/ UNIVERSITY OF CALIFORNIA AT LOS ANGELES MINERAL DEPOSITS BY WALDEMAR LINDGREN PROFESSOR OF ECONOMIC GEOLOGY, MASSACHUSETTS INSTITUTE OF TECHNOLOGY; FORMERLY GEOLOGIST, UNITED STATES GEOLOGICAL SURVEY SECOND EDITION REVISED, ENLARGED AND ENTIRELY RESET SECOND IMPRESSION McGRAW-HILL BOOK COMPANY, INC. NEW YORK: 239 WEST 39TH STREET LONDON: 6 & 8 BOUVERIE ST., E. C. 4 1919 6114 1 - COPYRIGHT, 1913, 1919, BY THE MCGRAW-HILL BOOK COMPANY, INC. PREFACE TO THE SECOND EDITION The last few years of scientific progress frequently loom up in magnified proportions. This is a view that may be corrected by a proper perspective but to the author of a text-book anxious to carry his readers up to date it is certainly disconcerting to measure the rapid accumulation of new facts and theories. In spite of the disturbed condition of the world during the last six years, students of minerals and mineral deposits have made many important contributions to science. Among these may be counted the investigations bearing on magmatic and con- tact-metamorphic ore deposits, on problems of oxidation and supergene sulphides, and particularly the application of metallo- graphic methods to ores by which the complexity of metallic replacements has been revealed. In this second edition all of the chapters have been revised and those on deposition of minerals, contact-metamorphism, oxidation and sulphide enrichment, have been largely rewritten. A discussion of metallogenetic epochs has been added, as well as an index by elements which will enable the student to coordi- nate rapidly the deposits of any given metal. Many new illus- trations have been introduced. Owing to war conditions it has not always proved easy to obtain late statistical data. In case of some European countries the latest figures available date from 1913. The progress of science emphasizes the difficult task of con- densation. In tha present volume some less essential descrip- tions have been^omitted so that its bulk has not been greatly increased. No one knows better than the author that errors may easily creep in and corrections and suggestions will be grate- fully received. The genetic arrangement has been preserved throughout. While it is realized that this may make the study less easy for beginners, it is believed that any merit that the book may have is due to this mode of treatment. CAMBKIDGE, April, 1919. PREFACE TO FIRST EDITION Mineral deposits are usually classified and described by the metals or the substances which they contain; for instance, de- posits of copper are described together, with little or no effort to separate them into genetic groups. Where a genetic treatment has been attempted it appears to me to have failed in not giving due weight to the physical conditions attending the genesis. Furthermore, it is the custom to divide the mineral deposits into two groups the metallic and the nonmetallic a line of division which can hardly be defended except on the ground of long-established habit. This book is the outcome of a desire to place the knowledge of mineral deposits on the broader and more comprehensive basis of a consistent genetic classification and thus bring it into a more worthy position as an important branch of geology. Opinions may differ as to whether our present knowledge is sufficient for such an undertaking. Believing that the time has come for a first attempt, I present this volume, in the hope that its short- comings may be judged leniently. The impetus to the work came during the preparation of a series of lectures a few years ago, and a course along the general lines followed in this volume has since then been presented an- nually at the Massachusetts Institute of Technology. The general plan has been to select a few suitable examples to illustrate each genetic group of deposits. These examples have been chosen regardless of their geographic location, and it was of course necessary to give up any attempt to describe deposits in detail or to present all known examples of any particular type. As the larger part of my experience has been within the United States of America, a considerable number of examples were gathered from this country. This experience I owe to the United States Geological Survey, in which I have had the honor to serve for many years. My indebtedness to my friends and associates in that organization is greater than can be expressed in words. BOSTON, August, 1913. VII CONTENTS PAGE PKEFACE v CHAPTER I INTRODUCTION 1 Economic geology Mineral deposits Definitions Technical utility Ore and gangue Distribution of the elements Composi- tion of the earth's crust Traces of metals in rocks General state- ment Copper Lead and zinc Gold and silver Tenor of ores Iron Copper Lead Zinc Silver Gold Tin, etc. Price of metals Production of ore and metal Weights and measures Conversion tables. CHAPTER II THE FORMATION OF MINERALS 22 Solution and precipitation General features Influence of pres- sure Influence of temperature Precipitation by evaporation of the solvent Precipitation by reaction between solutions Precipi- tation by reactions between aqueous solutions and solids Precipi- tation by reactions between gases or between gases, and solutions Crystalline minerals Colloids. CHAPTER III THE FLOW OF UNDERGROUND WATER 29 General statement Pores and openings in rocks Water in sands and gravels Water in rocks of uniform texture Water in sedi- mentary rocks Influence of fractures Influence of volcanism Conclusions Examples of movement of water Depth of water level Total amount of free water in earth's crust. CHAPTER IV THE COMPOSITION OF UNDERGROUND WATERS 42 Introduction Calcium carbonate waters in igneous rocks Calcium carbonate waters in sedimentary rocks Chloride waters in sedimentary rocks Chloride waters in igneous rocks Sulphate waters in sedimentary rocks Acid sulphate waters in igneous rocks Mine waters of sulphate type Sodium carbonate waters in sedimentary rocks Sodium carbonate waters in igneous rocks Sodium sulphide waters Summary Interpretation of water analyses. x CONTENTS CHAPTER V PAGE THE CHEMICAL WORK OF UNDERGROUND WATER 66 Metamorphism and mineral deposits Stability of minerals and rocks Metamorphism Metasomatism or replacement The law of equal volume General definition of the metamorphic zones Zone of weathering The intermediate zone The deeper zones Relation of mineral deposits to the metamorphic zones Deposits related to igneous activity Derivation of minerals Concentra- tion Underground temperatures. CHAPTER VI THE ORIGIN OF UNDERGROUND WATER AND ITS DISSOLVED SUB- STANCES 86 Origin of the water Meteoric waters Magmatic or juvenile waters Examples of springs in volcanic regions Salts from sedi- mentary rocks Salts from igneous rocks Salts of volcanic springs Origin of the dissolved gases Rarer elements contained in waters The igneous emanations. CHAPTER VII THE SPRING DEPOSITS AT THE SURFACE 99 Deposits of limonite and calcium carbonate Deposits of silica Deposits of other gangue minerals Summary. CHAPTER VIII RELATIONS OF MINERAL DEPOSITS TO MINERAL SPRINGS 109 CHAPTER IX FOLDING AND FAULTING 115 Folds Faults General terms General classification of faults- Faults of parallel displacement Faults in stratified rocks Slip Shift Throw Offset Faults classified according to the direction of the movement Classes of strike faults Extension of the words normal and reserve to diagonal and dip faults Special classes of faults Rotatory faults Mineralization of faults Complexity of faulting. CHAPTER X OPENINGS IN ROCKS 137 Origin of openings By the original mode of formation of the rocks By solution By fractures of various modes of origin Force of crystallization. CONTENTS xi CHAPTER XI PAGE THE FORM AND STRUCTURE OF MINERAL DEPOSITS 147 Syngenetic deposits Epigenetic deposits Spacial relations of veins Veins in relation to the country rock Vein walls Out- crops Length and depth of veins. CHAPTER XII THE TEXTURE OF MINERAL DEPOSITS 161 Filling and replacement Introduction Texture of deposits of igneous origin Texture of pegmatite dikes Texture of sedimen- tary deposits Concretions Texture of residual and oxidized deposits Texture of epigenetic deposits Primary texture of filled deposits Secondary texture and structure of filled deposits Metasomatism in mineral deposits Metasomatic processes Mode of replacement Texture of metasomatic rocks Replace- ments at high temperature Replacements at intermediate temper- ature Replacement at low temperature Criteria of replacement Role of colloids in filling and replacement. CHAPTER XIII ORE-SHOOTS 182 Form of primary ore-shoots Shoots of successive mineraliza- tions Superficial or secondary shoots Causes of primary ore- shoots Decrease of pressure and temperature Character of wall rock Impermeable barriers Intersections. CHAPTER XIV THE CLASSIFICATION OF MINERAL DEPOSITS 195 Classification by form and substance Genetic classifications Outline of proposed classification Detrital and sedimentary de- posits Concentration of substances contained in the rocks Residual weathering Deep circulating waters Regional meta- morphism Zeolitization Introduced ores not connected with igneous rocks Deposits genetically connected with igneous rocks Products of magmatic differentiation Metamorphism and surface enrichment of deposits. A classification of mineral deposits. CHAPTER XV DEPOSITS FORMED BY MECHANICAL PROCESSES OF TRANSPORTATION AND CONCENTRATION; DETRITAL DEPOSITS 207 Introduction Detrital quartz deposits Detrital clay deposits Fuller's earth Placer deposits Origin and distribution Gold placers Origin of placer gold Eluvial deposits Processes of CONTENTS PAGI concentration Eolian deposits Stream deposits Classification of fluviatile and marine placers Marine placers Buried placers Size and mineral association of placer gold Fineness and relation to vein gold Gold in relation to bed-rock Grade of auriferous watercourses The pay streak or "run of gold" Solution and precipitation of gold Relation to primary deposits Economic notes The gold-bearing conglomerates of South Africa Platinum placers Cassiterite placers Monazite placers Other placers. CHAPTER XVI DEPOSITS PRODUCED BY CHEMICAL PROCESSES OF CONCENTRATION IN BODIES OF SURFACE WATER BY REACTIONS BETWEEN SOLUTIONS . 247 Limestone Definition and origin Chalk Lithographic stone Hydraulic limestone Lime Uses Dolomite Importance of car- bonate rocks as related to ore deposits Cherts and diatomaceous earth Sedimentary sulphide deposits Sedimentary iron ores Limonites in swamps and lakes (bog iron ores) Occurrence Composition Origin Examples The siderites of marine and brackish-water strata Occurrence Examples The Jurassic siderites of England The oolitic marine limonites and hematites The oolitic limonites Occurrence Examples Origin The ma- rine oolitic silicate ores The marine oolitic hematite ores Occur- rence The Clinton ores The brazilian hematites The oolitic hematite-chamosite-siderite ores Review of the sedimentary iron ores Sedimentary manganese ores Bog manganese ore Manganese in lacustrine and marine beds Sedimentary phosphate beds Composition of the calcium phosphates Other phos- phates Phosphate deposits Use Production Origin of the phosphate rocks Occurrences of phosphate rocks. CHAPTER XVII DEPOSITS FORMED BY EVAPORATION OF BODIES OF SURFACE WATERS . . 287 The saline residues Introduction Types of water Normal suc- cession of salts Structural features Gypsum and anhydrite Occurrence Uses Stability and solubility Sodium sulphate and sodium carbonate Occurrence Sodium nitrate Borates Gen- eral occurrence Marine borate deposits Borax marshes Terti- ary lake beds Production and uses Origin Soxlium chloride Occurrence Examples The salt deposits of the gulf coast- Composition, production and use The German potassium salts Other sources of potassium salts Potassium in rocks and min- erals Potassium in brines Bromine and calcium chloride. CONTENTS xiii CHAPTER XVIII PAGE MINERAL DEPOSITS RESULTING FROM PROCESSES OP ROCK DECAY AND WEATHERING 319 General conditions Decomposition of minerals Total chemical changes by weathering Residual clay Occurrence Uses and properties Origin Residual iron ores (limonite and hematite) Origin Classification Brown hematites of the Appalachian region Iron ores of Bilbao, Spain Residual ores of Cuba Distribution and stability of residual iron ore Residual manga- nese ores Primary sources Manganese deposits in the United States Brazil India Origin Residual barite Residual zinc ore Residual ochers Residual phosphates Deposits of hy- drated silicates of nickel Bauxite Introduction Origin Occur- rences Uses and production. CHAPTER XIX THE HEMATITE DEPOSITS OP THE LAKE SUPERIOR REGION 357 General character and distribution Geology The "iron forma- tions" The iron ores Form of ore bodies Marquette range Menominee range Penokee-Gogebic range Cuyuna range Mesabi range Vermilion range Origin of Lake Superior iron ores Re'sume'. CHAPTER XX DEPOSITS FORMED BY CONCENTRATION OF SUBSTANCES, CONTAINED IN THE SURROUNDING ROCKS, BY MEANS OF CIRCULATING WATERS . 375 General statement Barite Modes of occurrence and origin Deposits in the United States Foreign deposits Uses and production Celestite and strontianite Sulphur Modes of occur- rence Origin of sulphur deposits in gypsum Examples Produc- tion Uses Sulphuric acid The magnesian deposits Serpen- tine Magnesite Origin Occurrence Production and use Meerschaum Talc and soapstone General occurrence and ori- gin Occurrences Production and uses Pyrophyllite Asbes- tos Amphibole asbestos Serpentine asbestos (chrysotile) Uses Ores of copper, lead, vanadium, and uranium in sandstone and shale General features Origin Copper and lead deposits in sandstone European occurrences American occurrences South America Africa Genesis of sedimentary copper ores Vanadium and uranium ores in sandstones Composition Occur- rence Genesis Production and use The copper-bearing shales of Mansfield Copper sulphide veins in basic lavas General features The Nikolai greenstone Copper sulphide veins in intru- sive basic rocks Other veins deposited by waters of the upper circulation. xiv CONTENTS CHAPTER XXI PAGE DEPOSITS RESULTING FROM REGIONAL METAMORPHISM 421 CHAPTER XXII DEPOSITS OF NATIVE COPPER WITH ZEOLITES IN BASIC LAVAS .... 425 General statement Origin of the zeolitic copper ores Probable source of copper The occurrence of zeolites and the process of zeolitization The Lake Superior copper deposits General occur- rence The Calumet conglomerate The amygdaloids The veins Mineral association Origin Mine waters Rock alteration Mining and smelting operations The copper deposit of Monte Catini Native copper with epidote in basic lavas (Catoctin type). CHAPTER XXIII LEAD AND ZINC DEPOSITS IN SEDIMENTARY ROCKS; ORIGIN INDEPEND- ENT OF IGNEOUS ACTIVITY 444 Characteristic features Origin Moresnet Silesia Alpine Trias Other European localities The lead-zinc ores of the Mississippi Valley. CHAPTER XXIV METALLIFEROUS DEPOSITS FORMED NEAR THE SURFACE BY ASCENDING THERMAL WATERS AND IN GENETIC CONNECTION WITH IGNEOUS ROCKS 465 Character and origin General features Successive phases of mineralization Zeolitic replacement Primary ore shoots, oxida- tion, and sulphide enrichment Types of deposits Older repre- sentatives of this class Genesis Proof of depth below surface Proof of temperature Relation to other veins Metasomatic processes Extent of alteration Types of alteration Metaso- matic processes at Thames and Waihi Metasomatic processes at Tonopah The development of kaolin Metasomatic processes at Silverton, Colorado Summary Quicksilver deposits The ores and their general occurrence Distribution, production and use Geological features Mineralogy of quicksilver ores Structure Genesis Relation to other ore deposits Stibnite deposits Mineralogy, production and uses Occurrence Gold- CONTENTS xv PAGE quartz veins in andesite Transylvania Hauraki Peninsula, New Zealand El Oro, Mexico Gold-quartz veins in rhyolite Argentite-gold-quartz veins Tonopah, Nevada The Comstoick lode Argentite veins Gold telluride veins Cripple Creek Gold selenide veins Occurrence of selenides Republic, Washington Sumatra The base-metal veins The San Juan region, Colo- rado General features Telluride district Silverton district Ouray district Rico district La Plata, Durango, and Needle Mountains quadrangles Lake City district Creede district Summary Gold-alunite deposits General features Goldfield, Nevada. CHAPTER XXV METALLIFEROUS DEPOSITS FORMED AT INTERMEDIATE DEPTHS BY ASCENDING THERMAL WATERS AND IN GENETIC CONNECTION WITH INTRUSIVE ROCKS 546 General features Metasomatic processes General character Al- teration of wall rocks adjoining gold-quartz veins Interior types Paragenesis Gold-quartz veins of the California and Victoria type Principal characteristics Gold-quartz veins of the Sierra Nevada The gold-quartz veins of the interior Cordilleran region Victoria, Australia New South Wales and Queensland Nova Scotia Gold-arsenopyrite deposits Gold-bearing replacement deposits in limestone Gold-bearing replacement deposits in quartzite Gold-bearing replacement deposits in porphyry The Silver-lead veins General features Quartz-tetrahedrite-galena veins Tetrahedrite-galena-siderite veins (Wood river type) Galena- siderite veins Lead-silver veins with calcite, siderite, and barite Pyritic galena-quartz veins The silver-lead replacement deposits in limestone General features Park City, Utah Tintic, Utah Aspen, Colorado Leadville, Colorado The Lead- ville-Boulder County belt The tungsten deposits of Boulder County Summary Deposits with native silver The zeolitic en- richments The silver-bearing cobalt-nickel veins of Saxony The silver-bearing cobalt-nickel veins of Ontario, Canada Quartz-adu- laria-zeolite veins (.Alpine type) Occurrence and mineral associa- tion Origin The copper veins Chalcopyrite-quartz veins Bornite-quartz veins Pyrite-enargite veins The pyritic replace- ment deposits Copper deposits of Shasta County, California The pyritic deposit of Mount Lyell, Tasmania The pyritic deposits of Rio Tinto, Spain General features Geological for- mations The ores Genesis The pyritic deposit of Rammelsberg, Germany Geology and structural features The ores Origin Cadmium ores Arsenic deposits Fluorite deposits Siderite deposits. xvi CONTENTS CHAPTER XXVI PAGE VEINS AND REPLACEMENT DEPOSITS FORMED AT HIGH TEMPERATURE AND PRESSURE AND IN GENETIC CONNECTION WITH INTRUSIVE ROCKS . . . . 651 General features High-temperature minerals Metasomatic proc- esses Temperature and pressure Classes of deposits Mode of fissuring and filling The cassiterite veins Mineral association Metasomatic processes General features Metasomatic proc- esses in the deposits of Cornwall Development of greisen Alteration of sedimentary rocks Origin of tin-bearing veins The cassiterite veins of Cornwall, England Literature Cassiter- ite veins of Saxony Tin deposits in other countries Wolframite veins Gold-quartz veins The veins of the southern Appala- chians The quartz veins of Ontario The pre-Cambrian gold veins of the Cordilleran region The gold-bearing veins of Brazil The gold-quartz deposits of Silver Peak, Nevada The gold-quartz veins of southeastern Alaska Metasomatic processes in veins of southeastern Alaska The gold-telluride veins of western Aus- tralia The gold-copper deposits Copper deposits The copper- tourmaline deposits Chile United States The copper-bearing veins allied to contact-metamorphic deposits and pegmatites Copper-titanium veins Copper molybdenum veins The lead- silver-zinc deposits Veins with tourmaline Veins with garnet The cobalt-tourmaline veins. CHAPTER XXVII DEPOSITS FORMED BY PROCESSES OF IGNEOUS METAMORPHISM .... 704 Introduction General features History Contact-metamorphism General features Form and texture Mineralogy Intensity of metamorphism Influence of composition of igneous rock Alteration of the intrusive rock Succession of events Suc- cession of minerals Volume relations Mode of Transfer Phys- ical conditions at the contact Depth of formation Piezo-meta- morphism Principal types of contact-metamorphic deposits Magnetite deposits General character Foreign occurrences Fierro, New Mexico Heroult, California Iron Springs, Utah Cornwall, Pennsylvania Chalcopyrite deposits General charac- ter New Mexico Clifton, Arizona Bisbee, Arizona Silver Bell, Arizona Cananea, Mexico Bingham, Utah Ketchikan, Alaska Zinc and lead deposits Gold deposits Gold-arseno- pyrite type Telluride type Cassiterite deposits Titanium deposits Scheelite deposits Graphite Properties General oc- currence and origin Occurrences Production and uses Garnet Deposits due to igneous metasomatism not distinctly related to contacts General features Boundary district Ducktown, Tennessee Franklin Furnace, New Jersey Metasomatic mag- netite deposits of Sweden Magnetite deposits in the United States. CONTENTS xvii CHAPTER XXVIII PAGE MINERAL DEPOSITS OP THE PEGMATITE DIKES 760 Introduction Mineralizers and the nature of their action Temperature of consolidation Occurrence and general character Types of pegmatites Acidic pegmatites Basic pegmatites Eco- nomic features of pegmatite dikes Feldspar and quartz Mica- Oxide ores Wolframite Columbite and tantalite Yttrium, thorium, cerium minerals Monazite and zircon Apatite Lithium minerals Cryolite Precious stones Native metals, sulphides and arsenides Molybdenite. CHAPTER XXIX ' I INERAL DEPOSITS FORMED BY CONCENTRATION IN MOLTEN MAGMAS 780 Constitution of magmas and their differentiation and consolida- tion General features Constitution of magmas Crystallization of magmas Differentiation in magmas Principal types of de- posits Diamonds Other precious stones Platinum and palla- dium Production and use Iron and nickel Chromite Ilmenite or titanic iron ore General features Microstructure of ilmenite Irregular bodies Dikes Occurrences Influence of pressure Magnetite The iron ores of northern Sweden The magnetites of the Ural mountains The magnetites of the Adirondacks Corundum General mode of occurrence Corundum in igneous magnesian rocks Corundum in syenite Other occurrences Production in the United States Uses Sulphide ores of igneous origin General principles Types of deposits Sulphides in peri- dotites and gabbros Sudbury, Ontario Cape Colony Bornite deposits Injected pyritic deposits General features Bavaria Sweden Norway. CHAPTER XXX METAMORPHOSED DEPOSITS 822 Processes involved Deformed pyritic deposits Regionally meta- morphosed iron ores General features Swedish "dry ores" Norwegian ores United States The zinc ores of Ammeberg, Sweden. CHAPTER XXXI OXIDATION OP METALLIC ORES 829 General conditions Depth of oxidation Outcrops Nomen- clature Principles of oxidation Textures and criteria of the oxidized zone Textures of the supergene sulphide zone Solu- tion Precipitation Supergene sulphides Criteria of super- gene sulphide enrichment Iron Copper Minerals Solution and precipitations Supergene copper sulphides Theory of super- xviii CONTENTS PAGE gene copper sulphides The relation of chalcocite, covellite and bornite Oxidation of chalcocite zones Examples of oxidation of copper deposits General features Rio Tinto Mount Morgan Butte Ely Bingham The southwestern chalcocite deposits Globe Ray Chuquicamata Zinc Minerals Solubility and mineral development Supergene shoots of zinc ore Supergene zinc sulphides Lead Minerals Reactions in the oxidized zone Supergene sulphides Oxidation in the Coeur d'Alene district Oxidation in the Mississippi valley district Gold Examples of oxidation of gold deposits Silver Minerals Solubility and mineral development Precipitation Supergene sulphide enrich- ment Zones of supergene deposition Enrichment at Granite- Bimetallic mine Enrichment at Georgetown Enrichment at Tonopah Enrichment at Chanarcillo Other metals Platinum and palladium Mercury Secondary sulphides of quicksilver Cadmium Nickel and cobalt Chromium Manganese Tin Tungsten Vanadium Molybdenum Bismuth Arsenic Anti- * mony Mine waters Chloride waters Carbonate waters Sulphate waters Oxidation of pyrite Examples. CHAPTER XXXII METALLOGENETIC EPOCHS 909 Introduction Main epochs Europe Pre-Cambrian epochs Paleozoic epochs Hercynian epochs Permo-Triassic epochs Jurassic and Cretaceous epochs Tertiary epochs Asia Africa Australasia South America Central America The Antilles North America The pre-Cambrian epochs Paleozoic sedimentary epochs Paleozoic intrusives Paleozoic epochs of saline deposits Epochs of Triassic copper deposits Cretaceous and later periods of lead and zinc concentration Tertiary and recent periods of rock decay The pre-Cambrian epochs The early Mesozoic epoch The late Mesozoic epochs The early Tertiary epoch The late Tertiary epoch The post- Pliocene epoch Cretaceous or later epochs of copper concentration in sedimentary rocks Index to mineral deposits by elements. INDEX . 929 MINERAL DEPOSITS CHAPTER I INTRODUCTION ECONOMIC GEOLOGY The application of geology to the practical problems of the industries and the arts constitutes economic geology. TTiis branch of the science includes as its most important division the study of deposits of useful minerals, but it also teache? the oc- currence of underground waters, explains the derivation and constitution of soils in relation to agriculture, and applies geo- logic principles to the planning of important engineering works. Only a part of the whole field of economic geology will be covered in these chapters. They will be confined to a descrip- tion, by classes and type examples, of the occurrence, structure, and origin of the principal deposits of metallic and non-metallic minerals of economic importance. The subjects of coals, mineral oils, and structural materials could not be included without unduly increasing the bulk of the volume. Little space has been given to statistics, while the problems of correlation and origin have been treated rather fully. A general part de- scribing principles of universal application precedes the detailed characterization of the various classes. A complete treatment of the subject should include discussions of distribution, occurrence, structure, origin, production, and valuation of deposits, as well as statements of the uses of the materials mined, processes of mining and reduction, and criteria for judging the value of the products. Such a complete presen- tation will not be found in this volume, for it is believed that by examining the subject from a scientific rather than from a utilita- rian viewpoint, the student will obtain a clearer insight into the geologic relationship of the various deposits. 1 2 MINERAL DEPOSITS Throughout its broad domain economic geology stands on the fundamental sciences of chemistry and physics. It is related on one side to theoretical geology, paleontology, mineralogy, and petrography; on the other side to mining, metallurgy, and many other technologic arts. A student who tries to approach the subject without the necessary knowledge of the allied sciences and arts is 'building on poor foundations. Even with this aid the study offers peculiar difficulties. The alteration of rocks close to many mineral deposits is intense and, as a result, the student who is familiar with only the fresh, unaltered speci- mens finds himself in the midst of puzzling and strange types that he is unable to classify with certainty. Altered andesites may assume the aspect of quartzites; a question may arise as to whether a silicified rock was once a limestone or a porphyry; diabases may at some places be converted into white fine-grained calcite-sericite-quartz rocks and at other places appear as ag- gregates consisting mainly of epidote and chlorite. These examples suffice to show that rock alteration is a subject of prime importance for the mining geologist. MINERAL DEPOSITS Definitions. The outer shell of the globe is commonly called the earth's crust. Of this shell only the most superficial part is accessible. The radius of the earth is about 4,000 miles. The deepest shaft attains only about 6,000 feet, 1 the deepest bore-hole 7, 350 feet. This crust consists of manifold mineral aggregates formed at different times and in various ways. Each individualized mass of mineral aggregates such as a stratum, a lava flow, an intrusive mass of igneous rock, a dike, a vein,' or a lenticular mass is called a "formation," a "member," or in general a "geologic body." Geologic bodies which consist mainly of a single useful mineral for instance, beds of pure gypsum or coal or which contain, throughout or in places, valuable minerals that can be profitably extracted for in- stance, veins containing disseminated gold are called "mineral deposits." Geologic bodies that are not worked for any particu- lar mineral or minerals, but for the aggregate of minerals the rock itself are usually designated as deposits of the particular 1 The gold mine of Morro Velho, Brazil, p. 190. INTRODUCTION 3 rock. Thus a mass of roofing slate is not spoken of as a mineral deposit, but as a slate deposit. Economic geology treats of the occurrence, composition, structure, and origin of those geologic bodies which can be technically utilized; it shows where they may be searched for and how their value may be ascertained. 1 Technical Utility. The limitation of technical utility must of course not be taken too literally, especially where questions of origin are concerned, for here, as in many other phases of the subject, applied geology merges into theoretic geology. More- over, it is no uncommon occurrence that the useless of yesterday becomes the useful of to-day. Examples are easily cited. About 1900 the cupriferous monzonite of Bingham, Utah, which yields an average of 30 cents in gold and 14 cents in silver to the ton and 1.5 per cent, of copper, would probably not have been classed as an ore, but with modern methods of treatment it is an important ore of copper. The zinc minerals of the western States, valueless and even causing loss in the marketing of ores, can now be profitably sold. The tungsten ores of Colorado, thrown over the dump not long ago, have attained a value of $200 a ton. Low-grade gold ores for instance, those of Mercur, Utah considered as hopelessly refractory before 1890, became rich assets with the introduction of the cyanide process. Many iron ores rich in phosphorus were neglected until the Thomas process provided means for their profitable reduction. Monazite containing thorium acquired importance with the invention of the incandescent mantle, for gas burners. New processes of reduction, the rising price of some commodity, inventions calling for rare and unused metals any of these may suddenly cause a geologic body that has previously been^ valueless to become of great importance. Titanic iron ores form vast de- posits which are now useless because of metallurgical difficulties but which some day will, no doubt, be utilized. This principle also works the other way. Decreasing prices may make a particular deposit unprofitable; that is what happened to many silver mines during the great decline in the price of silver which began in 1880. Great changes, mainly in the direction of rising prices have been brought about by the great war beginning in 1914. A large number of metals have doubled in price: They include silver, platinum, copper, lead, zinc, tin, antimony and 1 Stelzner and Bergeat, Die Erzlagerstatten, vol. 1, 1904, p. 1. 4 MINERAL DEPOSITS aluminum. Gold alone, being the standard by which other values are measured, remains stable. Ore and Gangue. These considerations bring us to the terms ore and gangue. "Ore" is a word which has been used in several meanings. An "ore mineral" is a mineral which may be used for the extraction of one or more metals. An "ore, "as the term is used here, is that part of a geologic body from which the metal or metals that it contains may be extracted profita- bly. Thus galena and malachite are ore minerals. An ore is practically always a mixture of minerals. Local usage has adopted several terms as substitutes for "ore." In the lead- zinc district of Missouri crude ore is called "dirt," while con- centrates are called "ore." In Michigan the ore is called "rock" and the concentrates are termed "mineral." Gold-bear- ing gravels are not usually referred to as ore. The use of the term " ore " is not quite consistent. Ordinarily it implies a metal, but the expression "sulphur ore," meaning py rite, is sometimes seen, 'and occasionally such terms as "sapphire ore" are found. The useless minerals occurring in the ore are termed "gangue." Thus, a gold ore may consist of quartz, calcite, siderite, native gold, auriferous pyrite, and galena. Here the first three are called "gangue minerals." The terms are not inflexible; for example, siderite may under some circumstances be utilized as an iron ore. Moreover, as stated above, what to-day is useless gangue may prove valuable ore to-morrow. It is therefore safe to make the definition of an ore rather wider than the present technical limits. 