 356 Rothbleierz, 534 Rotheisenerz, 147 Rothkupfererz, 530 Rothnickelkies, 539 Rothzinkerz, 585 Rottenstone, 179, 192 Roundness of grains in blo\\n sand, 256 Ruby, 158 Ruby silver, 538 Runn of Cutch , 240 Rutile, 14, 546 Ruttles, 492 SADDLE, 476 Sahara, 637 Sahlite, 107 Salband, 550 Salses, 381 Salt, common, 14; in sea-foam, 237 ; influence in promoting deposition, 214; rock, 161, 186, 238 Salt Lake of Utah, 239 Sand, 173 Sandberger on lodes of Black Forest, 559 Sand-dunes, 635 Sandstone, 173 ; calcareous, 10 Sanidine, 97 Sanidine-trachyte, 321 Saponites, 123 Sapphire, 158 Satin spar, 140 Saussurite, 331 Scandinavia, sinking of land in, 458 Scapolite, 130 Scale of hardness, 87 Scalenohedrons, 17; hexagonal, 58 ; tetragonal, 62 Schemnitz, aplite of, 447 Index. 727 Schilfglaserz, 537 Schiste made, 415 Schiste maclifere, 415 Schistose rocks, 163, 169, 403 ; intrusive, 424 Sc.hmirgel, 158 Schorl, 129, 130 Schorl-schist, 416 Schwarzgiiltigerz, 537 Schwarzkupfererz, 530 Schwefelkies, 149 Schwerspath, 144 Scoriaceous texture, 356 Scorise, 356, 359 Screes, 254 Scrope, his theory of up- heaval, 669, 671 ; his theory of volcanic action, 676; on subaerial denudation in Auvergne, 586 Scur of Bigg. 573 Seatearth, 258 Seatstone, 258 Sea-water, analysis of, 239 ; result of artificial evapora- tion of, 241 Sea-weeds, their bleaching action on red mud, 249 Secretionary nodules, 282 Sectibility, 86 Selaginella, 183 Selective metamorphism, 429 Selenite, 140 Selgemme, 161 Sepiolite, 120 Septaria, 279 Serapis, Temple of, 458 Sericite, 116 Sericite-schist, 416 Serpentinasbest, 121 Serpentine, 121, 335 ; origin of, 417 Shale, 177; sandy, 178 Shaler on mountain-chains and continents, 692 Sheffield, great flood near, 193 Shiver, 177 Shutlingslow, 486 Siderite, 138 Silberglanz, 537 Silberhornerz, 538 Silica, 13 ; colloidal, 83 Silicates, tests for, 92 Siliceous nodules in limestone, 231 Siliceous sinter, 243 Sillimanite, 125 Silver, tests for, 536 Silver-glance, 537 Simon's Seat, 477 Sinking of sea-bed during de- position, 517 Skaptar Jokul, 352 Slaggy structure, 342, 356 Slaty cleavage, 270; mechanics of production of, 272 Slickenside, 491 Slyne of coal, 276 Smaltine, 540 Smaragdite, 107, 331 Smell of minerals, 88 Smith, William, 4 Smithsonite, 535 Smooth fracture, 86 Snowfleld, 198 Soapstone, 120 Sodalite, 104 Soil, formation of, 203 Solfatara stage. 350 ; products of, 362 Solfatara di Tivoli, 250 Sollas, Prof., on formation of flint, 232 Solstices, 704 Sorby on artificial production of foliation, 422; on Cleve- land ironstone, 412 ; on fluid- cavities, 310, note ; on forma- tion of limestone, 230 ; on magnesian limestone, 411 ; on metamorphism, 440 ; on ripple-drift in mica-schist, 400; on rounded grains in blown sand, 256; on sinking of sand and clay in water, 209 ; soluble salts in dolo- mite, 244 South America, Darwin on foliated rocks of, 422 South Staffordshire, abortive boring for coal in, 527 South Wales, denudation of, 574 South Yorkshire contrasted with Lancashire in surface features, 601 Spain, rain-wash of, 204 Spatheisenstein, 138 Spathic iron ore, 14, 138 Specific gravity of minerals, 88 Speckstein, 120 Specular iron ore, 14, 147 ; for- mation of in lava, 362, 563 ; in microscopic sections, 155 Speisskobalt, 540 Spencer on North American lakes, 618 Sperenberg, bore-hole at, 644 Sphaerosiderite, 138 Sphserulites, 306 Sphalerite, 534 Sphene, 131 Sphenoid, 61 ; rhombic, 66 Spheroidal state of water, 354 Spheroidal structure, 281, 307 Spinel, 334, note Splint coal, 185 Splintery fracture, 86 Sporadosiderites, 664 Spotted schist. 