1 It is hardly necessary to call attention to the differences in prices of metals which cause wide disparity in the amounts of different metals necessary to constitute ores. An iron ore must ordinarily contain at least 30 per cent, of iron usually much more. A volcanic rock containing 15 per cent, of iron is far from being an iron ore, but quartz containing 0.05 per cent, of gold is a rich gold ore, worth $330 a metric ton; in fact, is little as 0.0001 per cent, of gold, equivalent to 1 gram to the metric ton, or a value of 66 cents a ton, if occurring in an ore with other useful substances, is ordinarily paid for by smelting works. 1 For a full discussion of the subject see J. F. Kemp, "What is an Ore?" Jour. Canadian Min. Inst., vol. 12, 1910, pp. 356-367. Also, Min. and Sci. Press, March 20, 1909. INTRODUCTION 5 DISTRIBUTION OF THE ELEMENTS To obtain data regarding the relative distribution of the ele- ments, several calculations have been undertaken on the basis of a great number of reliable rock analyses. Especially notable are the papers of Clarke, 1 Vogt, 2 and Washington. 3 Clarke used 1,200 analyses of American rocks; Washington 1,800 from various parts of the world. We are here chiefly concerned with the solid crust of the earth, although in passing^it is deserving of notice that enormous quantities of salts are dissolved in the sea water, among them sodium chloride, sulphates of calcium, magnesium, and potassium, and carbonates of calcium and mag- nesium. The volume of salts in the sea water, according to Clarke, would be enough to cover the entire area of the United States (exclusive of Alaska) to a depth of 1.6 miles, or the whole globe with a stratum of sodium chloride 112 ft. deep. Ac- cording to the same authority, the crust of the globe 10 miles thick, with an assumed average specific gravity of 2.5, contains about 93 per cent, of solid matter and 7 per cent, of sea water. Composition of the Earth's Crust. In calculating the average composition of the accessible portion of the solid crust it is nec- essary to consider the sedimentary and the igneous rocks. The sedimentary rocks form but a thin veneer compared with the igneous rocks. The average composition of the latter closely approximates that of the crust. Clarke calculates that the crust to a depth of 10 miles is composed of 95 per cent, of igneous rocks, 4 per cent, of shales, 0.75 per cent, of sandstones, and 0.25 per cent, of limestones. Van Hise and others arrive at somewhat different figures. The sediments average poorer in calcium, magnesium, and especially in sodium than the igneous rocks and thus show the effect of leaching. They also contain more potash and carbon dioxide, but on the whole they are similar in composition to the igneous rocks. According to Clarke the average of analyses of igneous rocks made in the laboratories of the United States Geological Survey is as follows : 1 F. W. Clarke, Geochemistry: Bull. 616, U. S. Geol. Survey, 1916, pp. 22-35. Many partial analyses are also included. 2 J. H. L. Vogt, Ueber die relative Verbreitung der Elemente, etc. : Zeit- schr. prakt. Geol, 1898, pp. 225-238; 314-325. 3 H. S. Washington, The distribution of the elements in igneous rocks: Trans. Am. Inst. Min. Eng., vol. 39, 1908, pp. 809-838. MINERAL DEPOSITS AVERAGE ANALYSIS OF IGNEOUS ROCKS o 47.29 S 0.103 Si 28.02 Cl 0.063 Al 7.96 F 0.10 Fe 4.56 Ba 0.092 Mg 2.29 Sr 0.033 Ca 3.47 Mn 0.078 Na 2.50 Ni 0.020 K 2.47 Cr 0.033 H 0.16 V 0.017 Ti 0.46 Li 0.004 Zr 0.017 C 0.13 Total 100.000 P 0.13 The eight elements first named above make up 98.56 per cent, of the igneous rocks. Among the six principal metals shown in the average composi- tion only iron, magnesium and aluminum are of economic import- ance. The lighter elements predominate, the atomic weight of each falling below 56 (Fe 55.9). In the average composition, the rarer metals titanium, zirconium, barium, strontium, manganese, nickel, chromium, and vanadium are represented, but except titanium, which amounts to 0.45 per cent., all these metals average below 0.1 per cent. Platinum, gold, silver, copper, lead, zinc, antimony, arsenic, tin, quicksilver, molybdenum, tungsten, and others are present in amounts less than 0.01 per cent. For some of these more definite estimates have been made by Clarke and Steiger 1 from careful analyses of large, composite samples of rocks and clays. The average percentages are as follows: CuO, 0.0130; ZnO, 0.0049; PbO, 0.0022; As 2 5 , 0.0005. These figures considered as orders of magnitude have a high degree of probability; possibly they are a trifle too high. Silver may constitute 0.00001 and gold perhaps 0.0000005 per cent, of the crust. The percentages of the useful metals given above do not by any means indicate the amount available for industrial use. That amount indeed is so infinitesimal in relation to the volume of the crust that it can not be expressed on the basis of per- centages. The metals in the deposits of useful minerals then comprise only a minute fraction of the quantity of metals in the 1 Jour., Washington Acad. of Sci., vol. 4, 1914, p. 57. INTRODUCTION 7 crust a fraction which has been locally accumulated by this or that process of concentration. In general igneous rocks contain more of the heavy metals than do the sedimentary rocks. We are well justified in regard- ing the former as the original source of these metals. Dissipa- tion by solution accompanies sedimentation and the many metals found in traces in the sea water furnish evidence of this. On the other hand, it is true that certain kinds of sedimentation will result in a concentration of metals, such as iron, zinc, cobalt, nickel and vanadium. Vogt and Washington have also formulated some rules con- cerning the relationship of certain metals with certain rocks. It is obvious, however, that a distinction should be made as to whether the metal is an integral part of the rock or whether simply deposits of the metal occur in the rock. Thus, for instance, lead deposits are characteristic of many limestones, but it may be doubted whether lead is a primary constituent of limestone; it is present because the rock had the power of pre- cipitating the metal from its solution. In highly siliceous rocks, especially in granites and in the pegmatite dikes accompanying them, we find minerals contain- ing fluorine, boron, lithium, zirconium, tin, tungsten, tantalum, molybdenum, thorium, and beryllium. Highly sodic magmas are also accompanied by a great number of rare metals. On the other hand, basic rocks in which the darker con- stituents predominate contain phosphorus, sulphur, chlorine, copper, chromium, nickel, cobalt, titanium and vanadium, and some of them, chiefly peridotites, contain platinum and diamonds. Gold and silver exist in minute quantities in many rocks, particu- larly in those of acidic types, like granite and rhyolite. These rarer metals are not everywhere present in similar rocks. Platinum, for instance, is contained in the peridotites of the Ural Mountains, but the peridotites of the Sierra Nevada are poor in that metal and the similar rocks in the Coast Ranges of California contain little platinum, as do the serpentines of Asia Minor and of Italy. Similar conditions characterize the occur- rence of rarer metals in granitic rocks. TRACES OF METALS IN ROCKS General Statement. In order to formulate a theory or a hypothesis of the origin of mineral deposits it is most desirable 8 MINERAL DEPOSITS to ascertain to what extent the different rocks contain the rarer metals. J. G. Forchhammer and L. Dieulafait began examinations for this purpose about 1860 and found traces of silver, copper, lead, bismuth, nickel, cobalt, zinc, arsenic, antimony, and tin in many rocks. Somewhat later F. von Sandberger followed up this line of investigation and ascertained that the dark silicates of many rocks contained lead, copper, tin, antimony, arsenic, nickel, cobalt, bismuth, and silver. Some doubt has been expressed as to a few of these results and it is believed that in some of Sandberger's specimens the metals were derived from adjacent veins or from the reagents or the vessels used in the analyses. However, in spite of analytical difficulties, the pres- ence of many of those metals in various igneous and metamorphic rocks is clearly proved. Many ordinary analyses show the pres- ence of chromium, cobalt, and nickel in basic rocks like perido- tites, serpentines, and pyroxenites. Some of these rocks con- tain as much as 0.76 per cent, of Cr 2 03 and up to 0.3 per cent, of (Ni,Co)0. Traces of nickel and cobalt are often found in diabases, gabbros, and basalts; occasionally in diorites. A little vanadium is common in all rocks usually only 0.01 to 0.05 per cent of V 2 O 3 . In the following paragraphs some of the most reliable data regarding traces of rarer metals are compiled. More extensive references will be found in Clarke's "Data of geochemistry." Copper. A. C. Lane 1 states that the Keweenawan ''traps" average 0.02 per cent, of copper. F. F. Grout 2 found 0.029 and 0.02 per cent, in fresh specimens of the same series from Minne- sota. J. Volney Lewis 3 says that some of the New Jersey diabases or "traps" contain chalcopyrite and that copper is also present in the pyroxene of these rocks. The average, according to numerous analyses, is 0.025 per cent, of CuO. R. C. Wells found 0.03 per cent, of copper in a perfectly fresh basaltic lava from The Dalles, Oregon. Analyses made for W. H. Weed in the laboratory of the United States Geological Survey show that Butte granite or quartz monzonite from a quarry near Walkerville, Montana, contains 0.006 per cent, of copper. The quartz and feldspar, forming 1 Mine waters, Proc., Lake Superior Min. Inst., June, 1908, p. 86 2 Econ. Geol, vol. 5, 1910, p. 471. 1 Econ. Geol, vol. 2, 1907, pp. 242-257. INTRODUCTION 9 91 per cent, of the rock, contain little or no copper; the mica and hornblende, which constitute 7 per cent, of the rock, yield 0.047 per cent, of copper; there is thus nearly eight times as much copper in the ferromagnesian minerals as in the rock. The altered rock surrounding the veins carries more copper than the fresh rock. E. T. Allen examined 18 samples of fresh gabbros and diorites from the Encampment district, Wyoming, and found copper in all, the largest quantity noted being 0.02 per cent, of CuO. 1 A composite sample of seventy-one Hawaiian lavas yielded Geo. Steiger 0.0155 per cent, of copper. H. I. Jenssen found 0.034 per cent, of copper in an andesite from Fiji. 2 E. Coman- ducci reported 0.0854 per cent, of CuO and 0.0038 per cent, of CoO in volcanic ash from Vesuvius. 3 J. B. Harrison 4 examined 36 igneous and metamorphic rocks from British Guiana and found that 6 contained no copper and 12 contained copper in traces only; those carrying most copper were diabases and porphyrites; a feldspathic tuff yielded 0.13 per cent. The average for the series was 0.025 per cent, of copper. In some of these rocks the copper may have been contained in secondary disseminated sulphides. In a fresh granodiorite from Steamboat Springs, Nevada, W. H. Melville 8 detected copper, lead, arsenic, and antimony. J. D. Robertson 6 found from 0.0024 to 0.0104 per cent, of copper in granite, porphyry, and diabase from the Archean of St. Francis Mountain, in Missouri. The average was 0.006 per cent. Lead and zinc were also recognized. The adjacent Silurian and Carboniferous limestones also contained these metals, but in smaller quantities. Native copper in minute scales is rather common in shales and copper sulphides, contemporaneous with the metamorphism, occur in many amphibolites. Copper has been repeatedly de- tected in sea water and is contained in the red and blue mud dredged from the deep seas. 1 A. C. Spencer, Prof. Paper 25, U. S. Geol. Survey, 1904, p. 49. 2 Chem. News, vol. 96, 1907, p. 245. 3 Gazz. chim. ital., vol. 36, pt. 2, 1906, p. 797. 4 Report on the petrography of the Cuyuni and Mazaruni districts. Georgetown, Demerara, 1905. 5 Mon. 13, U. S. Geol. Survey, 1888, p. 350. 6 Missouri Geol. Survey, vol. 7, 1894, pp. 479-481. 10 MINERAL DEPOSITS From this evidence the conclusion may be drawn that prob- ably all igneous rocks contain appreciable amounts of copper and]that acidic rocks contain less than basic rocks. The copper is largely associated with the ferromagnesian silicates, and in the lavas at least it appears to be present as a silicate. Lead and Zinc. In analyzing the quartzose porphyries of Leadville, Colorado, supposed to be free from sulphides, W. F. Hillebrand 1 found that of 18 carefully selected samples 15 con- tained lead, the richest carrying 0.0064 per cent, of PbO; the average was 0.002 per cent. The same analyst found 0.008 and 0.0043 per cent, of ZnO in similar rocks. J. D. Robertson 2 determined an average of 0.004 per cent, of lead and 0.009 per cent, of zinc in the Archean rocks from Missouri mentioned above. J. B. Weems 3 determined lead and zinc in the limestones and dolomites of the Dubuque region, Iowa. The average of 9 samples gave 0.00326 per cent, of lead and 0.00029 per cent, of zinc. J. B. Harrison looked for lead in 23 samples of rocks from British Guiana and was able to determine the metal in five; the maximum obtained was 0.02 per cent. L. Dieulafait detected zinc in hundreds of samples of Jurassic limestone from central France. On the other hand, W. F. Hillebrand 4 was unable to detect lead or zinc in samples of limestone from Mexico, near important lead deposits. Henry W. Nichols 5 found no lead, copper, or zinc in calcareous concretionary deposits of the Challenger Banks, near Bermuda. Zinc is reported by Dieulafait in sea water and in ashes of sea weeds. Lead apparently does not exist in the water of the ocean. Gold and Silver. A large number of experiments have been undertaken to decide the question whether fresh igneous or sedi- mentary rocks contain gold and silver. In mining districts where the solution of this problem has been frequently attempted it is difficult to obtain perfectly satisfactory samples, free from con- tamination by circulating water. Furthermore, contamination is possible from fluxes, from the dust of assay rooms, from mortars, or from bucking boards. The mere statement that the assay 1 Mon. 12, U. S. Geol. Survey, 1886, pp. 591-594. 2 Missouri Geol. Survey, vol. 7, 1894, pp. 479-481. 3 Iowa Geol. Survey, vol. 10, 1900, p. 566. 4 Oral information. 6 Econ. Geol., vol. 2, 1907, p. 309. INTRODUCTION 11 indicates gold and silver in a rock is not sufficient. It must be corroborated by a statement of the methods used and accom- panied by the evidence of microscopic examination as to the freshness of the rock. It is satisfactorily proved that many fresh, massive igneous rocks contain gold. The best evidence thus far brought forward is probably that afforded by the granite from the Altar district, Sonora, Mexico, described by G. P. Merrill. 1 The gold occurs embedded in fresh quartz and feldspar. W. Moricke 2 found gold in a pitchstone from Chile and believed the metal to be primary. Gold was found by W. F. Ferrier in a fresh syenite from Kamloops, British Columbia. According to R. W. Brock 3 probably primary gold was found in a porphyry dike on North Fork of Salmon River, West Kootenai, British Columbia. Of twelve dikes of porphyry at different points in West Kootenai, six contained gold, most of them being wholly unaltered. Brock also states that a sample of alkali syenite porphyry in the Valkyr Mountains, east of Lower Arrow Lake, British Columbia, con- tained gold that was visible to the naked eye. Many of the statements of this sort in the literature must be critically scanned, for it is not uncommon to find gold deposited by mineralizing solutions in massive rocks, especially in schists, under circumstances closely simulating original deposition. Statements regarding primary gold in the chloritic schists of the Sierra Nevada refer to occurrences of this class. It is frequently said that gold occurs in pegmatite, but few of the assertions have the requisite backing of complete evidence. The probability is strong, however, that gold is present in such dikes. One of the most definite descriptions of this mode of occurrence is furnished by J. Catharinet. 4 Sperrylite, an ar- senide of platinum, is stated by Catharinet to occur with the gold. Another case is reported by C. De Kalb 5 from Mohave, California. References in the literature to primary gold in rocks from the Ural Mountains and from Australia do not appear to be sufficiently substantiated. Having procured samples of various rocks, most of them far 1 Am. Jour. Sci., 4th ser., vol. 1, 1896, p. 309. * Min. pet. Mitt., vol. 12, 1891, p. 195. 3 Eng. and Min. Jour., vol. 77, March 31, 1904. * Eng. and Min. Jour., vol. 79, 1905, p. 127. * Trans., Am. Inst. Min. Eng., vol. 38, 1908, p. 312. 12 MINERAL DEPOSITS from mining districts, Luther Wagoner, 1 of San Francisco, assayed them with the results as given in the accompanying table. His method consisted in cyanide treatment of 40 or 50 grams of material followed by blowpipe cupellation. All par- ticulars of the operations are detailed. Wagoner found that the purest obtainable reagents, such as soda, borax, and cyanide of potassium, contain gold and silver. A sample of Merck's "C. P." carbonate of soda contained 3 grams of silver to the ton. A sample of cyanide of potassium yielded 147 milligrams of gold and 26.05 milligrams of silver to the metric ton. Wagoner's results merit attention, but it would be desirable to have them checked. His figures for silver seem high. It will be noted that gold values obtained by him are as low as 5 milligrams, or 3i s 888S8 S3 o o d d o 1-1 o o O O rH 3 ' \ II ' a d 1 -2 \ -S fc ""1 E.I BA: AR&GR S i & 1 *1 ''*H\ ^ 8" (' ! I*. in x - ra 7? ll 1 II 11 1 t5 ^ II -J "*!;;* ; ^ " G ro 'e/[ J in n \ il 1 11 o N J \ o. ^ s S "i 1 -i: i 1 ; 1 1 riJi ^ D* o / - 1 s | 1 1 t- [ill ds i 1 3 \ '.>. up 1 != 1 1 1 "i-2 1" t t -1 * iL 1 M -g ! c S it ^ 3| / r C eo ? * \ 2 2 ^ S M 1 ^\ / eg J fl * 01 St , M 4<4 O o 1 1 ? -5-2 --" H! J Ul ', ~ a ', . -g SOI/17/ OYSTEE R ^1 fiylF distribution of the de- 34 MINERAL DEPOSITS scending surface waters is likely to be irregular. Near the sur- face there may be a local water level but below this beds heavily charged with water may alternate with almost dry strata. Each bed may in a way be considered as a unit and if it outcrops it has its own zone of gather- ing, zone of discharge and its static zone. The Cre- taceous Dakota sandstone presents an excellent ex- ample of a porous stratum' in which a large amount of water can be stored. Throughout the Great Plains this is a veritable reservoir of water, which can be tapped by artesian wells as far as 300 miles from its outcrop and at depths of a few hundred to 3,000 feet (Fig. 2). But this stratum at present simply contains a stagnant body of water, and, as in most other artesian basins, the quantity is not inex- haustible. This very case proves how impervious the adjacent sedimentary beds are, for neither upward nor downward is an avenue of escape afforded in spite of the strong pressure. Should profound fissuring take place in the Great Plains a natural avenue of escape would, of course, be opened and a deep cir- culation established. Kemp and Fuller have both brought out the fact that the deep sedimentary beds are often remarkably dry. The well 4,262 feet deep at Wheeling, West Virginia, was in absolutely dry rocks for the lower 1,500 feet. Wells sunk at THE FLOW OF UNDERGROUND WATER 35 Northampton, Massachusetts, and at New Haven, Connecticut, to depths of 4,000 feet have failed to obtain water. A number of other instances are mentioned, and in many cases the dry part consists of sandstones or other porous rocks. Some time ago it was suggested by A. C. Lane 1 that part of the salt water in deeply buried beds is fossil sea water or "con- nate" water occluded in the sediments at the time of deposition. There can be little doubt that dryness as well as salinity increase with depth. 2 Influence of Fractures. The simple conditions outlined above are seriously disturbed where extensive fracturing has taken place and paths have been laid out on which the water may move under approximately normal hydrostatic conditions. There may be a comparatively slow descent of the water along devious joints and fractures and a rapid rise under hydrostatic head where the descending water reaches the open paths of D E FIG. 3. Section illustrating flow of water in jointed crystalline rocks. A, C, flowing wells fed by joints; B, intermediate well of greater depth, with no water; D, deep well not encountering joints; E, pump well adjacent to D, ob- taining water at shallow depth; S, dry hole adjacent to spring. After M. L. Fuller, U. S. Geol. Survey. important faults and fissures. Friction during the descent undoubtedly seriously diminishes the theoretical head, but the evidence is perfectly clear that in regions of dynamic disturbance, such as the Alps and the Rocky Mountains, strong ascending springs may result from these conditions. At the point of issue such springs may be warm and their temperature, in regions where no recent igneous action has taken place, may be a good indication of the depth attained by the water. Such springs seldom have a temperature higher than 65 C., and the composition of their salts corresponds to the character of the beds traversed. On the supposition, believed to be well founded, that only a moderate loss in heat takes place Bull. Geol. Soc. Am., vol. 19, 1908, p. 502. 2 See Chapter VI. 36 MINERAL DEPOSITS during the ascent, a water of the temperature named would be derived from a depth of about 5,500 feet. Large regions of the earth, such as the Scandinavian peninsula, contain no warm springs, and the eastern part of the American Continent yields very few of them. Fuller says : The results of drilling in sedimentary and crystalline rocks, as well as the studies of deep mines, show that in all probability water does not commonly exist in the rocks under great pressure, although such may be exerted in an occasional crevice. It is not believed that hydrostatic waters exist, except possibly in rare instances, at depths of over 10,000 feet, and that in reality the estimate of a depth of 6,520 meters, or 20,000 feet, as the limit of the zone of open cavities is closely approximate to the truth If waters were freely circulating at great depths, within the zone of fracture, hot springs would certainly be more common along the numerous faults of the Piedmont, Appalachian, and similar regions. 1 Van Hise suggests that the decreased density and viscosity of water at higher temperatures may lessen the head necessary for ascending springs, but it may be doubted whether these factors would ever offset the great friction encountered during the down- ward passage. Faulting and mountain-building processes de- velop heat and this disturbance of the conditions of temperature may result in convection currents and an increased circulation of the water stored in the rocks. Influence of Volcanism. When magmas are intruded into the zone of fracture the conditions become more complicated. It is thought by some that atmospheric waters are able to de- scend into the deep regions and become absorbed by the magmas, but this view appears improbable. Before its irruption into the zone of fracture the magma is assuredly far beyond the reach of any waters percolating from the surface. Daubree's well-known experiment has often been cited, as showing how water may pass through a disc of sandstone against a certain counter pressure of steam. Recent critical examination 2 has shown the fallacies involved in the experiment, and indicate that "the probabilities are all against the notion that appreciable amounts of meteoric waters can ever penetrate into deep-seated and highly heated rock masses." 1 Water-Supply Paper 160, U. S. Geol. Survey, 1906, p. 64. 1 John Johnston and L. D. Adams, Observations on the Daubree experi- ment, etc., Jour. Geology, vol. 22, 1914, pp. 1-15. THE FLOW OF UNDERGROUND WATER 37 The presence of a heated body in the zone of fracture would undoubtedly quicken the circulation of water by inducing strong convection currents and expelling the stored water from its reservoir. Whether this action is sufficient to account for the remarkable number and volume of hot springs rising in volcanic regions may well be doubted, and it is thought that the magma itself gives up most of its constitutional water, partly when moving up to higher levels, partly when crystallizing to solid rocks. Conclusions. In conclusion it is believed that water in quantities sufficient to supply an ascending circulation can only exceptionally attain a depth of 10,000 feet and that, except in regions of great dynamic movements, the active circulation is confined to the uppermost few thousand feet. More commonly the depth of active circulation is measured by the level of surface discharge and the water below that level is practically stagnant; the lower limit of the body of stagnant water then forms an irregular surface descending to greater depths along the fractures and rising higher in the intervening blocks of solid ground. Examples of Movement of Water. According to Fuller, water supplies in wells in crystalline rocks are usually found within 200 or 300 feet of the surface and it is ordinarily useless to go below a depth of 500 feet (Fig. 3). The occurrence of porous strata which are capable of holding immense quantities of water but in which none whatever is actually found is, according to Fuller, a common experience of drillers in this country, even where the upper strata contain a well-defined water table. Investigations of joints in the crystalline rocks of Connecticut have shown, according to Fuller, that the water occurs largely in the vertical joints, which have an average spacing of 3 to 7 feet at the surface. In depth these joints diminish rapidly or close up and it is therefore not advisable to go below 250 feet in search of water. It is estimated by E. E. Ellis that the water present in the upper 2,000 feet of the crystalline rocks is only 16 per cent, of their capacity, or 0.000007 of the rock volume. The evidence from many mining regions is of considerable importance. In the Sierra Nevada of California deep canyons are separated by broad-backed ridges capped with Tertiary gravels and andesitic tuffs. The abundant precipitation per- colates into the porous tuffs and gathers in the gravel basins, from the lower parts of which large quantities of cold springs 38 MINERAL DEPOSITS issue. This upper zone of gathering and discharge may be 1,500 feet deep and may lie the same distance above the bottom of the canyons. In spite of the depth of the percolating zone, the waters are potable, pure, and cold. A part of the water sinks into the underlying bedrock of slate or granite, but the quantity is far less than in the more porous Tertiary strata and it finds its lowest level of discharge along the beds of the rivers. For the Sierra as a whole the Great Valley of California forms the ultimate level of discharge. In the whole western part of the range there are no thermal springs and very few strong ascend- ing springs, in spite of the prevalent fissility and jointing in the rocks. Hot springs are encountered only along the eastern slope of the range, a region which in the late Tertiary and Quaternary was the scene of great dislocations and volcanic activity. In the gold-quartz veins contained in the old rocks of the western slope much water is found in fissures to a depth of about 800 or 1,000 feet. Below this little water is met and many stopes and drifts are entirely dry, and this applies both to mines high up on the slopes, as at Nevada City and Grass Valley, and to the Mother Lode mines of the foot-hill region. Cripple Creek, Colorado, is another interesting example. Here we have a granitic plateau at an elevation of 9,000 feet above the sea; this plateau contains a volcanic plug about 2 miles in diameter which is largely filled with porous breccias and tuffs. The water fills the volcanic rocks as in a sponge inserted in a cup and the mining operations to a depth of 1,500 feet have tapped heavy flows. But even in this water-logged mass there are solid intrusive bodies, for instance at the Vindicator mine, at a depth of 1,000 feet, which are so dry that water must be sent down for drilling. The data thus far available have led Ransome to the conclusion that even at Cripple Creek the water is slowly diminishing in quantity at increasing depth. The big drainage tunnel now under way, which will tap the veins at a depth of 800 feet below the present lowest tunnels, will afford more information on this subject. The granite which surrounds this water-soaked plug contains very little water and at most places is practically dry, in spite of the great hydrostatic pressure. The ultimate level of possible discharge would be in the valley of the Arkansas, 2,500 feet lower and many miles distant, but it may be gravely doubted whether any water from the Cripple Creek mines ever finds its way through the granite mass to this THE FLOW OF UNDERGROUND WATER 39 level. 1 Van Hise, after stating (Metamorphism, p. 1065) that during a'certain time the Portland mine, at Cripple Creek, yielded water to the amount of between 300 and 900 gallons per minute, asks whether better evidence could be required for proving the existence of an extremely active circulation. The answer to this is that the water was simply stagnant, stored water filling an underground reservoir. In the copper mines of Butte, Montana, where the granitic rocks are greatly faulted by movements of late date, much water was encountered, extending in places down to 2,400 feet, or the bottom of the mines. No ascending springs are found at the surface, nor any hot springs, although a high range adjoins the mines on the east and conditions seem to be favorable for deep circulation. The water is probably almost stagnant, and Weed mentions the existence of large bodies of dry rock. 2 One such body on the 1,600-foot level, 1,200 feet in width, is absolutely dry. Leadville, Colorado, is another place where the faulting is extensive and of comparatively recent date. At 1,500 feet, the greatest depth attained, there is still much water, mainly along the faults. At Rossland, British Columbia, according to Bernard McDon- ald, 3 the mine waters increase greatly during the spring months. The water level is at 40 feet and the quantity increases to a depth of 200 to 350 feet. Below 350 feet a decrease begins, slowly at first but soon more rapid, until at 900 feet there is only a slight seepage and below 1,000 feet the mine is dry. 4 Weed states that in the copper mine at Ely, Vermont, an incline shaft was carried down for a length of 3,600 feet, attaining a vertical depth of 1,700 feet. There is no water here below a vertical depth of 600 feet. At Przibram, Bohemia, the workings are dry and dusty at a depth of 3,000 feet. In Cornwall, and in New Found- land mines have been worked underneath the sea, and sometimes close to the sea bottom, without irruptions of salt water. 5 'Lindgren and Ransome, Prof. Paper 54, U. S. Geol. Survey, 1906, pp. 233-251. Finch, J. W., op. cit., p. 204. 2 M. L. Fuller, Water-Supply Paper 160, U. S. Geol. Survey-, 1906, p. 65. 3 T. A. Rickard, Min. and Set. Press, June 27, 1908. 4 M. L. Fuller, op. cit., p. 65. 6 For other examples see J. F. Kemp, The ground waters, Trans., Am. Last. Min. Eng., vol. 45, 1914, pp. 3-25. 40 . MINERAL DEPOSITS One of the most convincing examples is that furnished by the deep copper mines of Michigan and fully set forth by A. C. Lane. 1 He shows that the surface waters are of the normal, potable type and that they descend in diminishing quantities only to a depth of about 1,000 or 1,500 feet below the surface. Below this depth moisture is scant, but where it appears it consists of drippings of strong calcium chloride brine which cannot in any way be ex- plained as being derived from the surface water. Many levels are absolutely dry and water must be sent down for drilling. This case is particularly convincing, for we have here many fea- tures in favor of a strong circulation: Moist climate, inclined position of beds, and great permeability. No certain figure can be assigned to the depth of the ground water; it may be shallow or the water may descend on strong fractures for several thousand feet. At any rate the quantity is limited, and the water is largely stagnant and is much more likely to decrease than to increase at depths below 1,000 feet. Depth of Water Level. In moist climates the water level is usually found within 50 feet of the surface, but in regions with less rainfall there is great diversity in the location of this upper limit of the zone of saturation. In the more arid regions the water is often met 300 or 400 feet below the surface. In the valley of Hachita, New Mexico, no water is found in the sands and gravels until a depth of 500 feet is reached; at the Abe Lin- coln gold-quartz mine, New Mexico, a little water began to come in 1,300 feet below the surface. In the rich deposits of Tintic, Utah, the water level in limestone lies 1,700 to 2,000 feet below the surface, but in mines in andesite and porphyry in the same district water may be found at much less depth. When water is being drained or pumped from a mine the water level is artificially depressed, in the form of a flat funnel. The pump in this case does not merely drain the bottom level, but receives water from higher levels farther away from the shaft. It is important to note this, for the water thus obtained from the bottom of a wet mine may not have the same composition as that originally belonging to this level. Total Amount of Free Water in Earth's Crust. Several es- timates have been made of the total amount of uncombined water contained in the upper crust. The older estimates by Delesse, 1 Mine waters, Trans., Lake Superior Min. Inst., vol. 12, 1908, pp. 154- 163. THE FLOW OF UNDERGROUND WATER 41 Dana, and Slichter were very high. Chamberlin and Salisbury 1 believed that the water in the earth would be equivalent to a layer 800 feet deep over its entire surface. Van Hise 2 reduced the estimate materially and concluded that it would be equivalent to a sheet of water 226 feet (69 meters) thick over the continental areas. Fuller 3 estimates, after a careful study of the problem, that the total water would be equivalent to a uniform sheet 96 feet thick over the entire surface of the earth. This estimate is probably more nearly correct than any of the others. 1 Geology, 1904, vol. 1, pp. 206-207. 2 A treatise on metamorphism, Mon. 47, U. S. Geol. Survey, 1904, pp. 128-129, 570-571. 1 Op. tit., p. 72. CHAPTER IV THE COMPOSITION OF UNDERGROUND WATERS INTRODUCTION Water is continually evaporated from sea and land. From the gathered clouds it is precipitated as pure rain water, which, by the aid of absorbed oxygen and carbon dioxide immediately begins the attack on disintegrating rocks. The rivers finally carry suspended particles and dissolved salts to the sea. Con- sidered more closely, the rivers are the products of the weak solutions of the immediate run-off and the stronger ground water solutions from the zone of discharge (Chapter III) which have been in longer contact with the rocks and leached them more thoroughly. This inconspicuous process of decomposition of rocks and solution of resulting salts is one of scarcely realized geologic importance. F. W. Clarke 1 states that the Mississippi annually carries to the sea about 108 metric tons of salts from each square mile of territory drained; the Colorado abstracts about 51 tons from the same unit area. A smaller part of the ground water sinks to enrich the static zone of stagnant waters, and ultimately becomes highly charged with salts. A still smaller part of the water is permanently withdrawn by entering hydrated compounds like kaolin. To complete the picture we must not overlook the ascending hot solutions which come from great depths and which in part at least are derived from rising magmas. As most mineral deposits have been formed by aqueous solutions the composition of the waters of rivers, lakes and seas becomes a study of importance. Even more important is the composition of the underground waters of wells and springs. In considering them from a chemical standpoint it will be best 1 F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 113. R. B. Dole and H. Stabler, Water-Supply Paper 234, U. S. Geol. Survey, 1909, p. 78. 42 COMPOSITION OF UNDERGROUND WATERS 43 to attempt no artificial distinction between thermal or cold mineral or non-mineral waters. The substances dissolved in the ground water depend upon the formations which it traverses. At the immediate surface organic life may influence the composition. In general, each formation yields its characteristic salts to the precolating waters Each natural water is a chemical system of balanced constituents of more or less dissociated electrolytes, of colloids, and of gases. CALCIUM CARBONATE WATERS IN IGNEOUS ROCKS Igneous rocks, of deep-seated origin, as well as crystalline schists contain only small amounts of soluble salts. The surface waters penetrating them are charged with more or less carbon dioxide, which, at ordinary temperatures, gradually decomposes the silicates, particularly the pyroxene, amphibole, biotite, and the calcium feldspars; the alkali feldspars are more slowly attacked. As a result the springs in such terranes will have a low salinity, rarely above 1,000 parts per million, and will contain principally calcium carbonate, with more or less of the corre- sponding magnesium salt; a smaller amount of sodium carbonat'e and much less of potassium carbonate are present. There will be little of the chlorine and sulphuric acid radicles. The silica is relatively high. Such calcium carbonate waters are character- istic not only of superficial springs, but also of the deeper cir- culation in crystalline terranes; in the latter case the waters may be warm, though usually they are cold. The spring-fed rivers in such terranes have a similar composition. Where magnesian rocks like basalt and serpentine abound, the underground waters are richer in magnesia than usual, and this substance may even equal the calcium. Waters of this calcium carbonate type are common and, when encountered as ascending springs or elsewhere, justify the presumption of surface origin. 1 1 Some years ago, in a report on the gold-quartz veins of Nevada City and Grass Valley, Cal. (Seventeenth Ann. Rept., U. S. Geol. Survey, pt. 2, 1896, p. 121), I presented an analysis of an ascending spring found in the Federal Loan mine which carried some arsenic and hydrogen sulphide. At that time I held the opinion that this water might possibly have bad some connection with the genesis of the vein, but it is now apparent that it is simply water of the general surface circulation which happened to find its way up on the vein and which dissolved certain constituents from it. The analysis is quoted on page 44. 44 MINERAL DEPOSITS It often happens that hot springs which are not characterized by an abundance of calcium carbonate are accompanied by numerous other springs of somewhat lower temperature. A comparison of analyses will usually show that in proportion to the lowering of the temperature the quantity of calcium carbonate increases; this indicates a cooling admixture of surface waters bearing calcium. COMPOSITION OF SALTS AND TOTAL SALINITY OF SURFACE WATERS IN CRYSTALLINE ROCKS A. B C D CO 3 so, Cl 31.91 9.07 4.03 47.14 6.67 4 18 57.80 3.10 1 30 38.46 15.35 2 81 s 50 Ca 14 53 22 67 13 90 13 24 Mg Mn Na K (Al,Fe) 2 3 Si0 2 2.93 10.80 2.72 0.51 23.50 6.17 I 5.32 7.85 2.30 0.10 5.50 0.40 1.70 13.40 4.33 12.86 3.76 9.19 100.00 100.00 100.00 100.00 Salinity, parts per million . . i 37 280 245 282 NOTES RELATING TO ABOVE ANALYSES A. Cache la Poudre River, Colorado, above North Fork. In schist and granite. Analysis by W. P. Headden. See F. W. Clarke, Geochemistry, Bull. 616, 1916, p. 65. B. Aztec Spring, 4 miles east of Santa Fe, New Mexico. In schist and granite. Cold. Analysis by F. W. Clarke, Geochemistry, p. 64. C. Cold spring in Federal Loan mine, in granite and schist, Nevada City, California. Approximate analysis by W. F. Hillebrand; contains a little manganese and trace of lead. Deposits calcium carbonate, limonite, and some arsenic. Sulphur probably due to reduction from small amount of H 2 S after bottling. Seventeenth Ann. Rept., U. S. Geol. Survey, pt. 2, 1896, p. 121. D. Cold water from 500-foot level, Geyser mine, Silver Cliff , Colorado . Analysis by W. F. Hillebrand. In "The mines of Custer County, Colorado," by S. F. Emmons. Seventeenth Ann. Rept., U. S. Geol. Survey, pt. 2, 1896, p. 461. Free and, semi-combined CO, 38.8 parts per million. COMPOSITION OF UNDERGROUND WATERS 45 Waters of the kind described above are generally poor in the rarer metals. A little arsenic is found in some cases and traces of barium, strontium, lithium, boron and phosphorus are some- times recorded. Where they traverse mineral deposits, metals contained in the deposits will of course be dissolved, as in the water from the Federal Loan mine, Nevada City, California. The springs may, under favorable conditions, form crusts of calcium carbonate and hydroxide of iron, but as a rule their powers of solution and deposition are weak. Where the rocks contain much pyrite, as often is the case in mining districts, the sulphates, especially calcium sulphate, rapidly increase in the waters. CALCIUM CARBONATE WATERS IN SEDIMENTARY ROCKS Waters of the type described above are not confined to igneous rocks. They are often found in circulation in glacial drift and also in sedimentary rocks sandstones, limestones, and dolo- mites. Such waters sometimes contain hydrogen sulphide and carbon dioxide. The derivation of the latter is not always easily explained. In some cases the gas may emanate from a deep-seated magma, but more commonly it is formed by decom- position of carbonates. An example of such water is furnished by the cold Cresson Spring in Pennsylvania, which issues from a shale member between sandstones in a 3,000-foot series of Coal Measures, containing practically no limestone. This water is pure, its salinity being only 442 parts per million, and of this 272 may be calculated as calcium carbonate, 76 as sulphates of sodium, magnesium, and calcium, and 11 as sodium chloride. According to a careful analysis by Genth this water contains traces of nickel, cobalt, iron, manganese, copper, strontium, barium, and fluorine, 0.17 part per million of the last-named element being present. Several analyses of similar well-known waters are quoted on page 46. 46 MINERAL DEPOSITS COMPOSITION OF SALTS AND TOTAL SALINITY OF CALCIUM CARBONATE WATERS IN SEDIMENTARY ROCKS A * C CO 3 . 40 02 41 47 '48 64 SO, : : .... Cl 21.73 0.64 3.93 1 27 6.30 5 40 PO 03 NO, 23 Ca 23 35 23 54 16 56 Me .. 5 82 2 56 7 64 Na 1.81 2 38 10 36 K 2 04 80 NH 4 03 Mn ... 17 Fe,O, \ 40 A1 2 3 SKX 0.58 4.01 \ 0.10 22 85 0.20 4 50 BO 2 64 100.00 100.00 100.00 Salinity, parts per million 563 199 222 HCO 3 . NOTES RELATING TO ABOVE ANALYSES A. Virginia Hot Springs, Virginia. Analysis by F. W. Clarke. "Geo- chemistry," p. 193. Temperature tepid. Issues from Paleozoic sediments. See also Bull. 32, U. S. Geol. Survey, 1886, p. 61. B. Hot Springs, Arkansas. Spring No. 16. Temperature 62 C. Issues from sharply compressed folds of Silurian sandstone and shale. CO 2 from bicarbonates 28.34 cc. per liter; nitrogen 8.39 cc. per liter; oxygen 2.49 cc. per liter; H 2 S none. Arsenic none; trace iodine and bromine. Analysis by J. K. Haywood. The Hot Springs of Arkansas, Senate Doc. 282, Fifty- seventh Congress, First Session, 1902, p. 94. Recalculated by F. W. Clarke, "Geochemistry," 1916, p. 195. C. Cold water from well of Missouri Lead and Zinc Company, Joplin, Missouri. Depth 1,387 feet. In Paleozoic limestone. Analysis by Cleve- land and Millar Laboratory. Water-Supply Paper 195, U. S. Geol. Survey, 1907, p. 137. Recalculated. These waters frequently form ascending springs. Bartlett Springs, Lake County, California, the water of which is exten- sively used in that State, probably belong to this class. The COMPOSITION OF UNDERGROUND WATERS 47 water contains 782 parts of salts per million, of which 493 may be calculated as calcium carbonate. It is rich in free carbon dioxide and is low in chlorine, sulphuric acid radicle, sodium, and potassium, but contains some iron, probably as carbonate, a little barium, phosphoruSj and about 63 parts of silica per million. 1 Carbonate waters are undoubtedly active in solution and deposition in the upper part of the crust, and especially in the formation of concentrations from weathering rocks. They may deposit calcareous sinters and effect concentrations of iron and manganese. Some lead and zinc deposits in limestone may also be genetically connected with them; their power of solution and concentration of rarer metals appears to be weak, unless they contain carbon dioxide and hydrogen sulphide. Such waters in Kansas, Missouri, and Kentucky have been found to contain zinc and probably also lead and copper. The salts are surely obtained from the rocks traversed. CHLORIDE WATERS IN SEDIMENTARY ROCKS Infiltration from Present Oceans. Wells and springs along the sea coasts usually contain a higher percentage of sodium chloride than farther inland; this may be caused either by infiltration of sea water into sediments or porous igneous rocks, or by winds carrying finely divided salt from the spray of the waves. Solution of Saline Deposits. Many past geologic periods in- cluded epochs of desiccation and desert climate when salt was precipitated from evaporating waters of closed basins. Surface waters encountering such sedimentary deposits easily dissolve the sodium chloride, and wells and springs rich in this salt are characteristic of many regions. Besides sodium these waters contain calcium and magnesium, and they are often rich in calcium chloride. They are poor in silica and potassium and rarely contain much calcium which can be combined with carbon dioxide. The presence of bromine is almost character- istic; traces of iodine and boron are often found. Barium and strontium are almost always present, the former sometimes in considerable amount. Free carbon dioxide and hydrogen 1 Winslow Anderson, Mineral springs, etc., of California, 1892, p. 94. G. A. Waring, Springs of California, Water-Supply Paper 338, U. S. Geol. Survey, 1915. 48 MINERAL DEPOSITS sulphide are sometimes found, the latter especially where there is an abundance of calcium sulphate. Waters of this general type are characteristic of certain Paleozoic beds in the eastern United Slates, as. for instance, the Silurian of New York and Michigan and certain parts of the Carboniferous in Michigan. In the western States the "Red Beds," generally of Permian or Triassic age, are sometimes rich in salt and gypsum, and this combination appears in the waters of these terranes. There are many similar springs and wells in Pennsylvania, and in fact all through the interior Paleozoic basin, from Arkansas to Canada. The Saratoga Springs of New York, issuing from Silurian limestones, probably belong to this class. Their tem- perature is about 50 F.; the total solids amount to about 11,000 parts per million, of which the larger part is sodium chloride. Barium is conspicuously present, in some analyses to a max- imum of about 34 parts per million, likewise bromine at about 1.20 parts per million. Small amounts of silica, iron, and lithium, and traces of boron, iodine, and fluorine are recorded. The origin of the CO? so abundant at Saratoga Springs is uncertain. J. F. Kemp believes it to be of magmatic derivation. Examples of such waters are given in the table of analyses on page 50. Certain of these waters are abnormally rich in calcium chlo- ride, that most easily soluble salt which remains as the last liquid residue in evaporating brines. Several instances of such waters have been interpreted as residual or connate brines, remaining in early isolated Paleozoic basins. 1 In the lower peninsula of Michigan brines are obtained from deep wells in the Carboniferous and Silurian. One of the springs in this region contains 12,000 parts per million in total solids, with 6,000 calculated as NaCl, 1,600 as MgCl 2 , and 4,100 as Ca C1 2 . The researches of A. C. Lane have shown that the scanty waters in the deep levels of the copper mines near Houghton have a similar composition, except that here calcium chloride prevails. These waters, which are found in amygdaloid lava flows and associated sedimentary rocks of the Upper Algonkian (Keweenawan), are perhaps to be regarded as residual oceanic waters, which, in their long contact with the rocks, have under- gone considerable changes. An analysis is given below (p. 50) This water contains no barium. 1 E. M. Shepard, Underground waters of Missouri, Water-Supply Paper 195, U, S. Geol. Survey, 1907, p. 81. COMPOSITION OF UNDERGROUND WATERS 49 In oil bearing districts salt waters are of very frequent occur- rence. They are rich in sodium chloride and often also contain the chlorides of calcium and magnesium, as well as more or less bicarbonates. They are always poor in sulphates, and this is perhaps due to their reduction by -the hydrocarbons. Such salt solutions have been variously interpreted as connate waters, as solutions of saline deposits and as of magmatic origin. 1 In the western States many similar waters occur in the Red Beds, but, as stated they are usually also rich in calcium sulphate. As an example may be cited the tepid Quelites Spring in New Mexico, 2 which ascends through Red Beds and contains about 2.6 per cent, of solids; one-half is calculated as sodium chloride and the larger part of the remainder as calcium sulphate. Bro- mine, boron, and barium are present. On the Pacific coast such waters are not common. Byron Hot Springs, California, may be cited as an example. The temperature spring is 76 C. The water contains about 13,000 parts of salts per million, of which over 10,000 parts are sodium chloride. A large portion of the remainder consists of calcium chloride. Small quantities of bromine, iodine, and barium are present. 3 The Triassic strata of the French Alps and the Pyrenees are rich in similar waters, many of which are warm. The mineral combination is a characteristic mingling of chlorides and sul- phates, and undoubtedly all of the constituents are derived from the sedimentary rocks mentioned. The Spring of Mey in Haute Savoie, with a temperature of 39.8 C., may be taken as a typical example. It contains both carbon dioxide and hydrogen sulphide and yields a total of 5,000 parts per million of dissolved salts, of which 1,753 parts are calculated as sodium chloride, 1,773 as sodium sulphate, and 957 as calcium sulphate. Some bromine and traces of iodine, phosphorus, and arsenic are present. 4 A celebrated group of these chloride springs are found in Germany on both sides of the Rhine. Among them are the 1 C. W. Washburne, Chlorides in oil field waters, Trans., Am. Inst. Min. Eng., vol. 45, 1915, pp. 687-693. G. S. Rogers, Chemical relations of the oil field waters in the San Joaquin Valley, California, Bull. 653, U. S. Geol. Survey, 1917. 2 F. A. Jones, New Mexico mines and minerals, 1904, p. 309. 3 Winslow Anderson, Mineral springs and health resorts of California, 1892, p. 106. 4 Jacquot et Willm, Les eaux minerales de la France, Paris, 1894, p. 243. 50 MINERAL DEPOSITS waters of Soden, Homburg, Wiesbaden, Kreutznach, Kissingen, Nauheim. Most of them issue from or ascend through salt- bearing beds of Devonian, Permian, or Triassic age, and their composition is similar. The springs of Kreutznach are especially rich in calcium chloride. Some of the springs cited are hot, others cold; some are rich in carbon dioxide. In regard to Kreutznach and Wiesbaden there is room for doubt, for the former springs stand in intimate relation to eruptive rocks, while the latter issue from a gneiss and are by some authors con- sidered of juvenile origin. The majority of them, at any rate, have certainly derived their salts from sedimentary beds. The chloride waters, described above, are capable of dissolving and depositing many metallic substances and have strong dehydrating power. Their relation to mineral deposits will be mentioned later. COMPOSITION OF SALTS AND TOTAL SALINITY OF CHLORIDE WATERS (Cited from Clarke's Geochemistry, 1916, pp. 182-186) A B c D E F Cl 55.83 Br 0.04 1 0.03 SO 4 3.12 CO 2 63 58.79 trace 0.94 61 42.00 1.13 0.02 0.08 18 59 62.31 0.53 0.01 0.03 27 63.55 0.01 01 56.58 0.04 trace 0.78 3 13 B.O, 01 Na 33.09 K 0.27 Li NH 30.38 3.76 27.62 0.78 0.08 18.35 1.55 0.04 0.23 5.63 32.60 1.16 0.04 0.07 Ca 3.72 Ba 4.90 6.03 09 13.86 30.78 4.05 01 Sr . . 12 Mg 1.13 A1 2 O 3 0.40 0.02 3.41 2.53 0.02 0.01 0.61 Fe 2 O 3 06 03 Fe . 25 04 SiO 2 ! 0.08 0.20 0.14 0.02 0.01 0.76 100.00 100.00 100.00 100.00 100.00 100.00 Salinity, parts 10,589 per million. 23,309 12,022 309,175 212,300 8,241 COMPOSITION OF UNDERGROUND WATERS 51 NOTES RELATING TO ABOVE ANALYSES A. Cincinnati artesian well, Cincinnati, Ohio. Analysis by E. S. Wayne, cited by A. C. Peale, Bull. 32, U. S. Geol. Survey, 1886, p. 133. This water contains considerable quantities of free H 2 S and CO 2 . B. Utah Hot Springs, 8 miles north of Ogden, Utah. Temperature 55 C. Analysis by F. W. Clarke, Bull. 9, U. S. Geol. Survey, 1884, p. 30. C. Congress Spring, Saratoga, New York. Analysis by C. F. Chandler, cited by A. C. Peale, in Bull. 32, U. S. Geol. Survey, 1886, pp. 38, 39. Traces of F, P, B, Sr, and Al. Contains much free C0 2 . D. Brine from well 2,667 feet deep at Conneautsville, Pennsylvania. Analysis by A. E. Robinson and C. F. Mabery, Jour., Am. Chem. Soc., vol. 18, 1896, p. 915. A little H 2 S is present. E. Water from the deep levels of the Quincy mine, Hancock, Michigan. Analysis by George Steiger. F. The Kochbrunnen, Wiesbaden, Germany. Analysis by C. R. Fresen- ius. This water also contains traces of I, P, and As. CHLORIDE WATERS IN IGNEOUS ROCKS Waters rich in chlorine are sometimes found as ascending springs in igneous rocks, but almost always close to regions of comparatively recent volcanic activity. Their composition is somewhat different from the brines resulting from the dissolving of salts from sedimentary beds. Bromine is seldom present except in mere traces, wnile boron appears in considerable amounts. Such tepid salt waters arise, for instance, in the volcanic region around Clifton, Arizona. The Paleozoic rocks of this region are not known to contain either salt or gypsum. Another case is the Glen wood Hot Springs in western Colorado; the springs at this place issue from limestone, but the structural relations show that the basal granite underlies this limestone at slight depths. The temperature is 49.5 C; the water contains a large amount of sodium chloride and relatively small amounts of carbonates and sulphates. Hydrogen sulphide and free carbon dioxide are present. Still another case is Steamboat Springs, Nevada, which issue from granodiorite near the eastern base of the Sierra Nevada in a region of Tertiary volcanism. Many of these springs are rich in carbon dioxide and hydro- gen sulphide; they often contain many of the rarer elements, as shown in the analyses quoted below, and they usually appear in regions rich in ore deposits. Doubt as to the derivation of the salt may exist in many cases, as, for instance, in the springs of Kreutznach, Germany, which issue from a porphyry said by 52 MINERAL DEPOSITS Laspeyres to contain 0.001 per cent, sodium chloride. 1 Delkes- kamp, 2 on the other hand, holds that the salt is derived from sedimentary deposits. Another notable instance of chloride springs of this class is mentioned by Daubree 3 from the provinces of Antioquia and Cauca in Columbia, where they issue in great abundance from granite, crystalline schist, and late volcanic rocks. Great difficulties arise in attempting to trace the origin of the sodium chloride in springs of this class to surrounding rocks, even admitting that granite and other crystalline rocks may contain traces of this salt. Sinters of calcium carbonate and silica are often deposited at the orifices of these springs. COMPOSITION OF SALTS AND TOTAL SALINITY OF SODIUM CHLORIDE AND SILICA WATERS (After Clarke's Geochemistry, 1916, pp. 186 and 196) A B c D E Cl 35.00 36.61 31.64 13.52 37.52 Br 25 so, s - CO 4.58 0.22 5 08 1.84 15 1.30 8 78 9.01 0.32 10 16 4.96 PO 4 0.03 0.08 AsO 4 24 B 4 O 7 8 88 2 24 1 19 Na K Li NH 4 30.35 3.79 0.27 21.44 4.45 0.22 0.02 26.42 1.93 0.40 trace 19.71 1.88 0.28 24.22 0.36 Ca 25 39 11 2 59 Me 01 08 04 08 19 Fe trace trace trace As Sb 0.10 02 A1 2 O 3 0.01 0.76 12 0.35 SiO, 11 41 31 72 27 58 45 04 29 81 100.00 100.00 100.00 100.00 100.00 Salinity, parts per million . 2,850 1,830 1,388 1,131 2,735 1 Zeitschr. Deutsch. geol. Gesell., vol. 19, p. 854, and vol. 20, p. 155. 2 Verhandl'. Polytech. Gesell. (Berlin), 1903, II, p. 161. 3 Les eaux souterrainos. etc., II, 2, p. 106. COMPOSITION OF UNDERGROUND WATERS 53 A. Steamboat Springs, Nev. Analysis by W. H. Melville, given by G. F. Becker in Mon. 13, U. S. Geol. Survey, 1888, p. 349. Bicarbonate reduced to normal salts. Temperature 85 C. Contains free carbon dioxide and hydrogen sulphide. Traces of iron and quicksilver; deposits cinnabar and stibnite. B. Coral Spring, Norris Basin, Yellowstone National Park. Analysis by F. A. Gooch and J. E. Whitfield, Bull. 47, U. S. Geol. Survey, 1888. Tem- perature 73 C. H 2 S none. Free CO 2 42.5 parts per million. C. Old Faithful Geyser. Same locality and analysts. Temperature 84-88 C. H 2 S, 0.2 part per million. D. Great Geyser, Iceland. Analysis by F. Sandberger. E. Water of the pink terrace, Roturoa geyser. Analysis by W. Skey. Closely related to this group are the predominant springs in the great geyser regions of Yellowstone National Park, New Zealand, and Iceland. They are essentially sodium chloride waters with large amounts of silica, believed to exist in part as sodium silicate, a large quantity of free carbon dioxide, and a little hydrogen sulphide. Large amounts of boron, usually calculated as sodium borate, are often present, and also fre- quently arsenic. Bromine is rarely recorded in quantities approaching those in the brines from sedimentary formations. The waters are always hot and usually ascend through volcanic rocks, mostly rhyolite; from these the silica is supposed to be derived, but no such explanation seems sufficient to account for the predominating salt, sodium chloride, or for the boron. In the Yellowstone Park a number of the springs issuing near limestone bear evidence of their passage through this rock in increased quantities of calcium and magnesium. Others are rich in sulphate of sodium and other sulphates, but these springs give an acid reaction and the sulphates are in all probability due to the oxidation of hydrogen sulphide and the replacement of silica in sodium silicate by sulphuric acid. SULPHATE WATERS IN SEDIMENTARY ROCKS The waters which traverse sedimentary rocks are often rich in salts, particularly in sulphates. The gypsum waters have been mentioned and are connected with the sodium chloride waters in a manner corresponding to the association of gypsum and rock salt. By interaction of calcium sulphate and magnesium carbonate, the sulphate of magnesium may be formed, or it may be derived from the decomposition of a pyritic dolomite. 54 MINERAL DEPOSITS Sodium sulphate waters are almost characteristic of certain for- mations in the western Cretaceous, for instance; these formations consist mainly of sandstones and carbonaceous shales, the lat- ter often pyritiferous and the whole series mainly a product of near-shore deposition. The oxidation of the pyrite furnishes solutions containing free sulphuric acid, and by reaction between this and various other substances sulphates of calcium, magne- sium, and sodium will be formed. In land deposits contained in many series of sedimentary rocks sodium carbonate and sodium sulphate are formed by several well-established reactions, and percolating waters will easily abstract these salts. The inter- action of calcium sulphate and sodium carbonate results in sodium sulphate and precipitation of calcium carbonate. So- dium sulphate in the presence of free carbon dioxide will dissolve calcium carbonate, forming sodium bicarbonate and a'precipitate of gypsum. 1 Reactions in soils between sodium chloride and calcium sulphate result, according to Cameron, in calcium chloride and sodium sulphate, and similar reactions take place between sodium chloride and calcium carbonate. Sodium sulphate waters are, as stated, common in the western Cretaceous, especially in the shale formations. The lowest member of this series, the Dakota sandstone, is particularly noted as a water-carrying for- mation. The water, which is under artesian pressure, penetrates this formation for several hundred miles underground from its outcrop and in places contains so much sodium sulphate as to be unfit for irrigation purposes. There is no evidence that this water has formed mineral deposits in the sandstone. A well 1,400 feet deep, in Dakota sandstone at Pueblo, Colo- rado, contains, according to Darton, 2 1,337 parts per million of total solids, of which about one-half is calculated as sodium sul- phate and one-fourth as calcium sulphate. Very little silica and chlorine, but a little iron and carbon dioxide are present. This analysis appears to be typical. In many waters in sedimentary formations chlorides and earthy carbonates appear mixed with sulphates. Waters from artesian wells at Roswell, New Mexico, 1 F. W. Clarke, Geochemistry, 1916, p. 241. E. W. Hilgard, Am. Jour. Sci., 4th ser., vol. 2, 1896, p. 100. Cameron and Bell, Bull. 33, Bureau of Soils, 1906. F. K. Cameron, Butt. 17, Bureau of Soils. 2 N. H. Darton, Prof. Paper 32, U. S. Geol. Survey, 1905, p. 355. COMPOSITION OF UNDERGROUND WATERS 55 about 400 feet deep, derived from Permian limestones, 1 have a temperature of 64-70 F. and contain from 600 to 1,200 parts per million of solid salts, of which 300 to 576 are calcium and magnesium sulphates and the remainder carbonates and chloride of sodium. In regions of dislocations such waters may be hot and then the ordinarily low percentage of silica may increase con- siderably. The Arrowhead Spring 2 of San Bernardino Valley in southern California, issuing from Tertiary sediments, has a temperature of 184 F. and contains 1,086 parts per million of solids, of which 735 are calculated as sodium sulphate, 69 as potassium sulphate, 23 as calcium sulphate, 3 as magnesium sul- phate, 141 as sodium chloride, and 23 as calcium carbonate; 85 are present as silica. In well waters of the same valley the solids range from 191 to 260 parts per million and the relation is CaC0 3 >MgC0 3 = NaSO 4 > NaC0 3 > >NaCl. Silica amounts to 24 to 32 parts per million. PERCENTAGE COMPOSITION AND SALINITY OF SULPHATE WATERS DERIVED FROM SEDIMENTARY FORMATIONS A B cr D H SO (free) 9 37 Cl S0 4 CO 0.48 66.28 60 11.10 59.68 1 67 76.57 0.32 68.21 Na K o& Fe" 30.46 1.08 0.67 0.41 13.89 0.49 2.91 10.19 1.19 5.82 3.39 4 28 0.22 0.11 0.38 1.11 1.19 Al 7.36 11.08 SiO 2 0.02 0.07 1.39 7.11 100.00 100.00 100.00 99.10 Salinity, parts per million . . 74,733 15,682 3,303 464 A. Abilena Well, Abilene, Kansas, 130 feet deep. Analysis by E. H. S- Bailey, Geol. Survey, Kansas, vol. 7, 1902, p. 166. In Permian strata- From Clarke's Geochemistry, 1916, p. 187. 1 C. A. Fisher, Report on the Roswell artesian area, Water-Supply Paper 158, U. S. Geol. Survey, 1906. 2 W. C.^Mendenhall, Hydrology of San Bernardino Valley, Water-Supply Paper 142, U. S. Geol. Survey, 1905. 56 MINERAL DEPOSITS B. King's Mineral Spring near Dallas, Indiana. Twenty-sixth Annual Report, Indiana Dept! Geol., 1901, p. 32. Traces of Al, Fe, Ba, Sr, Li, Mn, Ni, Zn, Br, PC>4 and B 4 O7. Geological horizon Paleozoic shale. C. Alum Well, Versailles, Missouri. Analysis by P. Schweitzer, Geol. Surv. Missouri, vol. 3, 1892, p. 131. In Pennsylvanian shale. D. Rockbridge Alum Springs, Virginia. Analysis by M. B. Hardin. Cited in Clarke's Geochemistry, p. 98. From pyritic shale. Contains also 0.69 Mn, 0.01 Li, 0.05 Co, 0.07 Ni, 0.08 Zn, and traces of Cu, HNO 3 , and P0 4 . Waters percolating through oxidizing pyritic shales sometimes contain large amounts of the sulphates of aluminum and ferrous iron; evidently this happens only when comparatively large amounts of sulphuric acid, which is capable of attacking alumin- ous silicates, are set free. Such waters are not uncommon in the eastern and central States and usually contain small amounts of rarer metals; traces of nickel, zinc, and arsenic are common. The sulphate water, especially those rich in iron and aluminum, are of great importance in the genesis of deposits in the oxidizing zone, and the latter often form, at their orifices, large quantities of ocherous deposits. Many waters of this kind are known from Virginia, issuing from pyritic shales, and Peale 1 quotes some interesting and reliable analyses. A water from Alleghany Springs in Mont- gomery. County, analyzed by Genth, contained 3,129 parts per million of solids, of which the principal constituents were calcu- lated as 1,955 parts CaSO 4 , 255 parts MgSO 4 , and 61 parts CaC0 3 . Small quantities of strontium, barium, fluorine, and silica and traces of zinc, lead, copper, and cobalt are noted. Some free carbon dioxide and a trace of hydrogen sulphide are present. The Jordan Alum Springs in Rockbridge County, Virginia, of which several analyses by J. W. Mallet are recorded, contain from 306 to 935 parts per million of solid salts, of which the larger amount consists of aluminum sulphate, 35 to 85 parts of ferric sulphate, and from 8 to 17 parts of manganese sulphate. Small quantities of copper, zinc, cadmium, nickel, and cobalt are determined, also a trace of fluorine. One of the waters contained 102 parts of copper and 9 parts of zinc. In the Rockbridge Alum Springs, in the same State, small quantities of copper, nickel, cobalt, zinc, and a trace of lead were determined. 