417 Spring, Walth., experiments of on effects of pressure in meta- morphism, 438, note Springs, deep-seated, 196 Sprodglaserz, 537 Sprudelstein of Carlsbad, 282 Stages in volcanic action, 348 Stalagmite, 243 Stalagtites, 243 Stannite, 536 Stassfurt, rock-salt of, 241 Stassfurtit, 162 Statuary marble, 407 Staurolite, 125 Staurolite-schist, 415 Steam coal, 181, 185 Steam, motive force in volcanic eruption, 345 Steatite, 120 Steinmark, 134 Steinol, 382, note Steinsalz, 161 Steplmnite, 537 Sterry Hunt on formation of dolomite, 246 ; gypsum in Canada, 412; metamorphism, 439 ; volcanic action, 678 Stibnite, 542 Stigmaria, 258 Stilbite, 129 Stinkstone, 179 Stock-werke-porphyr, tin-bear- ing, of Altenberg, 562 Stockwerks-porphyr, 316 Stock work, 555 Stone bind, 178 Stone-cavities, 310 Stone coal, 185 Storm waves, force of, 206 Strahlkies, 150 Strahlstein, 107 Strata, 168 Stratification, 165, 219 Stratified rocks, 167, 169 Stratula, 168 Streak of minerals, 86 Strike, 460 Stromboli, augite from, 357 Strontianite, 145 Strontium, tests for, 145 Structure en eventail, 424 Sublimation, 339 Submarine volcanic eruptions, 352 Submergence produced by ice- cap, 459 Subsidence after volcanic eruption, 348 Sulphur, 13, 160 ; formation of deposits of, 250 Sun, a variable star, 717 Sun-cracks, 217 Sunderbunds, 259 Sutherlandshire, mountains of, 573 Swallow-holes, 191 Switzerland, former extension of glaciers of, 700 Syenite, 316, note, 324 Syenit-porphyr, 325 Sylvine, 161 Sylvite, 161 Symmetry, axes of, 29 ; of plane figures, 26 ; of solids, 27 ; planes of, 25, 29 Synclinal, 476 Syssiderites, 664 Szabo, his method of distin- guishing the felspars, 99 TABERG, "iron ore" of, 66 Tabular cleavage, 86; struc- ture, 307 Tachylite, 333 Talc, 120 Talc-schist, 415 Talkspath, 138 Tanan, mud volcanoes of, 382 Taste of minerals, 88 Temperature, rate of increase downwards, 643; stratum of, invariable, 643 Tenacity of minerals, 15, 86 Tennantite", 530 Tennessee, drainage of,613, note Tenorite, 530 ; formation of, 563 Tephrite, 332 Terra del Fuego, foliated rocks of, 418 Terrestrial deposits, 251 Teschenite, 333 Tetartohedral, 33 Tetarto-octahedrons, 69 Tetragonal scalenohedron, 62 Tetragonal trisoctahedron, 47 Tetrahedrite, 529 Tetrahedrons, 17; hexakis, 49; regular, 38 ; triakis, 47 Tetrahexadron, 42 7 28 Index. Thalassic rocks, 289 Turgite, 149 bution of, 686 ; extinct, 363 ; Thames, 607 ; dissolved matter Twin axes, 75 ; crystals, 75 ; mud-, 381 in, 194, 205 polysynthetic, 95 Vose on relation of pressure Tharsis Mine, 564 Twinned octahedron, 75 and metamorphism, 685, note Thickness of beds, 219 Twinning plane, 75 Vughs, 549 Thickness of strata, measure- Twisden, Rev. T. F., on change Vulcano, 351 ment of, 464 in position of the poles, 716 Thomson, Sir W., on geological Tyrol, dolomites of, 408 ; earth WALKERERDE, 134 time, 693 ; on rigidity of the pillars of, 190 Walkthon, 134 earth, 657, 658; on under- Wallace on combined effect of ground temperature, 650 ; on uniformitarianism, 695 UINTA Mountains, 609 Unconformity, 514; between geographical and astronomi- cal causes on climate, 713 ; Thonglimmerschiefer, 415 magnesian limestone and coal on estimating geological time Thonschiefer, 406 measures, 522 ; detected by by evolution, 694; on per- Thurston on bending rigid mapping, 520 ; incidental manence of continents, 688, bodies, 512 proofs of, 521 note Tidal friction, 649 Underclay, 258 Wallace on lodes of Alston Tid-holes, 549 Undercut rocks, 202 Moor, 559 Tietze on breaching of moun- Underground streams, 195 Warp, 219 tain-ranges by rivers, 613, Underlie, 492, 550 Warrant, 258 note Undulations, 473 Water, 13; in lava, 354; its share Tilestone, 168 Uniaxal crystals, 71 in metamorphism, 436 Till, 263 