1 A. C. Peale, Lists and analyses of the mineral springs of the Unite d States, Bull. 32, U. S. Geol. Survey, 1886, pp. 58-65. COMPOSITION OF UNDERGROUND WATERS 57 Free sulphuric acid is present in the Bedford Alum Spring to the amount of 70 parts per million, according to M. B. Hardin. The total solids are 1,207 parts, practically all sulphates, and about one-third consists of ferric sulphate. Small quantities, about 0.8 part per million of each, of nickel, cobalt, copper, and zinc were determined. Springs of similar composition are found in Pennsylvania and other eastern States. All these acid springs are poor in silica and contain very little chlorine. ACID SULPHATE WATERS IN IGNEOUS ROCKS Sulphate springs in connection with igneous rocks and vol- canism appear mainly as products of the oxidation of ascending waters of alkaline reaction, containing free hydrogen sulphide, but there is evidence that in regions of volcanic activity such oxidation takes place on a large scale and that these acid waters are of high importance in effecting rock alteration, particularly by attacking aluminum silicate and developing alunite. By similar reactions free hydrochloric acid may be generated, for instance by the decomposition of chlorides by free sulphuric acid. As a consequence it is common to find such waters near the orifices of hot springs, as well as at volcanoes. The develop- ment of free acid of course displaces the equilibrium and the oxidized water may differ greatly from its parent liquid; thus it happens that a single ascending hot spring may yield a whole series of derivatives of varying temperature and composition by mingling with other waters and by oxidation. A number of analyses of such waters are quoted in Clarke's Geochemistry. Some of them, especially from pools or lakes near volcanoes, are remarkably rich in hydrochloric acid. The peculiar water from the Yellowstone National Park known as the Devil's Inkpot contains, besides free acids, a large amount of sulphate of ammonia. The water from Roturoa, New Zealand, the analysis of which is quoted below, is a more characteristic product of the oxidation of normal thermal waters. The geysers of Sonoma County, California, of which there is a good series of analyses by Dr. Winslow Anderson, 1 form a most remarkable illustration of the oxidation of hot waters. There are at this place a great number of springs of varying temperature 1 Winslow Anderson, Mineral springs, etc., of California, 1892, pp. 136-154 G. A. Waring, Springs of California, Water-Supply Paper 338, U. S. Geol. Survey, 1915, p. 109. 58 MINERAL DEPOSITS and composition, all of them heavily charged with hydrogen sul- phide. The primary water at a temperature of 110 F. appears to contain chiefly carbonate of magnesium with some of calcium. The total solids amount to about 568 parts per million, most of which consist of the above-mentioned carbonates; there are 92 parts of silica per million. This water is probably of mixed origin; the carbonates are clearly derived from the serpentinoid rocks of the vicinity, but the hydrogen sulphide is most likely of magmatic origin. Free sulphuric acid is generated by oxidation and gives rise to a long series of peculiar sulphate waters, most of them rich in dissolved solids and of high temperature. An analysis of one of these shows 3,262 parts per million of total salts and acids, among which the sulphates of magnesium, sodium, and aluminum prevail. There are 544 parts per million of free sulphuric acid and 20 parts of free hydrochloric acid. Finally, acid water may result directly from the oxidation of deposits of pyrite or of sulphur. A water of the latter type is described by W. T. Lee from Beaver County, Utah. PERCENTAGE COMPOSITION OF SALTS AND TOTAL SALINITY OF ACID WATERS A B C HC1 free . . 0.18 5 60 H 2 SO 4 free HBO 1.29 2 73 59.11 ,46.39 Cl 08 SO 67 66 20 21 32 63 Na 73 8 35 \ K 24 32 \ 1.48 Li 01 NH 22 85 Ca 1 18 47 1 63 Mg 36 22 2 50 Fe" ' trace 5.76 Fe'" 0.33 8.25 Al 10 trace SiO 2 67 5 39 1 28 100.00 100.00 100.00 Salts and acids, parts per million 3,365 1,862 9,716 Includes some alumina. COMPOSITION OF UNDERGROUND WATERS 59 A. Devil's Inkpot, Yellowstone National Park. Analysis by F. A. Gooch and J. E. Whitfield, Bull. 47, U. S. Geol. Survey, 1888, p. 80. Contains also 65 parts of free CO 2 and 5 parts of H 2 S per million. Cited in Clarke's Geochemistry, 1916, p. 199. B. Cameron's Bath, Roturoa geyser district, New Zealand. Analysis by W. Skey, Trans., New Zealand Institute, vol. 10, 1877, p. 423. Contains 6 parts per million of H 2 S. Cited in Clarke's Geochemistry, p. 200. C. Water at Sulphur mine of Cove Creek, Beaver Valley, Utah. Analy- sis by W. M. Barr, Water-Supply Paper 217, U. S. Geol. Survey, 1908, p. 20. Contains also much free H 2 S. MINE WATERS OF SULPHATE TYPE Mine waters consist as a rule of the normal surface waters of the rock containing the ore deposit, modified by the salts re- sulting from the decomposition of the minerals of the deposit. In deposits free from sulphides, such as the copper and iron mines of Lake Superior, there is little difference between the mine waters of the upper levels and the normal surface waters of low salinity; both are comparatively high in silica and calcium carbonate. Where much pyrite or marcasite is present, as in coal mines and in most metal mines, the surface waters will con- tain sulphates of ferrous iron and aluminum and frequently also of the rarer metals. When these waters mingle with normal surface waters rich in calcium carbonate the iron and alumina may be precipitated as hydroxides and calcium sulphate remains in solution. Calcium sulphate waters often spread over a considerable area surrounding pyritic deposits. The mine waters will be discussed in the chapter on oxidation and secondary sulphides. SODIUM CARBONATE WATERS IN SEDIMENTARY ROCKS Waters containing sodium carbonate in large amounts are not common in sedimentary rocks, but here and there wells or springs of this character are encountered; they are usually cold and often contain some free carbon dioxide and hydrogen sulphide. The alkaline carbonate is probably, as suggested above, derived from a reaction between sodium sulphate and calcium carbonate or between sodium chloride and calcium carbonate. Waters of this kind occur at a few places in the eastern and central States. G. L. Gumming has described such waters from some artesian wells in Silurian limestone on the Island of Montreal. 1 They contain from 500 to 700 parts per million of solids, chiefly sodium 1 Mem. 72, Geol. Survey Canada, 1915. 60 MINERAL DEPOSITS carbonate with the remainder calculated as calcium chloride and sodium sulphate. Gumming shows that the waters are of com- plicated origin but believes that the sodium carbonate solutions are derived from dikes and intrusions in the limestone. A good instance is furnished by some Missouri waters in Carboniferous limestone, one of which is quoted under E in the following table. Similar waters are those of the wells at La Junta, Denver, and Greeley, Colorado. The artesian wells at Denver, about 1,200 feet deep, are in the Arapahoe Eocene, while the Greeley well, of the same depth, is sunk in Laramie sandstone. The maxi- mum of total solids is about 1,530, divided between sodium carbonate and sodium chloride. Some free CO 2 is present. Many artesian waters in New South Wales are rich in sodium car- bonate. Sedimentary beds containing volcanic tuffs often yield sodium carbonate waters. SODIUM CARBONATE WATERS IN IGNEOUS ROCKS Ascending sodium carbonate waters are most characteristic of regions of subsiding or expiring volcanism. During surface eruptions alkaline chlorides and carbonates always appear as sublimates and waters traversing tuffs, breccias and lava flows may dissolve these salts together with other volcanic exhalations such as borates. The characteristic sodium carbonate waters are, however, of deep-seated origin and usually break in through the older igneous or metamorphic rock underlying the lavas in regions where the active volcanism has ceased, and the pre- vailing opinion is that these waters with their charge are in whole or in part of magmatic origin. They rarely contain much calcium and they are poor in silica, but are usually heavily charged with carbon dioxide and sometimes hydrogen sulphide. They almost always contain many rarer substances such as boron, fluorine, iodine, arsenic and various metals. Such waters would attack silicates with great energy and it is suggested that their strong percentage of sodium may have been leached from walls of the fissures, during the conversion of sodium silicates to potassium silicates as in the process of sericitization, so common in mineral veins. It is certain that these waters are of the utmost importance in the genesis of orebearing veins. An excellent instance of a province of such waters is furnished by the volcanic district of central France. 1 An analy- ^acquot and Willm, Les eaux minerales de la France, Paris, 1894. COMPOSITION OF UNDERGROUND WATERS 61 sis of the celebrated Vichy Springs is given in the table on this page. Sodium preponderates as bicarbonate, but smaller quantities of sodium chloride and sulphate are also present. The whole region of the Central Plateau is rich in carbon dioxide, occurring both in springs and as exhalations (for instance, at the Pontgibaud lead-silver mines). The magmatic source of the gas is rarely questioned, whatever opinion may be held about the origin of the water (Fig. 4). In the- volcanic regions of Taunus and Vogelsgebirge on the Rhine in Germany are the springs of Ems and Fachingen. The springs of Ems issue with a temperature of 46 C. and contain about 2,870 parts per million of solids, of which about one-half may be calculated at sodium carbonate and a large part of the remainder as sodium chloride. PERCENTAGE COMPOSITION OF SALTS AND TOTAL SALINITY OF SODIUM CARBONATE WATERS (Cited from Clarke's Geochemistry, pp. 191, 193 and 197) Cl 6 17 8 85 11 52 13 57 6 63 4 01 F 0.19 0.03 so 4 s 3.75 5.77 31.19 0.32 6.21 06 4.26 CO, 45 57 41 91 19 15 22 38 44 yg 47 45 PO 4 AsO 4 1.52 0.04 0.01 0.01 BA Na 35 27 0.16 38 08 32 49 27.98 33 97 ;;;;;;;; 41 07 40 09 K 2.88 1.20 1 35 48 38 NH 4 05 Li 0.12 Ca 2.29 87 2 23 41 30 27 Sr.. 04 05 01 Mg Mn 1.11 0.41 0.65 01 0.11 0.12 0.15 Fe 02 14 Fe,O,. 04 06 A1 2 O 3 . . 02 20 SiO 1 32 2 30 1 34 73 85 3 05 100.00 100.00 100.00 100.00 100.00 100.00 Salinity, parts per million 5,249 2,614 5,431 5,096 2,069 1,668 62 MINERAL DEPOSITS NOTES RELATING TO ABOVE ANALYSES A. The Grand-Grille spring, Vichy, France. Analysis by J. Bouquet. Small quantity of fluorine present. Temperature 44 C. Issues from Tertiary beds. B. Ojo Caliente spring, near Taos, New Mexico. Analysis by W. F. Hillebrand. Trace of barium and arsenic. In lake beds and gneiss. C. The Sprudel, Carlsbad, Bohemia. Analysis by F. Ragzsky. Contains 0.76 gram free and half-combined CO2 per kilogram. Traces of Br, I, Li, B, Rb, and Cs. Temperature 72 C. In granite. D. Hot water from the Hermann shaft, Sulphur Bank, California. Analysis by W. H. Melville. A little H 2 S and a considerable amount of CO 2 present. Temperature 80 C. In basalt and sandstone. E. McClelland well, Cass County, Missouri, 45 feet deep, in Carboniferous limestone. Analysis by P. Schweitzer. Contains H2S. F. Artesian water, La Junta, Colorado. Well 386 feet deep. Analysis by W. F. Hillebrand. In Cretaceous beds. At the .foot of the Erzgebirge, in the Tertiary volcanic region of northern Bohemia, issue a series of hot springs, extending from Teplitz to Carlsbad and Eger. Most of these belong to the class of sodium carbonate waters with free carbon dioxide. They contain an abundance of salts, and in the Teplitz and Bilin springs sodium carbonate predominates. In the Carls- bad (C in table of analyses) and Marienbad spring the sul- phuric acid radicle is prominent and must largely exist in so- dium sulphate. The Carlsbad springs contain fluorine and barium with traces of many rarer metals which are mentioned on page 97. IntheCordilleran Ranges in North America and South America sodium carbonate waters are abundant and always closely con- nected with areas of Tertiary volcanic activity. In New Mexico the Ojo Caliente (B in table of analyses), Fay wood, and Las Vegas springs may be mentioned; in Colorado the Idaho Springs, Middle Park Springs, Poncha Springs, and the water in the Geyser mine at Silver Cliff; in Idaho the Boise Hot Springs. In California sodium carbonate waters are especially abundant and characteristic; they follow the Coast Range from San Diego to Mendocino County and appear to stand in some causal connection with the late Teritary or Quat- ernary eruptions of basalt. Some of the waters are clearly admixed with magnesium from the serpentinoid rocks which they have traversed, but in general the type is perfectly distinct. The following data are taken from the U. S. Geological Survey, Bulletin 32, by A. C. Peale. COMPOSITION OF UNDERGROUND WATERS 63 Source of water Salinity, parts per million Composition and quantity of principal salts San Juan Capistrano (T. 50 C.) . I Skaggs Springs (T. 54 C.) | Paso Robles Springs (T. 42 C.) . ! New Almaden Vichy (T. 17 C.). NapaSoda (T. 17 C.) ! Pacific Congress (T. 10 C.) Ukiah Vichy (T. 34 C.) . . . 290 HNaCO 3 > NaCl >SiO 2 111 105 70 2,556 HNaCO 3 > BO 2 >SiO 2 2,083 176 151 1,581 HNaCO 3 > NaCl >Na 2 SO 4 850 469 136 7,361 ! HNaCO 3 > CaSO 4 >CaCO 3 3,400 680 544 1,156 HNaCO 3 >MgCO 3 >NaCl 561 187 85 5,678 HNaCO 3 > NaCl >CaCO 3 2,091 1,923 289 4,624 HNaCO 3 > NaCl >MgCO 3 3,369 459 374 An interesting type of these waters is represented by the hot spring of Sulphur Bank (D in table of analyses), which contains boron and is depositing cinnabar. On the whole these waters are rich in unusual constituents and have great solvent powers. SODIUM SULPHIDE WATERS It is believed that in some of the springs already referred to for example, Steamboat Springs, Nevada sodium sulphide or other sulphur salts of sodium are present. In the Pyrenees of France and Spain is found a group of Springs in which sodium sulphide is constantly present. These springs have a high tem- perature and a low salinity, containing from 250 to 350 parts per million of salts; they usually issue in crystalline schists or on the contact of the schists with Paleozoic strata. A charac- teristic spring mentioned among others by Jacquot and Willm 1 contains total CO 2 52, S (in sulphides) 31, Na 97, CO 3 26, Cl 55, andJ3iC>2j)3 parts per million. Some organic matter is present and strong traces of boron, arsenic, copper, etc., are mentioned. There appears to be considerable difficulty in the explanation of this combination on the hypothesis of leaching from the sur- rounding country rock. Where contaminated by surface water 1 I*ea eaux minerales de la France, Paris, 1894. 64 MINERAL DEPOSITS or where locally issuing through Triassic strata they beocme calcic. By oxidation they acquire hyposulphites. SUMMARY In sedimentary formations; beyond the influences of igneous activity, the waters are of many differing types. Some con- tain mainly calcium carbonate; others are of the chloride type, with sodium or calcium as the prevalent base; still others, a very abundant class, are rich in calcium or sodium sulphates; a rarer type is that of the sodium carbonate waters. Naturally many waters show a mingling of these types. Most of these waters are cold; many are tepid; few of them are hot. Whether warm or cold, both hydrogen sulphide and carbon dioxide may be present. In older igneous rocks where the effects of volcanism have sub- sided the types vary less widely. The ordinary surface waters are always unless some disturbing influence interferes of the calcium carbonate type, often with sodium chloride, ferrous and magnesium carbonate, and considerable silica, but low salinity. These waters sometimes, but not often, appear as tepid ascending springs. If the rocks contain iron disulphide the waters may locally contain free sulphuric acid and the sulphates of calcium, aluminum and iron. The remaining classes of water in igneous rocks are ascending and confined to regions of recent or Tertiary volcanic activity. They are tepid to hot, though cold waters are also known. They easily fall into two classes: (1) the sodium chloride waters, of which the siliceous "geyser waters" form a sub-class; (2) the sodium carbonate waters, which are generally rich in free carbon dioxide. Transitions between the two classes are plentiful, and the latter class may in rarer cases also contain notable amounts of sodium sulphate; of this class the Carlsbad Springs form a prominent example. INTERPRETATION OF WATER ANALYSES Analyses of waters are usually stated in parts per million of radicles and metals. From this form a calculation will be neces- sary to ascertain whether the water is alkaline, neutral, or acid. Stabler 1 has suggested that for this purpose the quantities deter- 1 Herman Stabler, The mineral analysis of water for industrial purposes and its interpretation by the engineer: Eng. News, vol. 60, 1908, p. 356. Also, chapter on the industrial application of water analyses in Water- Supply Paper 274, U. S. Geol. Survey, 1911, pp. 165-181. COMPOSITION OF UNDERGROUND WATERS 65 mined may be multiplied by the reciprocals of the equivalents. The products are called the reacting values. If the water is neutral the reacting values of acids and basic radicles should balance. Palmer, 1 in his method of geological interpretation of water analyses, finds it convenient to express the reacting values in percentages, thus eliminating the factor of concentra- tion. Palmer's classification emphasizes the fact that a solution in which strong acids are exactly balanced with strong bases is relatively inert, whereas one in which either group exceeds the other is relatively active. It is of special use in showing the relationship and the nature of chemical action of different waters. Alkalinity and Salinity are the fundamental properties. Salinity is measured by the strong acid radicles (SO 4, Cl). If the basic radicles are partly or wholly alkaline metals their pro- portion of the salinity is said to be primary. The remaining salin- ity due to radicles Ca, Mg, Fe is called secondary. If the acid radicles are in excess, tertiary salinity or acidity results. The measure of primary alkalinity is the excess of alkaline metal radicles over the strong acids; the weak-acid radicles C0 3 and HCOs which balance any excess of the alkaline earth metals over the stronger acids produce secondary alkalinity. 2 In spite of an objectionable terminology Palmer's method furnishes a convenient basis for comparative study but as a classification of natural waters it is unwieldy and uncertain. It is not always a safe guide to the geological history of the water. The constants used in converting grains per gallon to parts per million and vice versa are as follows: 1 grain per U. S. gallon = 17.138 parts per million 1 grain per Imperial gallon = 14.285 parts per million 1 part per million = 0.0588 grain per U. S. gallon 1 part per million = 0.07 grain per Imperial gallon 1 Chase Palmer, Geochemical interpretation of water analyses, Bull. 479, U. S. Geol. Survey, 1911. 2 Cfr. F. W. Clarke, Geochemistry, 1916, p. 63. G. S. Rogers, The interpretation of water analyses by the geologist. Econ. Geol, vol. 12, 1917, pp. 56-88. CHAPTER V THE CHEMICAL WORK OF UNDERGROUND WATER METAMORPHISM AND MINERAL DEPOSITS Stability of Minerals and Rocks. The underground water plays a very important part in the changes which take place in rocks, and the majority of mineral deposits are formed by the aid of it. Near the surface it may completely saturate the rocks or move in large volumes on fractures. At greater depths where there is no active circulation it may be sparingly present as rock moisture. The great mass of underground water is of atmos- pheric origin but as all magmas contain water which is given off upon solidification some waters in the rocks may be of magmatic origin. Solution and precipitation go on continuously; one or the other may predominate at any given place. The reactions which take place in the underground solutions extend over a wide range as to temperature, pressure, substances, concentration and time, and they differ markedly under the varying conditions. The study of these reactions was first seriously undertaken by G. Bischof and Justus Roth 1 and these pioneers have been followed by many eminent geologists who have devoted themselves to the study of chemical geology. 1 G. Bischof, Lehrbuch der chemischen und physikalischen Geologic, 1863-1866. Justus Roth, Allgemeine and chemische Geologie, vol. 1, Berlin, 1879. C. R. Van Hise, 'A treatise on metamorphism, Man. 47, U. S. Geol. Survey, 1904. C. R. Van Hise, Metamorphism of rocks and rock flowage, Bull., Geol. Soc. America, vol. 9, 1898, pp. 269-328. C. R. Van Hise, Some principles controlling the deposition of ores, Trans., Am. Inst. Min. Eng., vol. 30, 19QO, pp. 27-177. U. Grubenmann, Die krystalHnen Schiefer, Berlin, 1910. F. Becke, Ueber Mineralbestand urid Struktur der krystallinen Schiefer, Ninth Session Internat. Geol. Congress, Vienna, 1903; also Sitz. Ber., k. k. Akad., Vienna, 1903. John Johnston and Paul Niggli, The general principles underlying meta- morphic processes, Jour. Geology, vol. 21, 1913, pp. 481-516; 588-624. C. K. Leith and W. J. Mead, Metamorphic Geology, New York, 1915. 66 CHEMICAL WORK OF UNDERGROUND WATER 67 One of the most fruitful conceptions developed in recent years is that of the limits of stability of minerals and rocks. Conforming to increasing heat and pressure, zones exist in the earth's crust, gradually merging into one another but each characterized by certain groups of minerals that are stable only under the conditions prevailing in that particular zone. No mineral is absolutely stable. If subjected to certain conditions of temperature or in contact with certain solutions it will melt, decompose, dissociate or dissolve. At the surface under the influence of atmospheric waters with oxygen and carbon dioxide practically no minerals are stable except a few oxides, hydroxides and native elements. In consequence of the reversible nature of chemical processes under changing conditions each mineral thus has its stability field or "critical level" which it can not leave without undergoing decomposition. The mineral aggregates, that is, the rocks, also follow this law and as the rock minerals have usually been formed in closely analogous ways most of the component minerals will become unstable more or less simultaneously. Certain minerals, few in number, are less sensitive than others to such changes and recur under the most different conditions. They are designated "persistent minerals" and are in general of simple composition and do not contain the hydroxyl mole- cule; among them are quartz, magnetite, pyrite, chalcopyrite, fluorite, calcite and native gold. Orthoclase, all plagioclases, biotite, augite, olivine, the spinels, cordierite, and garnets develop and are fully stable only at high temperatures. Minerals rich in water, like chlorite, serpentine, and talc, are characteristic of lower temperatures. Other minerals, like muscovite, zoisite, epidote, hornblende, and albite, develop by preference under strong pressure. The varied composition of the crust, the unequal distribution of the underground water, the changing pressure, and the great differences in temperatures even at the same horizon make it difficult to establish strict rules and well-defined zones. One merges into another. Besides, stability is a relative term. Some rocks, like granite, are really stable only shortly after their com- plete consolidation. Under the influence of percolating deep waters the minerals of the granite are unstable, as they are in the zone of weathering. But the changes take place so slowly that at many places they can scarcely be perceived. Other 68 MINERAL DEPOSITS rocks, like calcareous shales, are stable at moderate depths, but easily subject to recrystallization under pressure and rising temperature. The results of the reactions differ widely according to the composition of the waters. The minerals that develop in a rock charged with a slight amount of moisture are not the same as those that appear when the rock is penetrated by rapidly moving solutions, charged with salts and gases of foreign origin. Metamorphism. 1 The term metamorphism meaning strictly "a change in form," was proposed by Lyell in 1833 to express the changes of sedimentary beds to slates, quartzite, crystalline limestone, etc. Later it was extended to the development of schists and slates from igneous rocks by pressure and recrystalliza- tion. Still later, for instance, by C. R. Van Hise it has been employed in a wide sense so as to cover any change in the com- position and structure of any rock, through whatever agency and with or without gain or loss of substance. This would include weathering and the development of any kind of epigenetic deposit, such as mineral veins, in a rock. Geologists have not generally accepted this wide definition. Metamorphism is here reserved for the processes which result in a partial or complete crystalliza- tion or recrystallization of solid masses of rocks, 2 as in gneiss from granite or mica schist from clay shale. Though the mechanical effects of pressure may be conspicuous, metamorphism is always characterized by chemical changes in the component minerals. The composition of the rock as a whole may remain fairly constant. For practical purposes we may distinguish between static, dynamic, igneous and hydrothermal metamorphism. Static metamorphism proceeds without stress, at slight depths and under influence of a slight amount of water. At great depths and high temperatures a static recrystallization under great load may be recognized. 3 Dynamic metamorphism is effected under stress at higher or lower temperatures. These two are regional and proceed without marked changes in composition. Igneous metamorphism includes the effects of magmas on adjacent rocks and is a high temperature process. It is about 1 For a thorough discussion of the various uses of this term, see: R. A. Daly, Metamorphism and its phases, Bull. Geol. Soc. Am., vol. 28, 1917, pp. 375-418. 2 A. Barker, Geol. Mag., vol. 6, 1889, p. 15. 3 R. A. Daly,. op. cit., p. 400. CHEMICAL WORK OF UNDERGROUND WATER 69 equivalent to contact metamorphism but includes also the effect of igneous injection and pegmatitization. Hydrothermal metamorphism includes the changes effected in rocks by circulating hot ascending waters. Igneous metamor- phism may be local or regional and in part involves changes of composition. Hydrothermal metamorphism is local and almost always involves changes of composition. Metasomatism or Replacement. The geological importance of metasomatism or replacement has already been pointed out on p. 26. The word metasomatism, meaning a change of body, first used by C. Naumann to designate some kinds of pseudo- morphism, is now applied to the process of practically simultane- ous capillary solution and deposition by which a new mineral of partly or wholly differing chemical composition may grow in the body of an old mineral or mineral aggregate. The secondary minerals of any metamorphic rock result from metasomatic action. Rocks are termed metasomatic if their composition has been materially changed by replacement of the original minerals. Pseudomorphs and petrifications often furnish direct and in- controvertible evidence of processes of replacement. Metasomatism is met everywhere and at all depths in sedi- mentary and igneous rocks and shows that the rock minerals have been subjected to conditions under which they were unstable. The development of chlorite in augite, sericite or kaolin or calcite in feldspars, or galena in limestone is due to metasomatism. The typical metasomatic processes, traced with the highest magnifying power, show no space between the parent mineral and the metasome, as the newly developed mineral may be designated. The fibers and blades of sericite project into quartz without the slightest break in the contact. Rhombohedrons of siderite develop in quartzite, their crystal faces cutting across the grains without any interstices. Perfect prisms of tourmaline develop in feldspar grains, and sharp cubes of pyrite in primary feldspar or quartz. Metasomatic rocks, that is rocks which have suffered a change in composition, are very common in mineral deposits and are often produced by strong and rapidly moving solutions (usually a.que- ous, sometimes gaseous) which penetrate the material through veinlets and pores. There are many cases of complete or almost complete metasomatism, for instance of limestone by sulphides and quartz in which the chemical composition has been absolutely 70 MINERAL DEPOSITS changed. In contrast to this the ordinary metamorphic processes in rocks are carried on by the scant rock moisture and while there is metasomatism in detail, the composition as a whole is but little changed. For description of metasomatic processes and for criteria of metasomatism see Chapter XI. Dissemination refers to grains or crystals distributed in a rock and is without genetic significance. Impregnation is a genetic term and means that the mineral introduced is later than the rock; it may have developed by metasomatic processes or by filling of pore spaces or other cavities. Cementation is used to indicate the filling of interstices in porous or shattered rocks. The Law of Equal Volume. It is necessary to distinguish between (1) metasomatic changes proceeding in free crystals or grains, or in loose aggregates under light load, where the force of crystallization can easily overcome the restraining pressure, and (2) metasomatic changes proceeding in rigid rocks where the new mineral is forced to make room for itself by solution of the host mineral. In the first case the volume changes proceed according to the chemical formula. In the second case, the replacing mineral occupies exactly the space formerly filled by the primary mineral; the force of crystallization is of little or no direct influence, but as the pressure differs in intensity according to the crystallo- graphic directions and as solution proceeds most actively at points of greatest pressure the development of crystal faces is thereby explained. 1 The chemical formulas by which some kinds of replacement are usually expressed do not represent the actual change for these formulas are based on equal weights and will indicate definite changes in volume. The conversion of orthoclase to sericite is usually considered to take place according to the following reaction, 3KAlSi 3 8 + H 2 + C0 2 = KH 2 Al 3 Si 3 Oi 2 + K 2 C0 3 + 6SiO 2 Orthoclase Sericite which involves a decrease in volume of 15.5 per cent, even if the Si0 2 is assumed to have recrystallized as quartz. If, how- x W. Lindgren, The nature of replacement, Econ. Geol, vol. 7, 1912, pp. 521-535. Volume changes in metamorphism, Jour. Geol, vol. 26, 1918, pp. 542-555. CHEMICAL WORK OF UNDERGROUND WATER 71 ever, one volume of orthoclase has been replaced by an equal volume of sericite this equation is not correct, and by a calcu- lation of the quantities of silica, alumina, etc., contained in one cubic centimeter of orthoclase and sericite, respectively, it will be found that a considerable addition of alumina is necessary. The actual formula is probably very complicated and could be established only if all the reactions taking place in the solution during the conversion of one mineral to the other were known. Many kinds of metasomatism, for instance, galena or barite replacing calcite (Figs. 62 and 64) can not be expressed by chemical formulas. One crystal, for instance, of pyrite may simultaneously replace parts of adjacent grains of different minerals, or may replace an aggregate of minerals in a fine grained rock. These well known facts will at once show that replacement is not the expression of one definite chemical reaction. The law of equal volumes has been repeatedly verified by many independent observers and there is little doubt that it holds for most metasomatic processes, both on a large and a small scale, both in general metamorphism and in mineral deposits. The most fundamental changes in rocks take place with practical constancy of volume. A great deal has been written on changes of volume and energy liberated or absorbed, that is absolutely valueless as a measure of the processes that have been going on in rocks. The time will soon come when these relations are more clearly recognized. Metasomatism in solid rocks proceeds independently of molec- ular weight, molecular volume and specific gravity. It does not take place "molecule for molecule," and it is not expressible in simple chemical equations. At the same time it is molecular or at least sub-microscopic in the sense that complex processes of solution and precipitation constantly take place in the solution films. Metasomatism by equal volume takes place in most perfect form when a rock is permeated by stagnant or slowly moving solutions. When, as sometimes happens, the solutions move rapidly, the nice equilibrium is disturbed. Local excess of solution over deposition is then expressed in drusy or cellular structure, and the law of equal volumes may 'fail to hold. It is not meant that the processes of solution, precipitation and chemical reactions which make up metasomatism do not 72 MINERAL DEPOSITS obey the law of mass action and laws of LeChatelier and Van't Hoff. They undoubtedly do. With increasing pressure a denser mineral will tend to form but the saturated solutions that fill the rock will at once proceed to fill any available space created by the deposition of the denser mineral with the next combination ready to be precipitated. General Definition of the Metamorphic Zones. That part of the earth's crust which is within our observation is called the lithosphere. It may be observed directly by borings or mining operations or indirectly when deformation and denudation bring up to the surface rocks which we know were once deeply buried. The lower limit of the lithosphere can, of course, not be ac- curately fixed; F. W. Clarke has suggested that it may be defined as extending 10 miles, or 16 kilometers, below the surface. Below the lithosphere lies the centrosphere, concerning which we have little definite information. The conceptions of Albert Heim developed by C. R. Van Hise led to a division of the lithosphere into an upper zone of fracture and a lower zone of rock flowage, in which only sub- capillary openings exist (p. 30) and deformation is effected by granulation. and recrystallization. Between them intervenes a middle zone of combined fracture and flowage. The limits of these zones are very indefinite owing to the greatly differing plasticity of rocks, e.g., a granite and a calcareous shale. The experimental proof given by F. D. Adams 1 that in supported rocks in depth openings in granite can persist to depths of at least 11 miles, or about 58,000 feet, at a uniform pressure of 70,000 pounds per square inch and at temperatures supposedly corre- sponding, that is, 550 C. shows that the zones overlap widely and have only value as relative conceptions. Van Hise divided the zone of fracture in an upper zone of weathering and a lower zone of cementation. The zone of flowage corresponds to the deep metamorphic zones in which minerals form by replacement only and in which the temperature is high and the pressure largely stress. 2 1 F. D. Adams, Jour. Geology, vol. 20, 1912, pp. 97-118. 2 Van Hise called the upper two zones the realm of katamorphism and the lower that of anamorphism. In the zone of katamorphism (kata, down) complex silicates break down and simpler, less dense minerals form. In the zone of anamorphism (ana, up) silicates are supposed to be built up with forming of denser minerals and compact texture. Since Leith and CHEMICAL WORK OF UNDERGROUND WATER 73 Later investigations have shown that any rock may be de- formed under stress. 1 The thrust required to develop deforma- tion in marble at a pressure corresponding to 4.2 miles would be 66,400 pounds per square inch; in case of granite, 138,500 pounds per square inch. At greater depths the required stress increases markedly. The pressure necessary for plastic deforma- tion is very much greater than the crushing strength of the rock at the surface. Zone of Weathering. The best defined zone is that of weather- ing, the depth of which is determined by the level of the ground- water, or by the depth to which free oxygen can penetrate in large quantities. In the zone of weathering the water percolates downward more freely than in the underlying zone, there is a tendency to the destruction of the rocks as units, and active transportation and concentration are characteristics. Chemical work progresses by means of water solutions and gases, also extensively through the medium of organic life; me- chanical disintegration is also important. The chemical reac- tions are oxidation, carbonatization, desilication, and hydration, the two first named mainly through decomposition of silicates by water containing carbon dioxide. As a consequence of these reactions the volume should 'increase, but so much is carried away by solution that a great reduction of volume ensues. Disintegration works hand in hand with decomposition and in advance of it; calcium, magnesium, sodium, and potassium are leached; the final products are a small number of minerals, largely hydrated compounds with low specific gravity and, for the most part, comparatively simple molecules. Almost all rock-forming minerals are unstable, as are the sulphides. These processes give rise to many mineral deposits of oxidized ores, which will be described in a later chapter. The great extent of weathering and the intensity of the changes are justly emphasized, especially in regions of soluble rocks like limestone. It is well to bear in mind that this is not because of rapid attack by waters, but because of long-continued action by extremely dilute solutions. This is shown by the relative purity Mead have changed these conceptions (Metamorphic Geology, 1915) and now confine katamorphism to processes of weathering a confusion has been introduced that is best cured by the dropping of both terms. 1 F. D. Adams and J. A. Bancroft, Jour. Geology, vol. 25, 1917, pp. 597- 637. 74 MINERAL DEPOSITS of the surface waters, which contain calcium and magnesium carbonates with lesser amounts of alkaline salts. The soluble products mainly escape into the rivers through the zone of dis- charge, which lies below the zone of weathering, and finally into the oceans. The Intermediate Zone. The rocks immediately below the zone of weathering are often saturated with water which dimin- ishes in quantity with increasing depth. The small pressure permits fracturing and brecciation, and the openings created by these processes, as well as those resulting from porosity, are filled with minerals deposited by circulating solutions. To a small extent these minerals result from material abstracted from the zone of weathering, but that zone is shallow in comparison with the zone of cementation, and the salts available from the weathering are, to a large extent, carried away by the surface drainage. The larger part of the minerals deposited have been derived from the rocks themselves; to a considerable extent they are derived from deep-seated sources, as, for instance, in the ce- mentation by quartz veins and veinlets near igneous intrusions. Hydration and carbonatization are the principal processes. Minerals like chlorite, serpentine, talc, sericite, epidote, and cal- cite develop, largely by metasomatic processes. Replacement and filling work together. Where stress is present it is mainly in one direction and shearing and schistosity may develop; in metamorphic schists some of the minerals formed are muscovite, chlorite, talc, horn- blende, zoisite, epidote, and albite; also quartz, pyrite, and cal- cite, probably magnetite and specularite. The clay slates with muscovite and albite, the chloritic schists, and the talc schists belong to this zone. According to the views of Van Hise such schistose rocks can only develop in the deeper zones. The Deeper Zones. In the deeper belts (included by Van Hise under the name of the anamorphic zone) the pressure and temperature are high; the latter in general above 200 C. Very little water is present. Minerals are formed mainly by replace- ment. In the upper part of the zone temperature and pressure work in the direction of diminished molecular volume. The pressure is largely stress that is, acting in one direction but hydrostatic pressure (transmitted in all directions) is becoming of importance. The important reactions are dehydration, the development of silicates, and deoxidation. Heavy silicates, like CHEMICAL WORK OF UNDERGROUND WATER 75 wollastonite, garnet, tremolite, and diopside, form in siliceous limestones or in pure limestones where the silica is supplied by plutonic intrusions. The minerals produced are numerous, stable, heavy, and complex. The rocks formed are compact and strong. However, the temperature is not sufficiently high to break up the molecules in which hydroxyl is firmly contained. The recrystallization takes place according to the law of Riecke,. 1 so that the solution prevails at places of maximum pressure, and deposition at those of minimum pressure. The re- crystallized products may assume lamellar structure extending perpendicularly to the pressure; this results in a "schistosity by crystallization." Among the minerals of this zone are mus- covite, microcline, albite, microperthite, oligoclase, biotite, zoisite, epidote, hornblende, staurolite, garnet, cyanite, titanite. magnetite, and ilmenite. Most of the micaceous and horn- blendic gneisses containing garnet, staurolite, etc., belong to this zone; also the mica schists, amphibolites, and glaucophane rocks. Where there is no stress in this zone, many igneous rocks, like granite, basalt, and rhyolite, are stable. In the lower part of the deep-seated zone the temperature is high and the tendency is toward an increase of volume. The hydrostatic pressure is enormous and stress almost non-existent, but high temperature is the dominant feature. There are no minerals containing the hydroxyl molecule except biotite, and the characteristics are, therefore, the prevalence of anhydrous minerals of great molecular volume. Characteristic minerals in the crystalline schists of this zone are orthoclase, all plagioclases, biotite, augite, olivine, garnet, cordierite, sillimanite, magnetite, and ilmenite. Many of the minerals of this zone also appear in the massive igneous rocks and in the contact-metamorphic rocks. The rocks are mostly gneisses, gradually approaching granites; also granulites, eclogites, and augite gneisses. Most of the igneous rocks are stable in this zone. The orthoclase or microcline in the crystalline schists of the deepest zone tends to microperthite in the middle depths and to sericite in the upper zone. Plagioclases of the deep zone may be transformed into albite and anorthite and finally to albite and zoisite or sericite. The augites change to hornblende and finally 1 E. Riecke, Ueber das Gleichgewicht zwischen einem festen homogen deformierten Korper und einer fliissigen Phase, etc., Nachr., Gesell. d. Wissensch., Gottingen, 1894, 4, pp. 278-284. 76 MINERAL DEPOSITS to chlorite. Olivine of the deep zone is transformed to horn- blende or (with feldspar) to garnet and becomes serpentine in the upper zone. Carbon dioxide and water doubtless escape from the deep zones upward wherever calcareous rocks containing free water or hydrated compounds become submerged in it. Exceptional supplies of heat contributed by igneous intru- sions may carry the reactions of the lower zones close to the surface. Relation of Mineral Deposits to the Metamorphic Zones. Though certain kinds of mineral deposits have originated at the surface or in the zone of weathering, the largest number have undoubtedly been formed in the zone of fracture, where circulation of solutions is comparatively easy. It is safe to assert that the great majority of ore deposits have been formed within 15,000 feet of the surface. Ore deposits do not, as a rule, form in the zone of flowage or anamorphism where the passage of solutions is prevented. An exception to this is where hot emanations from intrusive bodies penetrate and impregnate certain rocks like limestone without the necessity of ducts and cavities. Ore deposits may form also in the hottest zone where the solutions consist of magmas in which the free rearrangement of molecules is possible. During the ordinary metamorphic processes under static or dynamic conditions extensive changes in mineral composition and structure may be effected with minimal changes in the chemical composition of the rocks, so that it is possible to trace the origin of highly metamorphosed rocks by the aid of analyses. Meta- morphism can be, and usually is, effected with the aid of minute quantities of rock moisture and during the process there is little opportunity for extensive concentration of rarer constituents. Mineral deposits due to simple hydration or chemical rearrange- ment within the mass may result. Examples: soapstone by hydration of magnesian minerals; magnesite from carbonati- zation of serpentine; sulphur from reduction of gypsum by organic compounds; garnets developed in crystalline schists; concentration of hematite from lean primary ores; and many similar instances. A comparison of the mineral records of ore deposits, formed at various levels in the earth's crust, with the results obtained CHEMICAL WORK OF UNDERGROUND WATER 77 by a study of general metamorphism soon brings out the fact that the same laws do not apply to both cases, although there are points of similarity. Attention was called to this important feature in a paper on the metasomatic processes in fissure veins, and increasing knowledge emphasizes the distinction. 1 Neither the rules of Van Hise nor the three zones of Grubenmann will fit closely the case of the ore deposits. The reason for this is not difficult to find. In metamorphism one deals with small quan- tities of solutions, free from large amounts of carbon dioxide and hydrogen sulphide. The majority of ore deposits, on the other hand, were formed by large quantities of waters rich in these gases and heavily charged with alkaline salts. A large number of silicates and other minerals, fairly stable under the in- fluence of ordinary deep ground water, are incapable of existence in many vein-forming solutions. Biotite, amphibole, soda-lime feldspars, often also chlorite, serpentine, and magnetite are included among these. Deposits Related to Igneous Activity. Several important groups of ore deposits must therefore be considered apart from the ordinary processes of metamorphism. This especially applies to those deposits of the rarer metals which stand in closest genetic connection with intrusion and eruption of igneous rocks. It will be found that classified according to gangue minerals, they form three groups: (1) Those characterized by gangue minerals such as garnet, biotite, hornblende, pyroxene, specularite, mag- netite, tourmaline, topaz, apatite, and scapolite, all associated with quartz; (2) those in which quartz, calcite, dolomite, siderite, barite, sericite, chlorite, and albite occur; (3) those in which quartz, chalcedony, opal, calcite, dolomite, barite, fluorite, seri- cite, chlorite, and adularia form the more abundant gangue minerals. Zeolites occur only exceptionally in these groups, and kaolin is considered almost wholly a product of descending waters. The first group may be called the high-temperature deposits, though the highest temperatures during their genesis probably did not exceed 500 C. Ordinarily, but not necessarily, they were 1 W. Lindgren, Trans., Am. Inst. Min. Eng., vol. 30, 1900, p. 601. W. Lindgren, The relation of ore deposition to physical conditions, Econ. Gcol., vol. 2, 1907, pp. 105-127. Also Compte Rendu de laaX eme session du Congrts Geologique international, Mexico, vol. 2, 1906, pp. 701-724. 78 MINERAL DEPOSITS formed at great depths and are sometimes designated as deep- seated deposits. The second group is formed under intermediate conditions of temperature and in general also of depth. The third group is formed at moderate temperatures, prob- ably rarely exceeding 150 C., within a few thousand feet of the surface; many of them were developed close to the surface. These groups are described in more detail in a later chapter. On the whole it is evident that the great majority of ore deposits have been formed relatively near the surface and well within the zone of fracture, probably well within the upper 15,000 feet of the crust, and most of them within 10,000 feet of the surface; this applies even to contact-metamorphic deposits, many of which have developed along intrusive masses injected high into the zone of fracture. Some instances are known of such deposits having been formed within 3,000 feet of the surface. Only one class, that of the magmatic segregations, may have its origin at abyssal depths; but it is thought that more commonly the differentiation of these ores was effected after the intrusion of the magmas into the zone of fracture. Derivation of Minerals. Regarding the derivation of the val- uable substances of mineral deposits we have to distinguish several groups: 1. Those which require little or no concentration, but merely involve chemical readjustment, as the formation of sulphur from sulphate. 2. Those which- are formed by precipitation from solutions, the origin of which is beyond doubt, as, for instance, the salt beds derived from evaporation of sea water. 3. Those which are concentrated by mechanical means from well-known sources, like the gold placers. 4. Those which are derived by the solution and removal of waste material, like residual manganese or phosphate deposits. 5. Those which are derived from concentration in magmatic solutions by processes of differentiation; for example, certain titaniferous magnetites. 6. Those deposits, mainly of rarer metals, which require great concentration and concerning the origin of which more or less uncertainty still prevails. Concentration. Regarding the last group, it should first be stated that almost all rocks and inferentially all magmas contain CHEMICAL WORK OF UNDERGROUND WATER 79 small quantities of these rarer metals. There are two ways in which concentration of the rarer elements is possible. The first is by solution, by means of descending surface waters, of the small traces contained in the rocks and by corresponding deposition during the subsequent ascent of the waters. This solution of minor constituents is a slow and imperfect process, but that it actually occurs is shown by examination of natural waters and their deposits. The smaller the metal traces the slower and less complete is this process of solution. J. F. Kemp has studied this problem in some detail 1 and concludes that the amount which can be extracted by water percolating through the cracks of a rock is only one-sixth to one one-hundredth of the total amount of the particular metal contained. The leach- ing of compact masses of rocks by underground water is, therefore, at best an exceedingly imperfect process, and one on which it does not seem safe to rely for the concentration of the richer ores of the rarer metals like gold and silver. Veins and other deposits, conceded to have been laid down by purely meteoric waters, contain in fact little or no gold, silver, molybdenum, tungsten, and other rare metals. Moreover, the descending waters are cool and dilute, and thus their chemical action is slow. To obtain effective concentration of such metals the first few thousand feet of percolated rock should probably be left out of consideration. The second way of concentration is by disturbing the equilib- rium of a molten magmatic solution. Such disturbances would take place by the irruption of magmas into higher levels of the crust or by cooling of the magma, or, in other words, by changes in pressure and temperature. From the study of volcanic phe- nomena it is known that under such circumstances certain substances are expelled from the magma, and that among these are water, halogens, alkaline salts, and a number of the rarer metals. From the study of plutonic phenomena we infer that a still more thorough expulsion of these substances was effected during the intrusions of deep-seated magmas. Gold, silver, lead, zinc, copper, molybdenum, tungsten, tin, and many other rarer metals certainly have a place in the list of magmatic emanations. Associated with these elements are the ions of sul- phur, carbon, chlorine, fluorine, boron, and other elements. 1 J. F. Kemp, Problem of the metalliferous veins, Econ, GeoL, vol. 1 , 1905-1906, pp. 207-232. 80 MINERAL DEPOSITS This concentration is effected automatically and with ease, and these elements, dissolved in water, ascend, propelled by the expansive force of the gases. The gaseous solutions will seek the fractures and fissures on their upward paths. Their high temperature facilitates the solution of other elements in the surrounding rocks. In upper and cooler levels the gases condense to liquid solution; precipitation begins by reduction of pressure and temperature or by reaction with the adjoining rock minerals. Finally, meteoric waters mingle with the mag- matic and this again causes deposition and ultimately the still warm waters issue as ascending springs at the surface. Both methods are used in the work of deposition by waters in the crust. The first is applicable to the more common gangue minerals and to the more abundant ore minerals. The second, it is thought, must be assigned as the main cause of the rich deposits of gold, silver, and other rare metals. The former class of minerals is found in all parts of the world ; the latter is confined to districts where igneous forces have been active. This conclusion is supported by an impressive array of ob- servations of the various mineral deposits related to igneous rocks. The weight of the cumulative evidence is exceedingly strong, but is perhaps not fully appreciated, except by those who have made the study of these deposits their specialty. UNDERGROUND TEMPERATURES 1 The increment in temperature in the upper part of the earth's crust is generally assumed to average 1 C. for 30 meters or nearly 100 feet. Beginning with a surface temperature of 11 C. at a depth of 100 feet, corresponding to the mean annual temperature of a place in the temperate zone, we would have at a depth of 1,000 feet, 20 C.; at 9,000 feet, 100 C.; at 20,000 feet, 210 C.; and at 35,000 feet, 360 C., which is near the critical tem- perature of water (364 C.). As a matter of fact but little is known about the increment at great depths. Actual measure- ments within the accessible zone or to depths of about 6,000 feet show considerable divergences from the average figure given above. In some cases the increase in temperature is not quite uniform. 1 J. D. Everett, Evidence before the Royal Commission on coal supplies, London, 1904. Also in Reports of the British Association, 1882-1904. An increment of 1 C. in 100 feet equals 1 F. in 55 feet. The Royal Com- mission considered that the average would be 1 F. in 64 feet. CHEMICAL WORK OF UNDERGROUND WATER 81 The best summary of the results obtained in widely separated parts of the world has been given by Koenigsberger, 1 who has also given important data regarding the influences which increase or diminish the geothermal gradient. The results best available for general statements of the nor- mal increment have, as a rule, been obtained from deep bore- holes in regions of slight relief, far from large bodies of water, and in little-altered rocks, with no Tertiary or post-Tertiary intrusions, and containing no large deposits of coal or oil. Obser- vations in deep mines are probably somewhat vitiated by the cooling effect of ventilation; in new drifts and stopes, ventilation should not greatly affect the results obtained from bore-holes in the rocks. The following data are abstracted from the tables of Koenigs- berger : GEOTHERMAL GRADIENTS IN UNALTERED ROCKS (NOT RECENT ERUP- TIVES) AND REGIONS OF FLAT RELIEF. BORE-HOLES Locality Gradient in meters Gradient infect Depth in meters Depth inl "' Author Martincourt, 1 France .... 31.0 101 1,200 3,937 Sperenberg, Berlin 32.5 107 1,268 4,160 Dunker. Sennewitz, Halle 36.6 120 1,048 3,438 Schladebach, Merseburg. . 35.7 117 1,236 4,055 Dunker. Paruschowitz,* Silesia .... 30.7 101 1,959 6,428 Czuchow,* Silesia 29.6 97 2,239 7,346 Michael and Quitzew. Bay City, Michigan 36.8 121 1,050 3,445 Lane. Marietta, West Virginia . . 37.9 124 1,360 4,462 Hallock. Homewood, Pennsylvania 36.7 120 1,309 4,295 Cummins. Wheeling, West Virginia 40.7 133 1,360 4,462 1 J. Koenigsberger and M. Miihlberg, Ueber Messungen der geother- mischen Tiefenstufe, Neues Jahrbuch, Bail. Ed. 31, 1911, pp. 107-157. (Contains also list of literature and technique of measuring temperatures.) Recent investigations relating to the measurement of temperatures in deep drill holes by maximum thermometers and thermo-electric methods are found in John Johnston and L. H. Adams, Econ. Geol., vol. 11, 1916, pp. 741-762. Note. The deepest bore hole in the world is that of the Goff Farm near Clarksburg, West Virginia. On January 24, 1918, this had reached 7,350 feet and the temperature gradient is 1 F. in 51 feet. The boiling point of water should be reached at about 10,000 feet. U. S. Geol. Survey, Press Bulletin 357, 1918. 2 At Paruschowitz, Czuchow, and Martincourt some coal beds are present. 82 MINERAL DEPOSITS The influence of cool bodies of water in lowering the earth temperature is shown in the following data: Locality Gradient in meters Gradient in feet Depth in meters Depth infect Author Port Jackson, N. S. W., 44.0 144 833 2,733 David. Australia, Tokio Japan 39.8 130 ' 361 1.184 Tanakadate. Pas de Calais, France 56.6 185 1,400 . 4,593 Le Prince-Rinquet. Copper mines of Lake Superior: a. Osceola, 8 km. from 42 138 303 994' the lake, b. Atlantic, 3 to 5 km. 52-55 171-180 276 905 Wheeler and from the lake, Supan. c. 1.5 km. from the lake, 67 220 508 1,667 d. Close to the lake 123 404 1,396 4,580 Underneath high ridges and mountains the increase is slow: Locality Gradient in meters Gradient in feet Depth in meters i Depth infect Author Mont Cenis (summit) 50 164 Gotthard (summit) 44 144 Stapff 43 5 143 Schardt In or near recent eruptive rocks the increase is often rapid. This rapid increase is even noticeable in Tertiary eruptions. The following data are from borings: Locality Gradient in meters Gradient in feet Depth in meters Depth in feet Author Sulz (Wurttemberg) Macholles, France Buda-Pest 24.1 14.2 15.0 79 46 49 710 1,005 903 2,329 3,329 2 963 Braun and Waitz. Michel-Levy. Szab6 In the vicinity of heat-producing waters, or where chemical processes of decomposition are active, the increase is especially rapid. CHEMICAL WORK OF UNDERGROUND WATER 83 Locality Gradient in meters Gradient in feet Depth in meters Depth in feet Author Idria, Austria 10 33 329 1,079 Scheinpflug and Holler. Comstock, Nevada 17.1 56 (457 \672 l,499l 2,205 / G. F. Becker. In coal mines and in borings in coal-bearing strata the increase is more rapid than the normal, owing to the chemical processes in the coal beds. Locality Gradient in meters Gradient in feet Depth in meters Depth in feet Author Charmoy, Creusot (bore- 26 85 1,168 3,832 Michel-Le"vy. hole). Paruschowitz, Silesia (bore-hole): Above coal 26 85 1,122 3,681 Hendrick. Below coal 35 115 1,959 6,427 Hendrick. Gelsenkirchen, Germany. . 23.5 77 705 2,313 Rosebridge, England 29.5 97 (cooled by ventilation). Similar relations appear to exist in regions producing petro- leum. 1 Locality Gradient in meters Gradient infect Depth in meters Depth infect Author Santa Maria, California. . 23.0 76 1097 3,599 Arnold and Ander- Berekei, Caucasus Apsheron, Russia 23.0 28.4 76 . 93 1000 300-700 3,281 984-2,297 . son. Kelickij. Solubjatnikow. In mines of various kinds the increase may be more rapid or more slow than normal. The cooling by ventilation reduces the temperature to some extent. 1 For later information on this subject see H. Hoefer, Temperature in oil regions, Econ. GeoL, vol. 7, 1912, pp. 536-541 and E. De Golyer, The sig- nificance of certain Mexican oil temperatures, Econ. Geol., vol. 13, 1918, pp. 275-301. 84 MINERAL DEPOSITS Locality Gradient in meters Gradient infect Depth in meters Depth in feet Author 31.0 102 457 ca. 1,500 D'Aubuisson. 33 108 610 ca. 2,000 Thomas. Bendigo, Victoria (New Chum 42.7 140 1,110 3,645 Jenkins. Railway). Ballarat, Victoria 44.2 145 634 2,080 Jenkins. Witwatersrand, S. A 115 377 1,200 3,900 Marriott. Althoughjin some mines the increase is about normal, in other mines it is remarkably slow. At Bendigo, 1 where gold- bearing quartz veins occur in Ordovician sandstone, the rock temperature at the greatest depth attained, 4,600 feet, is only 112 F. (44.5 C.). At this depth the water is salty and has .a temperature of 114 F. At St. John del Rey, 2 a gold mine in the schists of southern Brazil, at 4,000 feet below the adit tunnel, the temperature is only 95 F., or 35 C. On the Witwatersrand, in the Transvaal, 3 a temperature of 65 F. prevails at 500 feet; the increase down to 3,900 feet is regular at the rate of 1 C. per 360 feet; at the lowest depth the temperature is only 84.4 F., or about 29 C. According to Marriott the natural ventilation reduces the temperature near the workings 5 to 6 C. No reason is known for the slow increase of temperature in the Transvaal and in Victoria. Koenigsberger has suggested that the decided increase in the earth temperature near oil pools and beds of coal (except anthracites) may be utilized for the prognostication of the occurrence of these substances near a given bore-hole. In temperature measurements a maximum instrument con- structed on the same principle as a clinical thermometer is most practical; an instrument about 25 centimeters long is recom- mended. An ordinary high-class chemical thermometer reading to 0.2 C. may also be used. For the measurement of tempera- 1 W. J. Rickard, Deep mining at Bendigo, Mining Magazine, London, 1910, pp. 281-282. 2 Eng. and Min. Jour., July 3, 1909. > H. F. Marriott, An investigation of earth temperatures, etc., Trans., Inst. Min. and Met., 1906. See also The Mining Journal (London), April, 1906, p. 479. CHEMICAL WORK OF UNDERGROUND WATER 85 ture in mine workings H. C. Jenkins 1 suggests some rules summar- ized below: 1. Temperatures should be taken in new workings which are rapidly pushed. Otherwise cooling may affect the result. 2. When possible the rock should be free from easily oxidized sulphides. Considerable heat is developed when pyrite oxidizes, as is well illustrated in many mines. 3. Holes should be bored 6 feet and, if possible, horizontal. 4. Wet ground should be avoided, as the readings will generally be too low. 5. One or two days should be allowed to permit the heat of drilling to dissipate. 6. The thermometer should be inserted, supported upon cork mounts in an outer closed glass tube, and the bore-hole closed. Note 1. In measuring rock temperatures in some Alpine tunnels 2 a hole 1.5 meters deep was bored in the side of the tunnel about a meter above the floor and slightly inclined upward. An ordinary thermometer was used, its length approximating 25 centimeters. It was cemented in a glass tube and inclosed in a double metal cylinder with cork rings. The mercury bulb was inclosed in a mixture of wax and turpentine. The metal cylinder with the thermometer was pushed in by means of a metal wire. The hole was then closed by a long plug of wood wrapped with woolen cloth, and stoppered by a wooden plug covered with gypsum plaster. The readings were taken 24 hours after the insertion. The errors or differences from the actual rock temperature are 0.5 C. Note 2. A. C. Lane 3 doubts Koenigsberger's conclusion that the vicinity of Lake Superior affects the temperatures, and believes rather that climate changes may be responsible for the present slow increase. Lane gives the average gradient at the Calumet & Hecla as 189 feet for 1 C. 1 H. C. Jenkins, Rock temperatures in Victoria, Proc., Aust. Assoc. Adv. Sci., vol. 9, 1902, pp. 309-318. 2 E. Kiinzli, Geologische Beschreibung des Weissensteintunnels, Beitrdge zur Geologischen Karte der Schweiz, Neue Folge, 21. Lieferung, Bern, 1908, p. 128. 3 The Keweenaw series of Michigan, Lansing, 1911, p. 763. CHAPTER VI THE ORIGIN OF UNDERGROUND WATER AND ITS DISSOLVED SUBSTANCES Origin of the Water. There is no physical or chemical criterion by which the origin of a given water can be determined. A pure water might possibly rise from interior sources and acquire saline constituents during the ascent. A water of superficial derivation might be conceived to have become charged with magmatic products. If it is possible to distinguish between waters derived from the surface and those brought up from the interior of the earth, the evidence must be circumstantial and depend on geologic and physiographic testimony, such as geo- logic structure, igneous history, rainfall, and drainage basins. There are then two modes of derivation : 1. Meteoric Waters. (a) The water is derived from the rain that falls on the surface, or from the water courses, or from the lakes, or from the present oceans and has simply descended into the earth in the cavities, fissures, or capillary openings to ascend at suitable places under hydrostatic conditions, or to remain stored in the rocks and almost stagnant (Chapter III). The term meteoric waters, 1 or surface waters, is applied to this group. (6) The water was mechanically included in the sediments of ancient oceans and has for geologic periods been a constituent of these strata. The term "connate" has been proposed by A. C. 1 Reginald A. Daly, Genetic classification of underground volatile agents, Econ. Geol, vol. 12, 1917, pp. 487-504. Daly shows that Posepny's term "vadose" (vadus = shallow) was applied by him to the descending waters above the water level, that is to the zone of gathering (Chapter III). Authors have used it since with different meanings, in each case including a certain part of the atmospheric waters. It seems as if the science could dispense with the word vadose. Daly following Archibald Geikie has suggested "epigene" to cover the underground activities of both fresh and marine waters. The term phreatic, applied by Daubre"e to a somewhat indefinite part of the meteoric waters, may likewise be dispensed with. THE ORIGIN OF UNDERGROUND WATER 87 Lane 1 to cover the origin of such waters, which really like those of the preceding class, are of meteoric origin. 