Uniformitarian geology, 694 Wavellite, 159 Time, geological, 693 Unstratitied rocks, 169 Wave motion, 384 Tin, tests for, 536 Upheaval by intrusion of Weald of Kent, drainage of, Tin pyrites, 536 granite, 670 ; contraction 607 Tinstone, 536 theory of, 672 Wealden beds, formation of, Tin- white cobalt, 540 Uralite, 111 294 Titaneisen, 148 Uralite-porphyry, 330 Wedge-shaped sandstones in Titaniferous iron ore, 148 Uraninite, 547 Yorkshire, 212, 213 Titaniferous iron sand, 148 Uranium, tests for, 547 Weissbleierz, 533 Titanite, 131 Uranpecherz, 547 Werner, 4 Titanium, tests for, 148, 546 Utah, Salt Lake of, 239 Wernerite, 130 Titano-morphite, 155 Western Territories, lava-fields Toadstone, its effect on lodes, VARIATIONS of temperature, of, 352; metamorphic granite 554 pressure, and fusing-point of, 433 Tonalite, 329 within the earth, 654 Wey river, 608 Topaz, 125 ; Oriental, 158 Veinstuff, 548 Whitehaven, dolomitic rocks Topley on valleys of the Weald, 597, 638 Verglaste sandsteine, 404 Vesicular top of lava-flow, 356 near, 410 White iron pyrites, 150 Torbenite, 547 Vesicular structure, 311, 356 White lead ore, 533 Toughness, 86 Vesuvius, 350 Wilson on contortion, 508, note Tour, C. de la, liis experiments Vesuvian minerals, 358 Wind, denuding action of, 201 on boiling, 355 Virginia, Great Dismal Swamp Wismuthglanz, 544 Tourmaline, 129 of, 257 Wismuthocker, 544 Tourmaline-schist, 416 Viridite, 123 Witham river, 607 Trachyte, 325 Vitreous copper ore, 529 Witherite, 144 Trachyt-pitchstone, 322 Vitreous silver ore, 537 Wolfram, 152 ; tests for, 545 Transverse valleys, 595 Vitrified sandstone, 404 Wolframocker, 546 Trapezium, 76, note Vitriolbleierz, 532 Wood tin, 536 Trapezohedron, 18, 49; hexa- gonal, 59; tetragonal, 62 Vitro-porphyr, 322 Vivian ite, 153 Wurtz on relation of pressure to metamorphism, 685 Traprain Law, 373 Volcanic action, stages in, 348 Trass, 361 Volcanic agglomerate, 359 ; YAMPA river, 609 Travertine, 242 necks of, 372 Yar river, 60S, (311 Tremolite, 107 Volcanicash, 344, 358 ; altered, Yenite, 151 Triakis tetrahedron, 47 414 Triassic rocks, formation of, 301 Volcanic bombs, 359 ZARATITE, 539 Trichites, 305 Volcanic breccia, 360 Zeolites, 129 Trichroism, 86 Volcanic cones, 346, 631 ; trun- Ziegelerz, 530 Triclinic felspars, twinning of, cation of, 347 Zinc, tests for, 534 95 Volcanic eruptions, 343 ; sub- Zinc blende, 534 Triclinio system, 31, 68 marine, 352 Zincite, 535 Tridymite, 90 Trigonal trisoctahedron, 45 Trimetric system, 31, 63 Volcanic eruptive rocks, 370 Volcanic glass, 322 Volcanic group, 348 Zincspath, 535 Zinnkies, 536 Zinnober, 544 Trimorphism, 79 Volcanic mud-streams, 360 Zinnstein, 536 Trinidad, mud volcanoes of, 382 Volcanic necks, 370, 373 Zinnwald, tin-bearing greisen Triplosporites, 183 Volcanic rocks, definition of of, 562 Tripoli, 231 342; identity with lavas of, Zinnwaldite, 116 TronR 251 343, 378 ; intrusive and con- Zircon, 133 Trough, 476 Truckee Range, basalt ridges temporaneous, 369; irruptive and eruptive, 370 ; lithologi- Zircon-syenite, 325 Zirkel on included blocks in in, 353 cal varieties of, 378 granite, 395 Tufa, calcareous, 242 Volcanic tuff, 344 Zoisite, 128 Tuff, volcanic, 344, 360 Volcanic veins, 370 Zones, 25 Tungsten, tests for, 152, 545 Tungstite, 546 Volcanoes, association of with mountain-chains, 691 ; distri- Zwitter, 316 ; tin-bearing, of Altenberg, 562 MUIR, PATERSON, AND BRODIE, PRINTERS, EDINBURGH. 14 DAY USE RETURN TO DESK FROM WHICH BORROWED EARTH SCIENCES LIBRARY This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. DEC 2 8 1972 LD 21-40m-5/65 (F4308slO)476 General Library University of California Berkeley x &THS 1 >.*.\ ; -.