2. Magmatic Waters. The water existed in the solution con- stituting an igneous magma. Crystallization of the magma or its irruption into higher levels of the earth's crust liberated the water as one of the most volatile constituents, thus permitting its ascent to cooler levels. Such water may be called magmatic or juvenile. 2 Underground waters may then be meteoric or magmatic or a mixture of both. Large quantities of magmatic water are rarely found except in regions. of present or recent igneous activity. Smaller parts of meteoric or magmatic waters may permanently or temporarily be withdrawn from the circulation by being held firmly by capillarity, by forming inclusions in minerals or by entering chemical compounds. Heat, pressure and chemical ac- tion may release part of these imprisoned waters when rocks sink into warmer zones or are engulfed by rising magmas. Thus while no one may doubt that the magma contains primary water a certain small part of it may be derived by the melting of rocks immersed in magmas. 3 Magmatic or Juvenile Waters. Volcanic phenomena are almost always accompanied by the emission of large quantities of steam and other volatile substances, and geologists generally have agreed that part of this water is a contribution to the atmosphere and hydrosphere from the magmas. 4 More recently A. Brun 5 in a work of much merit on the volcanic exhalations has arrived at the result that the magmas are anhy- drous, a view which is difficult to accept, though undoubtedly some classes of lavas like basalt, when arriving at the surface, are relatively poor in water. The clouds of vapors attending volcanic eruptions are, according to Brun, mainly volatilized chlorides, mixed with dust from explosions. Day and Shepherd 6 1 Bull. Geol. Soc. Am., vol. 19, 1908, p. 502. 2 E. Suess, Verb, der Gesell. deut. Naturf. und Aertze, 1902, pp. 133- 150; Das Antlitz der Erde, Wien, Bd. 3, 2te Halfte, 1909, pp. 630, 655. 3 R. A. Daly, Am. Jour. Sd., 4th ser., 1908, p. 48; Igneous rocks and their origin, New York, 1914, p. 249. Daly terms such waters re-surgent. 4 T. C. Chamberlain, Jour. Geology, vol. 7, 1899, p. 559. * Recherches sur 1'exhalaison volcanique, Geneva, 1911. See also F. W. Clarke, Geochemistry, Butt. 616, U. S. Geol. Survey, 1916, p. 282. A. N. Winchell, Brun's new data on volcanism, Econ. Geol., vol. 7, 1912, pp. 1-14. 6 Arthur L. Day and E. S. Shepherd, Water and volcanic activity, Bull., Geol. Soc. Am., vol. 24, 1913, pp. 573-606. 88 MINERAL DEPOSITS recently disproved Brim's thesis by subjecting the gases of the Kilauea crater on the island of Hawaii to a very careful study, and ascertained that when free from contamination of air they consist of nitrogen, water gas, carbon dioxide, sulphur dioxide and hydrogen. They concluded that the water released from the liquid lava as it reaches the surface is entitled to be considered an original component of the lava with as much right as the sul- phur or the carbon. It follows logically that some of this water from cooling lavas, with associated gases must mingle with the waters of meteoric origin. Regarding plutonic rocks the direct evidence is lacking but indirect testimony is supplied by the inclusions of aqueous solu- tions so commonly found in granular rocks and by the presence of minerals like mica and amphibole which contain the hydroxyl molecule. The best general evidence of the existence of juvenile waters is furnished, not by observation of the present springs, but by the study of old intrusive regions. Here the granites merge into pegmatite dikes, the latter change into pegmatite quartz, and this into veins carrying quartz and metallic ores, such as cassiterite and wolframite. Here we have evidence difficult to controvert that dikes consolidated from magmas gradually turn into de- posits the structure and minerals of which testify to purely aque- ous deposition; this admitted, it is difficult to see what would prevent such waters from reaching the surface in the form of ascending springs. Elie de Beaumont 1 was the first to give full expression to this view. He believed that there were two classes of hot springs: The first (the more common) is intimately related to volcanism and derives its waters and dissolved solids from this source; the second, and more exceptional, derives its water from simple infiltration. This view was accepted by de Lapparent, but Daubree arrived at the contrary conclusion, that both volcanism and thermal springs result from the infiltration of water from the surface; similar views were held by Fouque and have more recently been adopted by de Launay. 2 The views of Daubree found general acceptance in other countries; in the United States they were accepted by Le Conte, Van Hise, and others. All 1 Bull., Soc. Geol. de France, Serie 2, 1847, Tome 4, p. 1272. 2 L. de Launay, Recherche, captage et ame'nagement des sources thermo- mine' rales, Paris, 1892. THE ORIGIN OF UNDERGROUND WATER 89 waters appearing at the surface were considered of atmospheric origin and their salts were dissolved from the rocks percolated. About the year 1900 the importance of magmatic exhalations for the formation of mineral deposits began to be reasserted by various mining geologists among them Vogt in Norway, and Spurr, Kemp, Weed, and Lindgren in the United States. In 1902 Suess, 1 the eminent Austrian geologist, announced his belief that many of the springs in volcanic regions were of "juvenile" origin that is, that they now reach the surface for the first time and yield a permanent addition of water and salts, carried up from magmas cooling at great depth. As an excellent example of this the Carlsbad Springs were cited. The question now arises whether it be possible to establish criteria by which the magmatic waters may be distinguished from those of meteoric origin. Delkeskamp in Germany has attempted the solution of this problem in a series of suggestive papers. 2 He rightly considers temperature of little value as a criterion and points out that many springs of meteoric origin are hot, while some, strongly suspected to be of juvenile origin, are cold. The constant admixture with vadose waters forms another diffi- culty, but accounts well for the many derivatives of varying characteristics which accompany every spring of deep-seated origin. Seasonal variations of temperature, salinity, and quantity of water constitute excellent proofs of superficial origin. A practical constancy of salinity, temperature and quantity is said to be the best proof of a juvenile origin. Among the juvenile springs are those of Carlsbad in Austria, Ems and Wiesbaden in Germany. It is doubtful whether these criteria can be accepted. Much more work must be done before we shall be able to establish the magmatic origin of any given spring. Examples of Springs in Volcanic Regions. As pointed out on p. 63 there are two types of ascending hot waters which may be 1 Verhandl. Gesell. deutscher Nat. u. Aerzte, Karlsbad, 1902. 2 R. Delkeskamp, Juvenile und vadose Quellen, Balneologische Zeitung, 16, No. 5, Feb. 20, 1905, p. 15. R. Delkeskamp, Die Genesis der Thermalquellen von Ems, Wiesbaden, und Kreutznach und deren Beziehungen zu den Erz und Mineralgangen des Taunus und der Pfalz. Verhandlungen Gesell. deutscher Nat. und Aerzte, 1903, 2, First Part. A. Gautier, Compt. Rend. vol. 150, 1910, p. 436. See also reference in Econ. Geol. vol. 1, 1905, pp. 602-612. 90 MINERAL DEPOSITS of juvenile origin. They are the sodium carbonate and the sodium chloride-silica types, both common in regions of expiring volcanism. The former appear, for instance, in central Germany, in central France, in California and at various places in our Western States. The latter characterize the great geyser regions of Yellowstone Park, Iceland and New Zealand. The two classes break up through volcanic rocks and through the underlying plutonic rocks or crystalline schists. Whether these waters are wholly or partly of magmatic origin is a doubtful question. Arnold Hague, who spent many years in the study of the Yellow- stone Park has expressed the decided opinion that the present hot springs at this locality are of meteoric origin. 1 Such an origin is probably more difficult to establish for the geyser district of New Zealand. On the other hand many geologists are of the opinion that some of .the dissolved salts and gases at all of these places are of magmatic or juvenile origin. Salts from Sedimentary Rocks. There is little difficulty in the large sedimentary areas, where volcanism is absent. The great majority of underground waters are here simply of atmos- pheric origin, and, in spite of great diversity, the saline constitu- ents, as well as the gases, are readily traced to the sediments traversed. It is evidently possible for atmospheric waters to attain sufficient depth to acquire a high temperature, though this rarely exceeds 60 C. The Hot Springs of Virginia, the Ar- kansas Hot Springs, the Arrowhead Springs of southern California, and the Utah Springs in the Salt Lake Basin clearly derived their saline constituents from the surrounding sedimentary rocks. Examples of this class from the French Alpine region are plentiful. In all these waters the principal constituents are those of the surrounding sediments calcium-magnesium carbonates from the limestones and dolomites, brines from the saline formations, calcium sulphate from gypsiferous Triassic formations, sodium sulphate from the Cretaceous shales, hydrogen sulphide from the reduction of sulphates by oil or other organic matter often present in the strata, carbon dioxide from reactions between carbonate of calcium and other salts. The presence of connate waters is difficult to prove. It is simply inferred from the occurrence of strong sodium chloride and calcium chloride brines in certain sedimentary rocks. Any marine beds must necessarily have con- 1 Origin of the thermal waters in the Yellowstone National Park, Bull., Geol. Soc. Am., vol. 22, 1911, pp. 101-122. THE ORIGIN OF UNDERGROUND WATER 91 tained occluded sea water, but many geologists doubt whether it would have remained undisturbed during long ages. Salts from Igneous Rocks. The various types of waters from sedimentary areas are closely paralleled by those from igneous rocks. The waters of the upper circulation in igneous rocks are characterized by their consistent content of calcium car- bonate, to which large amounts of magnesium and ferrous carbonates are sometimes added. The attack of carbon dioxide on alkaline silicates gives alkaline carbonates and the oxidation of pyrite affords a small amount of sulphates. Soluble silica is added from all silicates, suffering partial decomposition. Occasionally these waters are tepid or hot, but probably only where they have exceptional opportunities for deep descent in regions of strong dislocations or contact with hot eruptive rocks. Salts of Volcanic Springs.- Some of the hot ascending springs in volcanic regions carry much sodium carbonate as stated above. The long-continued action of the hot water saturated with carbon dioxide on the feldspars of the surrounding rock undoubtedly yields this salt in large quantities, and the scarcity of calcium and magnesium carbonates is explained by their pre- cipitation with increasing percentage of alkaline carbonates. Considerable quantities of sodium chloride are, however, always associated with the sodium carbonate and sometimes indeed predominate; to find an adequate explanation of this is more difficult. Igneous rocks average, according to Clarke's calculation, only 0.07 per cent, of chlorine, and while there are some exceptional rocks containing sodalite, the sodium chloride waters are by no means particularly associated with this mineral. Considering that the water could extract only a small part of this chlorine, it is not easy to estimate the amount of rock which must be percolated to obtain a sustained flow of chloride waters of the concentration often found in hot springs. The sam.e reasoning applies to the alkaline sulphates which are abundant in some waters. It might be imagined that surface waters moving downward could have become charged with sodium chloride or sulphate while traversing saline sedimentary rocks, but such an explanation seems somewhat forced in the case of springs which issue from granite in a region where no such sedimentary beds are known to occur. Boron is a common constituent of many of these springs, for instance, the Steamboat 92 MINERAL DEPOSITS Springs, Nevada, and Ojo Caliente, New Mexico, both of which issue from granitic rocks. It is still more difficult to find a reasonable explanation for the presence of this substance on any hypothesis of leaching. Tourmaline and datolite are of course present in some rocks, but the springs carrying boron exhibit no marked relation to areas where such boron minerals occur. It is true that boron occurs in saline sedimentary beds and that traces of it are often found in waters traversing them, but the quantities do not compare with those determined in many waters of volcanic associations. Similar statements can be applied to fluorine, though it is less abundant than boron. The geyser springs of Iceland, the Yellowstone Park, and New Zealand are rich in silica, and as most of them ascend through easily decomposed rhyolitic rocks, that substance may well be derived from leaching of the country rock. And yet when we note how veins rich in quartz are at places directly connected with pegmatite dikes, and how strong the evidence is against their deposition by leaching from surrounding rocks, we may well wonder whether this silica in the thermal waters is neces- sarily derived by solution of rock comparatively near the surface. And again, when we observe that chlorides form part of magmas, as indicated by the presence of sodalite, and remember that sodium chloride occurs as small crystals in the fluid inclusions of quartz phenocrysts, and finally note the abundance of chlo- rides at volcanic eruptions, would it not then be easier to account for this salt in the springs of volcanic regions by an easily effected concentration of volatile substances while the magma was still fluid, than by a laborious search for traces of chlorides in the congealed igneous rocks? Origin of the Dissolved Gases. Reduction of sulphates by organic matter and oil accounts satisfactorily for the hydrogen sulphide abundant in many sulphate and chloride waters in sedi- mentary rocks. Carbon dioxide occurs in many such waters, but the reactions by which this gas is produced are less well known. The rain water contains, of course, little carbon dioxide and far more is taken up during the percolation of the humus of the soil. This is soon absorbed in the formation of bicarbonates, but from the decomposition of these compounds gas may be set free. In igneous rocks similar processes may apply on a small scale, but many of the ascending waters in volcanic regions contain such enormous amounts of carbon dioxide that its explanation THE ORIGIN OF UNDERGROUND WATER 93 by chemical reactions within the water itself appears utterly insufficient. From the earliest times of geological science this difficulty has been recognized and the literature on the deri- vation of the carbon dioxide is extensive. It has lately been summarized by R. Delkeskamp. 1 In addition to the carbon dioxide absorbed by the water from the atmosphere and the soil humus, a possible source of superficial origin may be found in decomposing organic deposits such as peat and lignite, though the coals are more likely to give off hydrocarbons than carbon dioxide. Still another way in which the latter gas may be formed is by the reaction between acid waters, such as are often found in mines, and limestone. A more deep-seated source lies in the replacement near intru- sive contacts of limestone or dolomite by calcic or magnesic silicates. This process has taken place, on a larger or smaller scale, at the contacts of all intrusives in calcareous rocks, and there is no good reason to doubt that it is going on, in some localities, at present. The quantity of carbon dioxide available from this reaction is large and it is likely, indeed, that many thermal springs in regions of intrusives have been fed from such sources. But, on the other hand, coal beds and limestones cannot be supposed to exist underneath every volcanic region, and in large areas of granite rocks the hypothesis becomes decidedly improbable. It is known that many igneous rocks, particularly granites, contain liquid carbon dioxide as minute fluid inclusions, though it may be doubted whether they are as abundant as is assumed by some writers. On the assumption that the quartz contains a maximum of. 5 per cent, by volume of these inclusions Laspeyres has calculated the content of CO 2 in a cubic kilometer of granite as sufficient to furnish the springs of Nauheim, Germany, with carbon dioxide for 273,000 years. Such calculations carry little conviction to those who realize the difficulty involved in the absorption of any but a minimal quantity of this gas from the quartz grains by percolating waters; and besides, exhalations of carbon dioxide are not characteristic of areas of granite, but appear in regions of volcanic rocks without reference to the character of the basement rock traversed (Fig. 4). The interesting experiments of A. Gautier and others on the 1 R. Delkeskamp, Vadose und juvenile Kohlensaure, Zeitschr. prakl. Geol, February, 1906. 94 MINERAL DEPOSITS gases included or occluded (absorbed) in the minerals of a rock and set free on heating have been summarized by several writers, including F. W. Clarke 1 and F. L. Ransome. 2 A great number of exact analyses of these gases were made recently by R. T. Chamberlin, 3 who found in general that the various pulverized rocks yielded a total amount of gases (at C. and 760 mm. pressure) equal to from a fraction up to as much as 30 times the unit volume of rock; the gases determined were H 2 S, CO 2 , CO, * 10 20 30 4p 5jQ 60 Ki|flmeters FIG. 4. Carbon dioxide and sodium carbonate springs of central France. Black dots are springs. Shaded area shows extent of basaltic eruptions. CH 4 , H 2 , and N 2 , among which CO 2 and H 2 always predominated. The carbon dioxide is believed to be derived from the decom- position of small quantities of secondary carbonates, while in part it may also be included or occluded. The hydrogen and other carbon compounds are probably due to reactions of water vapor and carbon dioxide with some of the substances 1 Geochemistry. Bull. 616, U. S. Geol. Survey, 1916, pp. 276-281. J Econ. Geol, vol. 1, 1906, p. 688. 1 R. T. Chamberlin, The gases in rocks, Carnegie Inst. Washington, 1908, p. 80. THE ORIGIN OF UNDERGROUND WATER 95 contained in the rock, notably ferrous compounds. All of of these results are highly important and suggestive, and various hypotheses have been advanced showing how, by a sort of distil- lation, carbon dioxide and other gases might be given off and absorbed by ascending waters. There is indeed a possibility that some of the carbon dioxide in deep waters may have been derived in this way. But it is, perhaps, scarcely recognized that there is a great difference between heating a small quantity of pulverized rock in the air and obtaining the same amount of gases from a solid mass at great depth. It seems probable that pressure would prevent the escape of these gases, and if the mass of rock were heated to the melting-point it would undoubtedly acquire a capacity for absorption of far greater amounts of gases than those expelled by heating the powder to redness. The presence of liquid carbon dioxide in cavities in minerals of igneous rocks is proof of its occurrence in the molten magma consolidated in depth. Every eruption brings new evidence of exhalations from magmas congealing near the surface; and almost every volcanic district of recently closed igneous activity testifies to the persistence of this gas in escaping from the cooling lavas below. The Cripple Creek district, where gold-tellurium veins cut through the core of an old volcano, presents an excel- lent illustration of this condition. Imperceptible at the surface, exhalations of carbon dioxide become more marked in depth and their temperature, higher than that of the surrounding rocks, indicates that they came from below. In the extinct volcanoes of the Auvergne in France and of the Eifel on the Rhine, waters highly charged with carbon dioxide and exhala- tions of the same gas are extremely abundant. It seems difficult to escape the conclusion that the enormous quantities of this gas contained in the ascending waters of volcanic regions are of igneous origin a volatile constituent of the magma released when the magma was brought up to higher levels of less pressure in the earth's crust. These considerations apply equally well to the hydrogen sul- phide with which some of these springs are so abundantly sup- plied. The decomposition of sulphates by organic matter or other reducing agents may be appealed to in places, but in igneous rocks, like granite, it does not appear to be quantitatively suffi- cient, and as we know that this gas plays a prominent part in volcanic eruptions we may well feel justified in believing that the 96 MINERAL DEPOSITS waters ascending in regions of such eruptions may absorb this gas or alkaline sulphides and carry them to the surface. Rarer Elements Contained in Waters. In the meteoric waters in crystalline rocks the salinity is low and rarer metals are generally absent. There are usually some iron and manga- nese, and possibly refined methods might discover other heavy metals in extremely small amounts. Where these waters appear in mines they naturally take up certain amounts of the metals of the deposits. Careful examination of calcium carbonate waters in sedimentary rocks has disclosed traces of nickel, cobalt, copper, lead, zinc, strontium, and barium, much more rarely fluorine, boron, and iodine. Arsenic is often present and where the waters precipitate limonite from dissolved ferrous car- bonate, the ocher almost always contains traces of that metal. Where sulphates of iron and aluminum are among the saline con- stituents of waters in sedimentary rocks, determinable amounts of copper, zinc, cadmium, nickel, and cobalt may be found, with traces of lead, barium, and strontium. The quantity of cobalt usually exceeds that of nickel. Such waters issue at many places from beds of pyritic shale and owe their strong solvent power to the sulphuric acid generated by oxidation of pyrite. Chloride waters from sedimentary rocks are always com- paratively rich in bromine and barium, with traces at least of iodine and traces of boron, fluorine, and sometimes arsenic. From Wildbad, in Wiirttemberg, in a strong sodium chloride water, traces of tin besides the substances mentioned, have been reported. 1 From the calcium chloride springs of Cannstatt, the same authorities mention nitric acid, boron, iodine, fluorine, barium, arsenic, and manganese; in the ochery deposits also copper, lead, and antimony or tin. From the cold sodium chlo- ride water of Homburg, nickel, copper, arsenic, antimony, boron, and fluorine are reported; in the hot waters of Wiesbaden, copper, tin, arsenic, and boron. In almost all of the various waters mentioned above, traces of phosphoric acid are found. In hot ascending sodium chloride springs which issue in vol- canic regions rarer elements have often been determined; such waters are often rich in boron. Steamboat Springs, Nevada, contain notable amounts of arsenic and antimony with traces of quicksilver. The springs of the Yellowstone Park carry boron and arsenic, but are poor in other rarer constituents. 1 Stelzner and Bergeat, Die Erzlagerstatten, 2, 1905-06, p. 1220. THE ORIGIN OF UNDERGROUND WATER 97 The ascending sodium carbonate springs in volcanic districts also frequently contain boron and fluorine in notable amounts. Arsenic and copper have been found in the springs of Ems, and the same metals with lead also at Vichy. The Carlsbad Sprudel contains, according to Gottl, 1 traces of bromine, iodine, fluorine, selenium, phosphorus, boron, barium, strontium, lithium, titanium, tin, arsenic, antimony, copper, chromium, zinc, cobalt, nickel, and gold. The statements summarized above show clearly that traces of metals are by no means confined to springs of supposedly deep- seated, "magmatic" or "juvenile" origin, but that they occur in many different kinds of water. The presence of silver has appar- ently not been recorded, and that of gold only from the Carlsbad Springs. Quicksilver and large quantities of antimony seem to occur only in sodium chloride or sodium carbonate waters of the volcanic type, of which also higher amounts of boron and fluorine are characteristic. Arsenic is probably the most common of the rarer metals and has been found in all kinds of water. Copper, zinc, nickel, and cobalt are not uncommon, both in waters of sedimentary and in those of igneous origin. Lead is of rare occurrence. In minute quantities it is contained, together with zinc, in some calcium carbonate waters issuing from Paleozoic rocks of the Mississippi basin. Iron is characteristic of meteoric waters and occurs only in minute quantities in the hot ascending sodium springs. The Igneous Emanations. 2 At several places above the igneous emanations have been mentioned and the inference drawn that the waters in the crust of whatever origin must at places have absorbed such volatile substances. It may be well to describe briefly the character of these emanations. The active volcanoes constantly emit volatile matter from lava flows, craters, and fumaroles. Some of the less volatile materials crystallize as sublimates near the gas vent; other parts escape into the atmosphere. Chlorides 3 are given off in abundance by the erupted lavas 1 J. Roth, Allgemeine und chemische Geologie, Berlin, vol. 1, 1879, p. 570. 2 For summary on this subject see F. C. Lincoln, Magmatic emanations, Econ. Geol, vol. 2, 1907, pp. 258-274. 3 According to Brun congealed lavas, upon heating, give off considerable quantities of chlorides, chlorine, hydrochloric acid, sulphur dioxide, and nitrogen. This statement still awaits confirmation. 98 MINERAL DEPOSITS and crystallize near the fumarolic vents; they comprise the salts of sodium, potassium, aluminum, ammonium, iron, copper, lead, and manganese. Sulphates come next in abundance; more rarely are fluorides or oxyfluorides found. Boric anhydride and sulphur are common products of sublimation at many places. Selenium and tellurium have been recognized in sulphur of volcanic origin. Arsenic in the form of realgar is reported, and the presence of cobalt, tin, bismuth, and molybdenum has been established, most of the detailed work having been done at the Italian volcanoes. Among the sulphides, pyrite, pyrrhotite, and galena 1 are mentioned. Specularite is a common product of the reactions effected by the volcanic gases. Even silicates, such as leucite, augite, hornblende, and sodalite, may be formed by sublimation. Among the volcanic gases nitrogen, carbon dioxide, and hydrogen sulphide 2 are the most important and their emission, particularly that of carbon dioxide, may continue long after the igneous activity has subsided and even after the volcanic cone has been eroded. Evidence of this is furnished by the exhala- tions of carbon dioxide and nitrogen in the mines at Cripple Creek; of nitrogen at Creede and Tonopah; of carbon dioxide in the Tertiary gold deposits of New Zealand. Some fumaroles in volcanic regions give off superheated steam with many associated salts and gases. Very .interesting are the "soffioni" 3 of Toscana, Italy. These are vents emitting super- heated steam of a temperature of up to 190 C. and a pressure of 2 to 4 atmospheres. They contain much boric acid undoubtedly of volcanic origin. If all these substances are given off at the surface the intruding masses in depth must part with still more of their volatile con- stituents. All of these, whether from lavas or intrusives, must to greater or lesser degree mingle with the ground water. Beyond all doubt, ascending waters in regions of igneous activity must in places carry a considerable load of magmatic eamnations. 1 F. Zambonini, Mineralogia Vesuviana, Naples, 1910. A. Bergeat, Die Aeolischen Inseln, Abh. math-phys. Klasse, K. bayer Akad., 20, 1, 1899, p. 193. 2 Oxygen is usually and marsh gas (CH 4 ) sometimes present. 3 P. Toso, Bolletino del R. Comitate Geol., vol. 42, Rome, 1913, p. 122. CHAPTER VII THE SPRING DEPOSITS AT THE SURFACE Processes of solution and precipitation are in continual progress in the underground waters. By a comparison between the prod- ucts of the laboratory and those of nature we have arrived at the conclusion that a majority of mineral deposits have resulted from reactions in the underground water channels. Only at the surface, however, is it possible to study the actual progress of these chemical changes, and great interest, therefore, attaches to the deposits formed by the natural waters where they issue as springs from their underground path. The precipitation taking place in rivers, lakes and seas will be described in later chapters. On the whole the composition of the material deposited by springs is simple. Three main divisions are recognized: De- posits of limonite (iron hydroxide), calcium carbonate and silica. Mixtures of two or all of these substances are frequently ob- served. The deposits are known as 1. Ochers, 2. Tufas, traver- tines or calcareous sinters, 3. Sinters or siliceous sinter. The precipitation is in part due to cooling or escape of carbon dioxide but algae and micro-organisms frequently aid by secreting silica jelly, calcium carbonate, colloidal ferric hydroxides or manganese oxide. 1 Deposits of Limonite and Calcium Carbonate. Limonite is frequently deposited by superficial meteoric waters which con- tain ferrous carbonate and ferrous sulphate. Many such ochers contain a little manganese and traces of arsenic, nickel and cobalt. Other waters also deposit some limonite so that many sinters and tufas are stained by this compound. Analyses of such ochers are quoted by F. W. Clarke. 2 Calcium carbonate is probably the most common spring deposit, though the ordinary dilute surface waters rarely are 1 W. H. Weed, Ninth Ann. Rept., U. S. Geol. Survey, 1889, pp. 613-676. H. Molisch, Die Eisenbakterien, Jena, 1910. Jour. Chem. Soc., vol. 92, pt. 2, 1907, p. 888, abstract. 2 Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 205. 100 MINERAL DEPOSITS able to form important precipitates. Hot carbonated waters issuing from limestone often deposit large masses of such tufa, covering many acres with thick terraced beds. The Mammoth Hot Springs in the Yellowstone Park offers a beautiful example of such tufa. The precipitates are almost pure calcium carbonate with a little magnesium carbonate. In many of these springs the calcium carbonate is the least soluble constituent which remains after the others have been carried away. Thus, the sodium chloride springs of Glenwood, Colorado, yield a con- siderable deposit, and the sodium carbonate springs of Ojo Caliente, New Mexico, which are very poor in calcium, deposited at their former point of issue a porous tufa containing over 90 per cent, of calcium carbonate. This carbonate is no doubt deposited in crystalline form, though it is usually fine grained. Such deposits are not always calcite, for the presence of aragqnite has been proved in many spring deposits, for in- stance those of Hammam Meskoutine, in Algeria, and of Carlsbad, in Bohemia. 1 Deposits of Silica. At hot springs containing much silica, this substance is abundantly precipitated because of evaporation, through mixture with other waters, or, according to W. H. Weed, by the action of certain hot-water algae. The material is de- posited as a colloid jelly which subsequently hardens to opaline or chalcedonic silica. Such sinters are formed by the hot springs of the Yellowstone Park and may contain up to 95 per cent, of silica. Sodium is often present as chloride or carbonate. The Steamboat Springs of Nevada 2 deposit a sinter of pure silica or mixtures of calcium carbonate and silica, the latter being present as chalcedony, or small crystals of quartz. (See Fig. 5.) This sinter contains weighable quantities of sulphides of mercury, lead, copper, arsenic, and antimony; the presence of gold and silver was also determined, and traces of manganese, zinc, cobalt, and nickel were found. Antimony sulphide, Sb 2 S 3 , is deposited as the amorphous " metastibnite " in quantities large enough to color the sinter red in places. In a shaft sunk into the gravel immediately adjoining the granite hill from which the springs issue, Lindgren 3 discovered delicate crystals of stibnite covering 1 H. Vater, Zeitschr. Kryst. u. Min., vol. 35, 1902, p. 149. 2 G. F. Becker, The quicksilver deposits of the Pacific coast, Mon. 13, U. S. Geol. Survey, 1888, p. 341. 3 W. Lindgren, Trans., Am. Inst. Min. Eng., vol. 36, 1906, pp. 27-36. THE SPRING DEPOSITS AT THE SURFACE 101 the pebbles and associated with thin crusts of black opal and grains of pyrite or marcasite. The sinter of the Yellowstone Park often contains arsenic, especially in the form of scorodite (FeAsO 4 .2H 2 O), and near one of the springs which was impregnated with pyrite Weed 1 noted rhyolite that contained traces of gold and silver. On the whole, however, the Yellowstone spring deposits are poor t n the rarer metals. The same author, associated with Pirsson, 2 FIG. 5. Section of chalcedonic spring deposits from Steamboat Springs, Nevada. White areas microcrystalline quartz. Magnified 29 diameters. Crossed nicols. reports the occurrence of orpiment and realgar with native sulphur in a siliceous sinter from the Norris geyser basin. De Launay mentions a deposit containing orpiment at St. Nectaire, Puy-de-D6me, France. A calcareous sinter deposited by an ascending sodium car- bonate spring in the Geyser mine, Silver Cliff, Colorado, on the 1 Mineral vein formation at Boulder Hot Springs, Montana, Twenty- first Ann. Rept., U. S. Geol. Survey, pt. 2, 1899-1900, pp. 233-255. 2 Occurrence of sulphur, orpiment, and realgar in the Yellowstone National Park, Am. Jour. Sci., 3d ser., vol. 42, 1891, pp. 401-405. 102 MINERAL DEPOSITS 2000-foot level, yielded traces of lead, copper, zinc, nickel, and cobalt. At Hammam Meskoutine, in Algeria, a similar spring, according to Daubree, deposits tufas and pisolitic sinters in which, in the concretions, shells of calcium carbonate alternate with shells of pyrite; strontianite is deposited by the same spring. Quicksilver, gold, and silver have been recognized in the spring deposits of the geyser districts in New Zealand. From the Whakarewarewa hot springs at Roturoa (sodium chloride-silica FIG. 6. Section of chalcedonic spring deposits, from De Lamar, Idaho, showing vegetable remains. Magnified 35 diameters. Ordinary light. type) suiters have lately 1 been analyzed which yielded nearly 5 ounces of silver and about $1 in gold per ton. At De Lamar, Idaho, Lindgren found in rhyolite spring de- posits of flinty chalcedony, which included casts of vegetable remains and yielded traces of gold and silver. 2 (Fig. 6.) S. Meunier 3 reported 0.5 per cent, of cassiterite in siliceous 1 J. M. Bell, First Ann. Rept. N. Z. Geol. Surv., 1907, p. 100. 3 The gold and silver veins of Silver City, etc., Twentieth Ann. Rept... U. S. Geol. Survey, pt. 3, 1898, p. 187. 3 Compt. Rend., vol. 110, 1890, p. 1083. THE SPRING DEPOSITS AT THE SURFACE 103 sinter deposited by a hot spring at Selangor, in the Federated Malay States, but this statement has lately been challenged. 1 It has been shown that springs, hot or cold, may deposit lim- onite in abundance, with arsenic, manganese, and traces of other metals; and it is likewise proved that the carbonate and silica sinters of hot springs, particularly those of the NaCl or Na 2 C0 3 type, contain small quantities of the rarer metals, including gold, silver, copper, lead, zinc, antimony, arsenic, tin, and quicksilver. In very few instances has commercial ore been obtained from spring deposits at the surface. Quick- silver ores have been mined in New Zealand and ores of iron and manganese have been utilized in rare instances. The evidence that such waters have formed workable ore deposits is therefore strong but hardly conclusive; the remarkable poverty in metals of the deposits of the springs in the Yellowstone National Park, for instance, will to many seem an argument against the hydro- thermal theory of the genesis of ore deposits. Deposits of Other Gangue Minerals. 2 Calcite, quartz, chal- cedony, and opal are common products of deposition at the sur- face, but besides these the mineral deposits often contain such minerals as barite, ankerite and siderite, fluorite, and more rarely gypsum, strontianite, celestite, and zeolites. It will be necessary to examine the competency of the various waters to form these gangue minerals. Fluorine is present in traces in many waters, both vadose and deep, but appears in larger quantities in waters of the sodium carbonate type. Few authenticated instances of actual deposi- tion of fluorite by springs are recorded; the substance rarely occurs in crystallized form and the chemists have probably often neglected to test the sinters for fluorine. The Carlsbad Springs deposit a pisolitic sinter of aragonite and calcite. Accord- ing to Berzelius 3 and later chemists this contains a notable quan- tity of calcium fluoride. A limonitic variety of the spring deposit yielded 0.272 per cent, arsenic. 4 The analyses on page 104 also demonstrate that various phosphates may be precipitated as well as the carbonates of iron and strontium. 1 J. B. Scrivenor, Origin of tin-deposits, Perak, Chamber of Mines, p. 5. 2 The best resume' of the older data regarding spring deposits are found in Roth's Allgemeine und Chemische Geologie, vol. 1, 1879, pp. 564-596. 3 Pogg. Ann., 74, 1823, p. 149. 4 Blum and Leddin, Am. Chem. Pharm., vol. 73, 1850, p. 217. 104 MINERAL DEPOSITS COMPOSITION OF DEPOSITS OF CARLSBAD SPRINGS Berzelius, Analyst A B 12.13 Ferric oxide Manganese oxide 19.35 53 20 0.43 trace 96 47 30 Basic ferric phosphate Aluminum phosphate . 1.77 0.60 0.10 06 Calcium fluoride 0.99 Silica 3 95 Water . ... 9.00 1 59 , At Plombieres, in the French Vosges, springs with a tempera- ture of 70 C. issue from granite. They have a low salinity (360 parts per million) and contain mainly sodium sulphate and silica, believed to be present in part as sodium silicate, also traces of arsenic and fluorine. The derivation of these salts is doubtful and the springs are apparently not directly related to volcanic rocks. They issue from well-defined fissure veins containing quartz and fluorite, and DaubreV found that the waters had actually deposited calcite, aragonite, and fluorite. The bricks and cements used by the Romans 2,000 years ago in the construc- tion of the baths at Plombieres were found to contain zeolites, chiefly apophyllite (containing fluorine) and chabazite, with opal and chalcedony. Chabazite is also reported by Daubre"e as deposited from springs at Luxeuil and at Bourbonne-les-Bains, which have a temperature of 46 and 68 C. respectively. In some of the mines of Cripple Creek, Colorado, gypsum is found suggesting deposition by ascending springs. Crystals of gypsum occur commonly near springs charged with calcium sulphate. Weed 2 has described how the Hunter Hot Springs, near Livingston, Montana, deposit this mineral in fractures in Tertiary sandstone; stilbite, a zeolite, is forming with the gypsum. The springs have a temperature of 64 C. and are weak mineral 1 Les eaux souterraines, 3, p. 31. 2 Economic value of hot springs and hot springs deposits. Bull. 260, U. S. Geol. Survey, 1904, pp. 298-604. THE SPRING DEPOSITS AT THE SURFACE 105 waters. According to a somewhat doubtful analysis they are rich in silica and alumina, but poor in calcium sulphate, so that Weed believes that at present they deposit more stilbite than gypsum. The presence of stilbite has also been noted by Weed 1 in the vein-like deposits, containing gold and silver, believed to be made by the present Hot Springs at Boulder, Montana; the stil- bite occurs in the predominating quartz, chalcedony and calcite. Lindgren noted the presence of a little adularia in the material. According to Lindgren 2 a spring deposit in New Mexico contains about 89.60 per cent, of calcium carbonate and 0.9 per cent, of calcium fluoride. There are no springs now at this place, but it is probable that the sodium carbonate of Ojo Caliente, a short distance lower down in the valley, formerly issued here. As shown by an analysis on page 61, the water contains a notable amount of fluorine. A vein of white crystalline fluorite is opened by a shaft close by the calcareous tufa and is believed to have been formed by the same waters. Both tufa and vein contain traces of gold and silver, and a few crystals of barite were observed in the vein material. W. H. Emmons and E. S. Larsen 3 have described a similar case from Wagon Wheel Gap, Colorado. Veins and replacements of fluorite in quartz porphyry and Cre- taceous sandstone near the sodium carbonate springs of Teplitz, Bohemia, have been described by J. E. Hibsch 4 and the evidence is convincing that fluorite was deposited by these thermal waters. Barite is deposited far more abundantly than fluorite. As shown above, many carbonate and even sulphate waters contain a little barium. It has been proved that alkaline bicarbonates with an excess of carbon dioxide can hold barium in solution, notwithstanding the presence of sulphates; sodium chloride and other salts also seem to retard the formation of barium sulphate. Haidinger observed that barite was deposited by the hot waters at Carlsbad, 8 and Becke noted the same at Teplitz. 6 At Idaho 1 Mineral vein formation at Boulder Hot Springs, Montana. Twenty' first Ann. Rept., U. S. Geol. Survey, pt. 2, 1899-1900, pp. 233, 255. 2 Econ. Geol, vol. 5, 1910, pp. 22-27. 3 Econ. Geol, vol. 8, 1913, pp. 235-246. 4 Tsch. M. u. P. Mitt, 25, 1906, pp. 482-488. 5 Jahrb. K. k. Reichsanstalt, 5, 1854, p. 142. See also Delkeskamp, R., Entstehung und Wegfiihrung des Baryts. Notizblatt d. Ver.f. Erdkunde (Darmstadt) (4), 21, 1900, pp. 55-83. Tsch. M. u. P. Mitt., 5, 1883, p. 115. 106 MINERAL DEPOSITS Springs, in Colorado, a hot sodium carbonate water issues from granitic rocks, and barite crystals are found in abundance in a cellular and decomposed dike rock at the mouth of the spring. Spurr 1 shows that the barium is contained in this dike rock and believes that the barite resulted from the reaction of the water on the rock. Barium is, however, far more commonly contained in sodium chloride waters, particularly in the brines of sedimentary strata. Many writers record the deposition of barite from such waters, and it is probable that wherever this mineral appears in large quantities in mineral deposits waters of this type have been active. An excellent example is reported from a mine near Clausthal, Germany, 2 where a spring of strong brine, which undoubtedly derived its salts from sedimentary strata, was encountered at a depth of 1,200 feet; this brine contained in grams per liter 67.555 sodium chloride, 10.509 calcium chloride, 4.360 magnesium chloride, 0.350 potassium chloride, 0.314 barium chloride, and 0.854 strontium chloride. When it was mingled with the ordi- nary mine waters, which carried sulphates, abundant precipita- tion of barium and strontium sulphate took place in the pumps and elsewhere, so that within a few years the capacity of the pipes became much reduced by this deposit. This compact material contained 92.44 per cent, barium sulphate and 4.32 per cent, strontium sulphate. The reaction is believed to be retarded by the presence of sodium and magnesium chlorides. According to P. Krusch, 3 barite is precipitated in the pumps at some Westphalian coal mines by a similar reaction between strong salt brine from the Triassic sandstones and potable water with sulphates, each ascending on separate faults and each deriving its contents of salts from sedimentary strata. Veins of barite with small amounts of galena, pyrite, etc., appear in the Carbon- iferous and in the Devonian. At a lower horizon quartz veins contain galena and zinc blende, both kinds of deposits having been made, according to Krusch, by these saline waters. Simi- ^J. E. Spurr, Prof. Paper 63, U. S. Geol. Survey, 1908, p. 165. 2 Lattermann, Die Lautenthaler Soolquelle und ihre Absatze. Jahrb. Preuss. geol. Landesanstalt, 1888, pp. 259-283. Ref. in Stelzner and Ber- geat, Die Erzlagerstatten, II, 1905-06, p. 1218. 3 Monatsberichte Deutsch. geol. Qesell., 1904, p. 36; Ref. in Zeitschr. prakt. Geol, 12, 1904, p. 252. THE SPRING DEPOSITS AT THE SURFACE 107 lar deposits of barite in the pipes of the pumping apparatus have been described from English coal mines. 1 An account by Headden 2 describes an interesting group of springs on the North Fork of the Gunnison River, Delta County, Colorado. They are cold, but contain free carbon dioxide and a little hydrogen sulphide and are essentially sodium chloride waters. At least one spring contains barium and all of them yield a little strontium. The Drinking Spring has a total salinity of about 1,656 parts per million. Small quantities of calcium, potassium, magnesium, barium (0.0132 gram per liter), strontium (0.0066 gram per liter), lithium, manganese, ammonia, iron, aluminum, also a trace of zinc, are present in the order stated; also sulphuric acid radicle (0.6254 gram per liter), silica, boron, and bromine, the latter three in very small amounts. The spring deposits a calcium carbonate sinter, which was found to contain 5.42 per cent, barite, but no gypsum or sulphur. Ferrous carbonate, or siderite, is sometimes observed, as in the Carlsbad " Sprudelstein " and in deposits of limonite formed in bogs and peat. Deposits of magnesium minerals are rare. H. Leitmeier 3 describes a deposit of hydrous carbonate of mag- nesia from the springs of Lohitsch in Styria; many springs, especially those whose waters have traversed sedimentary beds, contain organic matter and are probably competent to deposit hydrocarbons. The more common gangue minerals in certain classes of veins are thus deposited by spring waters, particularly by the warm sodium chloride and sodium carbonate springs. There are, of course, a great number of gangue minerals like tourmaline, garnet, feldspars, and similar silicates which cannot be expected to develop in water under the conditions of temperature and pressure prevailing at the surface. Summary. The deposits of ascending springs of undoubted meteoric origin contain opal, chalcedony, calcium carbonate, limonite, hydroxide of manganese, barite, siderite and pyrite. They often deposit sulphur by the oxidation of hydrogen sulphide. The ochery deposits very frequently yield small quantities of arsenic, copper, lead, zinc, nickel and cobalt. The springs of the sodium carbonate and sodium chloride- 1 J. T. Dunn, Chem. News, vol. 35, 1877, p. 140. 2 The Doughty Springs, etc., Proc., Colo. Sci. Soc., 8, 1905, pp. 1-30. 3 Zeitschr. KrystaU. u. Min., vol. 47, 1909, p. 104. 108 MINERAL DEPOSITS silica type in volcanic regions yield abundant deposits of opal, chalcedony, quartz, calcium carbonate, limonite, barite, siderite sometimes also pyrite. They also deposit fluorite which is rarely if ever found in the sinters of meteoric waters and yield smaller quantities of quicksilver, antimony, arsenic, lead, copper, zinc, tin, silver and gold. The rarer metals are thus more promi- nent and the waters are particularly characterized by a relative abundance of borates, arsenates and fluorides. The list of recognizable minerals deposited by springs at the surface is as follows : Sulphur, quartz, opal, chalcedony, limonite, wad, calcite, aragonite, siderite, strontianite, barite, gypsum, celestite, fluorite, scorodite, pyrite, realgar, orpiment, cinnabar, stibnite, chabazite, apophyllite and stilbite. CHAPTER VIII RELATIONS OF MINERAL DEPOSITS TO MINERAL SPRINGS The deposition of many valuable minerals can be directly ob- served in nature: limonite, for instance, from the evaporation of water containing iron, or from precipitation in bogs and lakes; sulphur by the decomposition of hydrogen sulphide dissolved in water; residual deposits of limonite, nickel silicates, and pyrolusite by the decomposition of rocks by meteoric waters; com- mon salt and borax by the evaporation of lake waters. A large class of deposits, such as the deep-seated veins containing metals and ores developing near intrusive contacts, we can never hope to observe in nature in the process of formation. Ascending mineral springs are not uncommonly observed in mineral deposits, particularly in those which follow fissures, but caution must be used in attributing a genetic role to these springs. If we find such a spring in a contact-metamorphic deposit or in a vein of deep-seated origin, as a cassiterite vein, it would be unlikely indeed that this spring had anything to do with the formation of the deposit, for it could scarcely be as- sumed that the circulation of underground waters could be maintained in the same path during the many vicissitudes of deep erosion, involving the laying bare of rocks once many thou- sand feet below the surface. The formation of ore deposits usually occupies comparatively short epochs, and the agencies to which they owe their origin are evanescent phenomena. In a rather large class of veins, however, of which we know that they were formed near the surface and in recent geological times, we may look with more confidence for a maintenance of the originating solutions, but even here it is well to investigate carefully; the spring may simply be a water of the upper cir- culation which selected the fissure as a convenient path. The case of Plombieres has already been mentioned (p. 104) and there seems to be little reason to doubt that the quartz- fluorite veins at this place have been deposited by the same hot 109 110 MINERAL DEPOSITS waters which, still issue from the fissures. The Triassic sand- stone, covering the granite in that vicinity, is in part replaced by silica and also contains fluorite and barite. DaubreV cites the frequent occurrence, in the Triassic beds of the Central Plateau and the Vosges, of veins and extensive silicification similar to that at Plombieres. Barite, fluorite, and sometimes galena accompany the quartz. According to Jacquot and Willm, 2 the sodium chloride springs of Bourbon-l'Archambault, at the north end of the Central Plateau region, issue from a fracture in Triassic strata, which contains quartz with galena, barite, and fluorite. Dikes of mica- ceous porphyry (minette?) follow the fissures. The waters have a temperature of 53 C. and the total solids aggregate 3,186 parts per million, of which 1,770 are sodium chloride. Bromine, iodine, fluorine, arsenic, and copper are present, and the saline constituents are attributed to the Triassic and Permian strata. The spring deposits contain earthy carbonates and 0.07 per cent, copper oxide. The springs of Contrexeville, which issue from the Triassic and carry traces of fluorine and arsenic, have a temperature of 11 C., and the salts, among which calcium sul- phate and calcium carbonate prevail, also bear marks of deri- vation from the sediments. The springs of Lamalou, near Montpelier, southern France, have a temperature of 34-47 C. and 1,500 parts per million of total solids; the alkaline carbonates prevail, but they also contain calcium and magnesium carbonates, suggesting an admixture of the meteoric waters. Traces of barium, arsenic, copper, lead, nickel, and cobalt were determined. 3 These springs are believed to be genetically connected with the erup- tion of a neighboring basalt area and stand in close relation- ship to veins containing pyrite, arsenopyrite, and chalcopyrite in a gangue of quartz and barite, the exploitation of which had to be stopped on account of the fear of tapping large volumes of water. Barite is believed to be deposited by the present waters. The sodium carbonate springs at Ems, 4 according to Delkes- kamp, issue from a fissure which forms the extension of a quartz vein and contains chalcopyrite. Basalt occurs in the same vicinity. 1 Les eaux souterraines aux epoques anciennes, p. 151. 2 Les eaux mine"rales de la France, Paris, 1894, p. 107. 3 L. De Launay, Recherche, etc., des sources thermomine'rales, 1892. 4 Verhandl. Gesell. deutscher Nat. u. Aerzte, 1903, 4 2,^first part. RELATIONS TO MINERAL SPRINGS 111 Sandberger and Delkeskamp state that the hot sodium chlo- ride springs of Wiesbaden are closely connected with a quartz vein containing tetrahedrite; veins of barite and calcite are common, as are impregnations of barite ; the latter are attributed to earlier (Tertiary) spring waters. Close connection with ore-bearing veins is also, according to Delkeskamp, indicated by the sodium chloride springs of Kreutz- nach, which issue close to a number of veins containing cal- cite, barite, and fluorite with ores of copper and quicksilver. Here, also, Tertiary strata higher than the springs are impregnated with barite, suggesting a considerable age and a formerly higher point of issue of the springs. The saline constituents of the water are believed to be derived from sedimentary rocks. The evidence presented by Delkeskamp does not suffice to establish direct connection of the springs with the ore deposition, but the widespread occurrence of barite, in close association with strong sodium chloride springs, is assuredly suggestive. Mineral springs with a maximum temperature of 26 C. have been opened at several places in the mines of Freiberg, Saxony, and are described in some detail by Stelzner and Bergeat, 1 but there is little reason to believe that they are genetically connected with the deposits. The same authors describe weak sodium carbonate springs, which were encountered in the veins of Joachimsthal, in Bohemia; their highest temperature was 28 C. This is only about 10 miles from Carlsbad and at a lower level. Posepny has suggested that the waters may be derived from the same general source which supplies the springs at Carlsbad. Finally should be mentioned the hot waters which broke into the copper mine of Bocheggiano 2 at Massa Maritima, Tuscany, at a depth of over 1,000 feet, and which had a temperature of 40 C. They contained from 769 to 2,053 parts per million of total solids, mainly sulphates of calcium and magnesium, with a not- able amount of boric acid. The waters, except for the boron, are of the type of ordinary mine waters and may be of meteoric origin with an admixture of boron from volcanic exhalations. Warm springs have been encountered in the mines of Corn- wall ; and one of them in a tin vein near Redruth is said to have contained much lithium, which is not surprising considering the general distribution of lithium-bearing muscovite in that 1 Die Erzlagerstatten, 2, 1905-06, p. 1227. 2 B. Lotti and K. Ermisch, Zeitschr. prakt. Geol, 1905, pp. 206-247 112 MINERAL DEPOSITS region. It seems difficult to believe that these springs are the remains of the waters which deposited the veins, for the veins were formed at a great depth and under high pressure and tem- perature at a remote geological time. In the Cordilleran region of the United States examples of mineral springs in mineral veins are not so common as might be expected from the coexistence of a late mineralization and pres- ent abundance of thermal waters. One reason for this lack lies probably in the great physiographic changes which in most parts of this region have taken place in relatively late times and which would tend to lower or divert the discharges of the springs. At Silver Cliff S. F. Emmons 1 found issuing from the 2,000-foot level of the Geyser mine a strong spring of sodium carbonate water with free carbon dioxide, yielding small quantities of copper, lead, and zinc; the temperature was 26.5 C. The shaft was sunk to a depth of 1,850 feet in rhyolite tuff; at this depth, at the contact between the tuff and pre-Cambrian gneiss, a vein was found containing galena, zinc blende, tetrahedrite, argentite, etc., in a gangue of calcite, barite, and quartz. The water deposited a calcium carbonate sinter with traces of lead, zinc, copper, nickel, and cobalt. In this instance it is possible that the ascending water may have had a genetic connection with the deposit. At the Comstock lode 2 hot waters were encountered at an early date and have made exploitation difficult. It can scarcely be doubted that these waters stand in causal relation to the vein and it is certain that they now dissolve and precipitate gold and silver, as well as pyrite. The heat of the lode has been attributed to oxidation of pyrite or to kaolinization of the feldspars of the country rock, but Becker has shown that it is clearly due to the ascending waters. Reid 3 has examined the evaporated residue from water collected on the 2,250-foot level of the C. and C. shaft. He found 2.92 milligrams of silver and 0.298 milligrams of gold per ton of solution. This water, which has a temperature of 46 to 81 C., contains 965 parts per million of solids, mostly sulphates of calcium and sodium but including 1 The mines of Custer County, Colorado, Seventeenth Ann. Rept., U. S. Geol. Survey, pt. 2, 1896, p. 461. 2 G. F. Becker, Geology of the Comstock lode, Mon. 3, U. S. Geol. Survey, 1882, p. 230. 3 John A. Reid, The structure and genesis of the Comstock lode, Bull. 4, California Univ. Dept. Geology, 1905, pp. 177-191. RELATIONS TO MINERAL SPRINGS 113 133 parts of silica. This water is assuredly not one of the pure types of ascending waters; its composition is in the main the same as that of the ordinary mine waters and it may be a mix- ture of meteoric mine waters with a very hot ascending water. Particularly convincing of the competency of the ascending "vol- canic" springs to deposits gold and silver-bearing veins are the data given on p. 105 in relation to the Ojo Caliente springs of New Mexico and those of Wagon Wheel Gap in Colorado. To this is added the evidence of the gold and silver-bearing sinters of New Zealand (p. 102) and Steamboat Springs, Nevada (p. 100). The widely cited occurrence at Sulphur Bank, in Lake County, California, 1 is considered to furnish good proof of deposition of cinnabar by hot sodium chloride waters, also heavily charged with boron (analysis on page 61). The springs issue through Quaternary basalt in which cinnabar was deposited with opal as crusts along crevices, sometimes as delicate crystals loosely attached to the walls, or as impregnations of the porous basalt ; the pyrite or marcasite was mostly disseminated in the rock, but occurred also as crusts alternating with cinnabar and opal. Melville found traces of gold and copper in the marcasite. At the surface no cinnabar was observed, but sulphur, derived from the oxidation of H 2 S, was present. A few feet below the surface the cinnabar appeared and continued down to about 300 feet, into the sandstones on which the basalt rested. No quicksilver was found in the water, but no one who has studied the occur- rence has doubted that cinnabar, pyrite, and opal have been pre- cipitated from the water which still ascends in these channels. The gases dissolved in the water consist mostly of carbon dioxide, with hydrogen sulphide, hydrocarbon, nitrogen, and some am- monia. The evidence gains in importance when it is realized that the mineral combination and general mode of occurrence cited are characteristic of the quicksilver deposits of the Coast Ranges. A number of other instances of deposition of cinnabar by as- cending waters are given in Chapter XXIV. 1 G. F. Becker, Geology of the quicksilver deposits of the Pacific slope, Man. 13, U. S. Geol. Survey, 1888, pp. 251-268. Joseph Le Conte and W. B. Rising, The phenomena of metalliferous vein formation now in progress at Sulphur Bank, Am. Jour. Sci., 3d ser., vol. 24, 1882, pp. 23-33. F. Posepny, The genesis of ore deposits, 2d ed., Pub. by the Am. Inst. Min. Eng., 1902, pp. 32-36. 114 ' MINERAL DEPOSITS Summary. There is then, convincing testimony that deposits of quicksilver, antimony, arsenic, gold and silver, may be formed close to the surface by hot ascending waters of the kind related to volcanic phenomena. It is probable, indeed, that the majority of fissure veins which contain notable amounts of gold and silver together with sulphides of the baser metals have been formed by these waters. Of this, more conclusive evidence is yielded by the many water deposited veins which so frequently, like a metallic aureole, surround the areas of igneous intrusive rocks. On the other hand it is certain that warm and even cold waters of the meteoric circulation in non-volcanic regions are likewise competent to form mineral deposits of the baser metals containing oxides and carbonates of iron and manganese, and sulphides of copper, lead and zinc with very small quantities of gold and silver. There is even evidence that such waters may develop deposits of minerals of vanadium, and uranium with radium (Chapter XX, p. 399). The waters most competent to per- form this work appear to be the calcium carbonate solutions and the chloride brines which at the same time contain carbon dioxide and hydrogen sulphides. 1 1 C. E. Siebenthal, Zinc and lead deposits of the Joplin region, Bull. 606, U. S. Geol. Survey, 1915, p. 154. CHAPTER IX FOLDING AND FAULTING 1 FOLDS Sedimentary beds and ore deposits contained in them are often bent, corrugated, and folded in more or less complex manner. Extensive folding is usually effected by horizontal or "tangential" thrust, but minor bends and monoclines (Fig. 7) may originate by thrust in any direction. In extreme cases any fold or bend may pass over into a break or fault. In folding on a large scale it is necessary that the sedimentary complex have FIG. 7. Monocline near Gallup, New Mexico. After E. Howdl. beds of sufficient strength (competent beds) to transmit the thrust and support the structures; if the complex is plastic it will be deformed by flowage and no regular folds will result. Folds are synclinal (Figs. 8 and 9), trough-like; or anticlinal (Fig. 10), shaped like a saddle. A plane which bisects the average angle between the limbs is called the axial plane of the fold. By complex movements the axial plane may become a curved surface. If this axial plane is vertical the limbs dip at like angles; if the axial plane is inclined the limbs have unequal dips. In close folding the limbs dip steeply (Fig. 11). When the axial plane of folds inclines strongly in one direction we speak 1 E. de Margerie and A. Heim, Les dislocations de Pecorce t^rrestre, Ztirich, 1888, pp. 49-63. Bailey Willis, The mechanics of Appalachian structure, Thirteenth Ann. Rept., U. S. Geol. Survey, 1894, pp. 211-281. C. R. Van Hise, Principles of North American pre-Cambrian geology, Sixteenth Ann. Rept., U. S. Geol. Survey, pt. 1, 1896, pp. 589-633. 115 116 MINERAL DEPOSITS FIG. 8. Open syncline showing Carboniferous phosphate beds uncon- formably covered by Eocene beds. Beaver Creek, Utah. After E. Black- welder, U. S. Geol Survey. FIG. 9. Eroded syncline, Georgetown Canyon, Idaho, showing phosphate bed. After H. S. Gale, U. S. Geol. Survey. FEET Above sea-leve -8500 -7500 -6500 -5500 -4-500 FIG. 10. Eroded anticline, Montpelier, Idaho, showing bending of copper- bearing beds of Triassic age. After H. S. Gale, U. S. Geol. Survey. FOLDING AND FAULTING 117 of overturned folds, and these by further compression may easily pass over into overthrust faults (Fig. 12), causing a part of the 1000 2000 3000 *ooo 5000 Feet Upper diabase Lower diabase FIG. 11. Close folding with overthrusts and thickening of strata by duplication. After M. Koch. folded series to slide over the other. In flat overthrust faults the horizontal movement may amount to many miles. 118 MINERAL DEPOSITS Synclines and anticlines extend naturally about perpendicu- larly to the direction of compressive stress. Their direction is in- dicated by a line passing through all the highest or lowest'points of a given stratum. These crest lines or trough lines have usually a distinct dip; the angle of this line with the horizontal is called the pitch of the fold. Minor plications on the limbs FIG. 12. Diagram showing development of an overthrust fault from a fold. After C. R. Van Hise, U. S. Geol. Survey. often indicate the pitch of the fold. Thrusts in two directions result in cross-folding with, the development of bending in forms known as canoes, domes, and basins. When beds are lifted in dome shape so that they dip away from a central point they form a quaquaversal. In sharp folding of a sedimentary complex, the strata become thicker by compression at the points of greatest bending (Fig. 13). The harder strata of sandstone or limestone will yield to tension by breaking or tearing; the softer, shaly strata do not break, but yield to deformation. Strata of differing hardness may slide over one another at such points, and openings may be produced which, for instance, may later be filled with quartz. In slates and crystalline schists which have been deformed at great depths by flowage the harder or "competent" layers, like quartz veins, may be corrugated in an extraordinarily complicated manner. Some quartz veins of Nova Scotia, called "barrel quartz," are believed to owe their form to such conditions (Fig. 14). FOLDING AND FAULTING 119 FIG. 13. Overturned anticline of crystalline limestone, Lenox, Mass., showing thickening and breaking of strata at points of bending. After T. Nelson Dale, U. S. Geol. Survey. FIG. 14. Gneiss with corrugated veinlets of quartz. After C. R. Van Hise, U. S. Geol. Survey. 120 MINERAL DEPOSITS FAULTS^ Sedimentary beds and deposits, as well as deposits of later origin which persistently follow a certain horizon in a sedimen- tary series, are sometimes abruptly cut off by structural planes. b FIG. 15. Faulting of Mendota vein, Silver Plume, Colorado, a, granite; b, quartz; c, galena and zinc blende. After J. E. Spurr, U. S. Geol. Survey 1 E. de Margerie and Heim, Bailey Willis, C. R. Van Hise, op. cit. J. E. Spurr, Geology applied to mining, New York, 1904, pp. 149-163. J. E. Spurr, Geology of the Tonopah mining district, Nevada, Prof. Paper 42, U. S. Geol. Survey, 1905, p. 144. J. E. Spurr, Measurements of faults, Jour. Geology, vol. 5, 1897, p. 723. J. E. Spurr, The relation of ore deposition to faulting. Econ. GeoL, vol. 11, 1916, pp. 601-622. &F. L. Ransom e, The direction of movement and the nomenclature of faults, Econ. Geol., vol. 1, 1906, p. 777. C. F. Tolman, Jr., Graphic solution of fault problems, Min. and Sci. Press, June 17, 1911, et seq. Reprinted, San Francisco and London, 1911. C. F. Tolman, Jr., Econ. Geol, vol. 2, 1907, pp. 506-511. H. F. Reid, Geometry of faults, Bull, Geol. Soc. Am., vol. 20, 1909, pp. 171-196. H. F. Reid, W. M. Davis, A. C. Lawson, and F. L. Ransome, Proposed nomenclature of faults, Bull, Geol. Soc. Am., vol. 24, 1913. C. K. Leith, Structural Geology, New York, 1913. F. H. Lahee, Field Geology, New York, 1916, Chapters VII and VIII. FOLDING AND FAULTING 121 When such an occurrence is found it is safe to conclude that the interruption is due to a fault -that is, to a fracture along which movement has taken place and that the continuation of the bed exists somewhere beyond this break (Fig. 15). FIG. 16. Sketch showing replacement of shale by pyrite. Natural size. The small fissures are older than the pyrite and are crossed by its banded structures. After F. L. Ransome, U. S. Geol. Survey. In the case of epigenetic deposits not closely following the original lines of structure in the rocks, such a conclusion is often, but not 'always justified. The interruption of the ore-body may FIG. 17. Plan of a vein in the Homer mine, Idaho Springs, Colorado, showing deflection of the vein upon meeting a dike. After J. E. Spurr, U. S. Geol. Survey. be due to an actual post-mineral dislocation, or it may be caused by a cessation of mineralization on account of structures existing before the mineralization began. The replacement of 122 MINERAL DEPOSITS limestone by galena or shale by pyrite (Fig. 16) may stop sud- denly at a clay-coated seam, which offered a barrier to the solu- tions. A vein-filled fissure may terminate abruptly or split up within a few feet upon encountering softer and more plastic rocks, such as clay shales, thick gouge seams, or soft tuffs. A vein traversing formations of varying hardness often suffers abrupt deflection at rock contacts. It may also be deflected by encoun- tering older dikes or fissures, either barren or filled with vein material (Fig. 17). The distinction between faults and deviations is most impor- tant. The appearance of detached fragments of the ore usually Scale 2 feet FIG. 18. Sketch of faulted quartz vein in andesite, showing "drag." After J. E. Spurr, U. S. Geol. Survey. termed "drag" on the faulting plane (Fig. 18), the direction of the striations, and the interrelations of dip and strike of fault- ing, fissure, and ore-body are all valuable data which must be interpreted in each case. There are many geometrical rules for the finding of a faulted ore-body, but they are of little value unless the character of the interruption is known. Each case must be considered and judged by itself. FOLDING AND FAULTING 123 Too often faulting is considered in only two dimensions that is, as either normal or reverse movement in a vertical plane. The fact is that most faulting movements have lateral as well as vertical components; every mining engineer knows the frequent occurrence of inclined or horizontal striation on fault planes. Several proposals have been made looking to a uniform nomen- clature of the various elements involved in faulting movements; the best of these are advocated by J. E. Spurr, C. F. Tolman, Jr., and H. F. Reid. Lately a committee of the Geological Society of America has been instructed to examine this question in more detail, and their conclusions, in large part based on the work of Reid, will probably be adopted by American geologists. An abstract of their report 1 will be found in the following pages. Measurements of fault movement are made in the fault plane itself, in a plane normal to the trace of the faulted body. on the fault plane, in any normal plane, and in a horizontal plane. General Terms A fault is a fracture in the rock of the earth's crust, accompa- nied by a displacement of one side with respect to the other in a direction parallel with the fracture. The fracture is usually not an open crack, and an open crack would not be a fault unless one of the sides had moved parallel with the crack relatively to the other. As we pass from one part of a fault to another, we find that certain characteristics vary. Definitions descriptive of charac- teristics must therefore be considered as referring to the parts of the fault to which they are applied and not necessarily to the fault as a whole. A closed fault is one in which the two walls of a fault are in contact. An open fault is one in which the two walls of a fault are separated. The same fault may be closed in one part and open in another. The fault space is the space between the walls of an open fault. A. fault surface is the surface along which dislocation has taken place; this may be called a fault plane if it is without notable curvature. A fault line is the intersection of a fault surface with the 1 Reid, Davis, Lawson, and Ransome, op. cit. 124 MINERAL DEPOSITS earth's surface or with any artificial surface of reference, such as a level of a mine. U When a fault is made up of a number of slips on closely spaced surfaces, the section of the earth's crust containing these minor faults is called the shear zone. This name would also apply to the brecciated zone which characterizes some faults. The fault breccia, or fault rock, is the breccia which is frequently found in the shear zone, more especially in the case of thrust faults. Gouge is a fine-grained, impervious clay, usually a mixture of minerals which sometimes occurs between the walls of a fault. A horse is a mass of rock broken from one wall and caught between the walls of the fault. The fault strike is the direction of the intersection of the fault surface, or the shear zone, with a horizontal plane. It is measured from the astronomic or from the magnetic meridian. The fault dip is the vertical inclination of the fault surface, or shear zone, measured from horizontal plane. The hade is the inclination of the fault surface, or shear zone, measured from the vertical; it is the complement of the dip. The hanging watt is the upper wall of the fault. The foot watt is the lower wall of the fault. General Classification of Faults Faults of parallel displacement are those in which all straight lines on opposite sides of the fault and outside of the dislocated zone, which were parallel before the displacement, are parallel afterward. Rotatory faults are those in which some straight lines on oppo- site sides of the fault and outside of the dislocated zone, parallel before the displacements, are no longer parallel afterward that is, where one side has suffered a rotation relative to the other. Determinations of throw are almost always relative, and hence we can rarely tell which side of the fault has moved; therefore the terms "upthrow" and " downthrow," which are used accord- ing to the side from which the fault is viewed, are objectionable, as they suggest that a particular side of the fault has actually been moved. They are in very general use and should be retained, but it should be definitely understood that they refer merely to a relative and not to an absolute displacement. FOLDING AND FAULTING 125 Faults of Parallel Displacement. No faults of any magnitude consist of simple parallel displacements over their whole length. Faults die out at their limits, and the displacement is not uni- form along their courses, so that there is necessarily some slight rotation, varying in amount in the different parts of the fault's course. Probably the greatest number of faults, certainly of large faults, are of this character. The variations in rotation and displacement are permitted by slight plastic deformation. If, however, we confine our attention to a small length of the fault, we may describe the displacement there as if the rock were rigid; and if the rotation is very small, we may describe it as if a parallel displacement had occurred. It sometimes happens that the strikes on the opposite sides of a fault are different; the strata are then said to "strike at each other." This suggests a rotation, but it may be due to local variation of strike or to an unconformity. The word "displacement" should receive no technical mean- ing, but is reserved for general use; it may be applied to a relative movement of the two sides of the fault, measured in any direction, when that direction is specified; for instance, the displacement of a stratum along a drift in a mine would be the distance between the two sections of the stratum measured along the drift. The word "dislocation" will also be most useful in a general sense. There are two methods of defining the displacement due to a fault; we may define the apparent relative displacement of a bed by naming the distance between its two disrupted branches measured in any chosen direction, such as the vertical distance between the branches, measured in a shaft, or the perpendicular distance between the lines of intersection of the two branches with the fault plane; or we may define the actual relative dis- placement and its components in important directions. The ap- parent displacements are those usually measured; the actual displacement must be worked out for a complete understanding of the fault. Only four important technical words are used to denote the various displacements caused by faulting, qualifying words being added to indicate the component of the displacement referred to. These words are: Slip, which indicates the relative displacement of formerly adjacent points on opposite sides of the fault, measured in the 126 MINERAL DEPOSITS fault surface. The qualifying words relate to the strike and dip of the fault surface. Shift, which indicates the relative displacement of regions on opposite sides of the fault and outside of the dislocated zone. The qualifying words relate to the strike and dip of the fault surface, except in the expression "vertical shift," which is self- explanatory. Throw, which indicates a displacement not related to the strike or dip of the fault plane. Offset, which indicates the horizontal distance between the outcrops of a dislocated bed. By keeping in mind the general meaning of these four words, all confusion in the uses of the proposed nomenclature can be avoided. There is no generally accepted word in present use to denote the slip. Willis and Tolman use "displacement;" Spurr uses "throw." We have reserved "displacement" for general use, and the word "throw" is here used in quite a different sense. The word "shift" also suggests the meaning attached to it; there is no distinctive word now in use to describe the shift. In mines, where the fault surface itself its visible, the slip will generally be determined; it is of paramount importance in min- ing. In field surveys, where the fault is studied by the dislocation of the outcrop of strata, or dikes, often at a considerable distance from the fault, the shift is determined. In the larger problems of geology the shift is of greater importance than the slip. The dis- tinction between the slip and the shift is important; it has not heretofore been recognized in the nomenclature of faults. The perpendicular throw is of the greatest importance. It is fre- quently the only displacement determined, and in strike faults all the displacements in a vertical plane at right angles to the fault strike that is, all the displacements which have heretofore received the most attention can be expressed in terms of perpendicular throw. The offset is often the most important surface measurement made. Faults in Stratified Rocks. Among stratified rocks the character of the displacement of the strata due to a fault is so much influenced by the relation of the strike of the fault to the strike of the strata that syecial subclasses are generally recognized. A strike fault is one whose strike is parallel to the strike of the strata. FOLDING AND FAULTING 127 A dip fault is one whose strike is approximately at right angles to the strike of the strata. An oblique fault is one whose strike is oblique to the strike of the strata. These terms are, of course, not directly applicable in regions of unstratified rocks; but they might be used in such regions with respect to the strike of a system of parallel dikes if this were distinctly stated in the description of the faults. Similarly with regard to the general structure of the region: A longitudinal fault is one whose strike is parallel with the general structure. A transverse fault is one whose strike is transverse to the gen- eral structure. 1 Slip. The word " slip " indicates the displacement as measured on the fault surface; the qualifying words refer to the strike and dip of the fault. The slip or net slip is the maximum relative displacement of the walls of the fault, measured on the fault surface, along the line of the movement; it is given by ab in Figs. 19 and 20. 2 FIG. 19. The slip. FIG. 20. The shift. The strike-slip is the component of the slip parallel with the fault strike or the projection of the net slip on a horizontal line in the fault surface; ac in Figs. 19 and 20. 3 The dip-slip is the component of the slip parallel with the fault dip, or the projection of the slip on a line on the fault surface perpendicular to the fault strike; be in Figs. 19 and 20. 4 Shift. It frequently happens that a fault has not a single sur- face of shear, but consists of a series of small slips on closely 1 See the word "flaw" further on. Spurr and Tolman call this the "total displacement." 2 Tolman calls this the "horizontal displacement." Tolman calls it the "normal displacement." 128 MINERAL DEPOSITS spaced surfaces, and in some faults the strata in the neighbor- hood of the fault surface are bent, so that the relative displace- ments of the rock masses on opposite sides of the fault may be quite different from the slip and not even parallel with it. The word "shift" is used to denote the relative displacements of the rock masses situated outside of the zone of dislocation; the quali- fying words relate to the strike and dip of the fault, with one ex- ception, in which the meaning is clear. The shift, or net shift, is the maximum relative displacement of points on opposite sides of the fault and far enough from it to be outside of the dislocated zone; de in Figs. 20 and 21, where d is the position of a selected point before and e after the faulting. The strike-shift is the component of the shift parallel with the fault strike; df in Figs. 20 and 21. FIG. 21. The shift. FIG. 22. The throw. The dip-shift is the component of the shift parallel with the fault dip; fe in Figs. 20 and 21. (The triangle def is parallel with the fault surface in Fig. 20. *) The bending of the strata near the fault may be so great that the direction of the shift is no longer even nearly parallel with the fault surface; it is better then to use the three following terms for the components of the shift: The strike-shift is the horizontal component of the shift parallel with the fault strike, as already defined. The normal shift is the horizontal component of the shift at right angles to the fault strike. It equals the horizontal shortening or lengthening of the earth's surface at right angles to the fault strike, due to the fault. The vertical shift is the vertical component of the shift. These 1 The dip-shift and strike-shift are not accurately shown in Fig. 20, because the net shift, de, is not parallel with the fault plane, and the lines de, df, and fe would not lie in one plane. But the definitions are clear and the figure illustrates them fairly well. FOLDING AND FAULTING 129 terms may evidently be used equally well when the shift is parallel with the fault plane. Throw. The word "throw" will apply to components of the displacement having no immediate bearing on the strike or dip of the fault plane. The throw is the vertical component of the slip; eg in Figs. 20 and 22, de in Figs. 23 and 24. The word is almost universally used in this sense, but A. Geikie uses it to designate the vertical distance between the two parts of a dislocated bed, projected if necessary a very different thing. Geikie's "throw" would be represented by df in Figs. 23 and 24. Spurr uses "throw" to designate the distance between the two parts of a dislocated bed measured on the fault plane. FIG. 23. Section of a normal fault. FIG. 24. Section of a reverse fault- The heave 1 is the horizontal component of the slip, measured at right angles to the strike of the fault; bg in Figs. 20 and 22, eg in Figs. 23 and 24. The word "heave" has been used in many senses; J. Geikie, Willis, Scott, and Fairchild use it as denned above; A. Geikie and Spurr use it to designate what we have called the "offset" of a bed (see below); Jukes-Brown apparently used it for the strike-slip (De Margerie and Heim, page 72) ; so did Ransome; and Scott also uses it in this sense when he re- fers to "heave faults." The perpendicular throw of a bed, dike, vein, or of any recog- nizable surface, is the distance between the two parts of the disrupted bed, etc., measured perpendicularly to the bedding plane or to the plane of the surface in question. It is measured, therefore, in a vertical plane at right angles to the strike of the disrupted surface. 2 The importance of the perpendicular throw 1 Sometimes called the"horizontal throw." 2 Spurr calls it the "perpendicular separation." Tolman's "perpendicu- lar throw" would under certain conditions correspond in meaning with our expression. 130 MINERAL DEPOSITS of the strata is so great that it is convenient to have special terms for it; these are given below. The stratigraphic throw is the distance between the two parts of a disrupted stratum measured at right angles to the plane of the stratum; oh in Figs. 23 and 24. The stratigraphic throw is in general the simplest throw to determine; it can be found from the distance between the outcrops of the two parts of the same stratum, the dip of the stratum, and the slope of the ground. The dip throw is the component of the slip measured parallel with the dip of the strata; cb in Figs. 23 and 24. The throws have been defined as components of the slip. Where we are dealing with a simple fault in plane strata, the shifts will be the same as the slips, and the term throws will apply to both equally well; it is only in plane strata that the per- pendicular throw is important. Where there is a dislocated zone about the fault, the term perpendicular throw would necessarily apply to the shift; but we cannot detach the word throw from its accepted meaning and apply it generally to the shift. FIGS. 25 and 26. Plan of an oblique slip. Offset. The offset of a stratum is the distance between the two parts of the disrupted stratum measured at right angles to the strike of the stratum and on a horizontal plane. 1 If Figs. 25 and 26 represent the ground plans of oblique faults on a level surface, ab, and not ac, would be the offset of the stratum; ac would be the horizontal displacement of the stratum parallel with the fault strike. Some confusion of nomenclature results from the non-observ- ance of the fact that the distance between the dislocated parts of a stratum measured in a certain direction is not the same as 1 A. Geikie and Spurr use the term "heave" for this offset. FOLDING AND FAULTING 131 the component of the slip in the same direction; for instance, let Fig. 27 represent a reverse strike dip-slip fault 1 in section. A. Geikie calls ad the throw and ef the heave, whereas the most general usage seems to be to call ac the throw and be the heave, as adopted above. The distance ad has not been defined, but it FIG. 27. Plan a reverse fault. is readily described as the vertical displacement of the stratum, without limiting the word "displacement" to a technical mean- ing; ef is the offset. Let Figs. 28 and 29 lie in the fault plane and let the point a move by faulting to c, then ac will be the net FIGS. 28 and 29. Section in a fault plane. slip, ad the strike-slip, cd the dip-slip, and ab the displacement of the stratum parallel with the fault strike; ab is not necessarily at right angles to the strike of the strata. Faults Classified According to the Direction of the Movement. Faults may be classified, according to the direction of the move- ment on the fault plane, into three groups, as follows: 1 This means a reverse fault whose strike corresponds with the strike of the strata and in which the displacement has been in the direction of the fault dip. 132 MINERAL DEPOSITS Dip-slip faults, where the net slip is practically in the line of the fault dip. Strike-slip faults, where the net slip is practically in the direc- tion of the fault strike. J. Geikie calls such faults "transcurrent faults," or "transverse thrusts." Jukes-Brown calls them "heaves." Scott calls them "horizontal faults," or " heave faults." A vertical fault is one with a dip of 90 degrees (see below) ; and, by analogy, a horizontal fault should be one with a zero dip and the term should not be applied to strike-slip faults. Oblique-slip faults where the net slip lies between these directions. Classes of Strike Faults. Most geological text-books and books on field methods have confined themselves almost exclu- sively to the discussion of dip-slip faults, and have given little attention to the other two classes. Strike faults have usually been treated as if they were also dip-slip faults and classified into Normal faults, where the hanging wall has been depressed relatively to the foot wall. Reverse faults, where the hanging wall has been raised relatively to the foot wall. Vertical faults, where the dip is 90 degrees. The relative displacement has usually been determined by means of a dislocated bed. Although exception may well be taken to these terms, their retention is recommended, because they are in general use and are well understood. The word "reverse" is preferable to "reversed" (which has been almost universally used), as the latter implies the reversal of an action. The horizontal distance between two points on opposite sides of a fault, measured on a line at right angles to the fault strike, is always shortened by a reverse strike fault, lengthened by a normal strike fault, and unchanged in length by a vertical fault. It can be shown that normal faults may be formed without the existence of tension and indeed under some pressure, but the definitions we are giving are geometric and not dynamic. Extension of the Words Normal and Reverse to Diagonal and Dip Faults. The expressions "normal" and "reverse" may be used in connection with oblique and dip faults, even when these are strike-slip or oblique-slip faults, provided they are applied to designate the apparent relative displacement of the two parts of a dislocated stratum, or other recognized surface, in a vertical FOLDING AND FAULTING 133 plane at right angles to the fault strike. It does not follow, in the case of oblique-slip faults, that a horizontal line at right angles to the fault strike would be lengthened by a normal or shortened by a reverse fault. This has been pointed out by Ransome 1 and can be illustrated by Figs. 30 and 31. In Fig. 30 a reverse fault, as determined by the displacement of the stratum, has caused an extension at right angles to the fault strike. It is evident that if the hanging wall had moved, as in Fig. 31, with the stratum dipping as there represented, we should have had a normal fault and a contraction at right angles to the fault strike. The relations of the two parts of the disrupted stratum in Fig. 30 are exactly the same as if we had had a simple reverse FIG. 30. A reverse fault due to FIG. 31. A normal fault due to an an oblique slip. oblique slip. dip-slip fault, and in Fig. 31 as if we had had a simple normal dip-slip fault ; and if there are no disrupted dikes or other means of determining the amount of the strike-slip, the movements described could not be distinguished from simple dip-slip faults. 2 It very frequently happens that nothing more than the apparent displacement of the strata can be determined, and we recommend that the terms "normal" and "reverse" faults as denned be used purely for purposes of description and not for the purpose of indicating extension or contraction, tension or compression, vertical or horizontal forces. Special Classes of Faults. There are two kinds of faults which have played such important roles in altering the structure of some regions that they have received special names. *Econ. Geol. vol. 1, 1906, pp. 783-787. 2 The methods of determining the complete displacement at a fault are given in Reid's Geometry of faults, Bull. Geol. Soc. Am., vol. 20, 1909, pp. 170-196, and in Tolman's Graphic solution of fault problems, op. cit. 134 MINERAL DEPOSITS Overthrusts. These are reverse faults with low dip or large hade. In some cases the dip-slip has been enormous, amounting to tens of kilometers. Scott calls them "thrusts" and separates them entirely from faults of high dip; but the word "thrust" has been used for ordinary reverse faults of high dip. The word "overthrust" has been generally used for this kind of fault and is very descriptive. It should be adopted. Flaws. Suess has described with care certain faults in which the strike is transverse to the strike of the rocks, the dip high and varying from one side to the other in the course of the fault, and the relative movement practically horizontal and parallel with the strike of the fault. He has used the word "Blatt" (plural, "Blatter") to designate them. Miss Sollas has used the word "flaw" in the English translation of Suess. The gold-quartz veins of the Tauern in Austria, investigated by Posepny, follow such dislocations. FIG. 32. Vertical section of a faulted vein, Berlin mine, Nevada, showing also its probable original position. After Ellsworth Daggett. Rotatory Faults. When a rotation of one side of the fault occurs, the amount of the rotation and the direction of the axis should be given. Rotatory faults have been but little studied, and it is not considered advisable to suggest a special nomen- clature at present. Mineralization of Faults. Any fault may become a fissure vein by filling and replacement along its course. However, it is rather unusual to find large structural faults, normal or over- FOLDING AND FAULTING 135 thrusts, which have been extensively mineralized. Shear zones, sheeted zones, and "flaws" (Blatter) often result in veins or lodes. Complexity of Faulting. During mining operations excellent and detailed instances of the complexity of faulting are often found. Normal and reverse faults may occur in close proximity . A fault consists more frequently of a series of closely spaced breaks than of a single fracture. Displacement occurs usually along each of these breaks, the result being a distortion of the deposit within the faulted zone. Fraction_No_[ shaft- FIG. 33. Horizontal plan showing faulted vein, Tonopah, Nevada. Scale 50 feet to one inch. After J. E. Spurr, U. S. Geol. Survey. Fig. 32 shows a case of complicated normal faulting from the Berlin vein, Nevada. 1 Besides the faults indicated there are a great number of other dislocations with horizontal dis- placement. The deposit is a filled quartz vein, 2 to 3 feet wide, carrying 2 per cent, of sulphides with silver and gold. The great complications ensuing where faulting takes place along two intersecting fault systems have been described by Spurr 2 in his report on the Tonopah district, Nevada. The 1 Ellsworth Daggett, The extraordinary faulting at the Berlin mine, Eng. and Min. Jour., Mar. 30, 1907. 2 Prof. Paper, 42, U. S. Geol. Survey, 1905. 136 MINERAL DEPOSITS result of such structures is likely to be a zigzag distribution of the fragments of the faulted vein with an average movement determined by the two components. Repeated small disloca- tions practically result in a deflection of the vein (Fig. 33). Overthrusts of great magnitude, such as are found in the Alps, may have had most important results as to the continuation in depth of ore deposits. As these dislocations may be measured in miles, it follows that whole groups of deposits contained in the overthrust portion of the strata may have been cut off entirely from their continuation in depth. 1 The relations of ore deposits to dynamic metamorphism is described in Chapter XXX. 1 B. Granigg, Ueber die Erzfuehrung der Ostalpen. Leoben, 1913. CHAPTER X OPENINGS IN ROCKS Chemical processes and alteration in general may go on in a rock without cavities other than pore space and capillary or sub-capillary openings. Such processes are, however, metamor- phic rather than metasomatic ; they simply effect a mineralogical rearrangement without much chemical change; the composi- tion of the rock remains constant. The formation of epigenetic mineral deposits usually implies a considerable addition of for- eign material by solutions and these solutions must be guided to the place of deposition by open spaces, such as fissures, joints, or cracks. As a matter of fact the great majority of mineral de^ posits were formed where the path of the solution was prescribed by openings in the rocks other than those of ordinary pore space. After the solutions have gained access to the rock they may of course enter the pores and capillary openings and effect metaso- matic changes. The sizes of capillary and subcapillary openings are given on p. 30. The discussion which follows relates mainly to openings of supercapillary size. Such openings are chiefly found in the zone of fracture (p. 72). Few of our mineral deposits have been formed at depths much greater than 15,000 feet. Small openings may, however, exist in hard rocks at a distance below the surface much greater than the figure just indicated (p. 73). The possi- bility is, therefore, shown that solutions from great depths may gain access to the upper zone of fracture. ORIGIN OF OPENINGS Rock cavities may originate in various ways : 1. By the Original Mode of Formation of the Rocks. Many volcanic flows contain abundant gas pores, or blow holes produced by the expansive force of gases escaping from the magma. Zeolites and calcite, sometimes with native copper, often accumu- late in these pores, and such rocks are usually termed "amygda- loids" and the filled cavities "amygdules" (Fig. 34). Some 137 138 MINERAL DEPOSITS sandstones and conglomerates contain much pore space in which solutions may deposit ores or other substances. 2. By Solution. Solution cavities are found mainly in easily soluble rocks, such as limestone, dolomite, gypsum, and salt. Posepny justly maintains that the solvent power of water suffices to produce long galleries or passages in rock salt and mentions several examples. 1 Joints in limestone are often irregu- larly enlarged by solution and when subsequently filled with ores FIG. 34. Photomicrograph of basalt showing blowholes filled with chlorite, calcite and native copper. Black areas represent copper. After Volney Lewis. such cavities are known as gash veins or pipe veins. Caves in limestone are likewise made by atmospheric water of the upper circulation, containing dissolved carbon dioxide. Such caves are generally formed above the ground-water level in the zone of oxidation, though cases are known which suggest that the process 1 Genesis of ore deposits, 1902, p. 20. York. Pub. by Am. Inst. Min. Eng., New OPENINGS IN ROCKS 139 can go on also below this level. Caves occur in all limestone regions and are sometimes of enormous extent; the Mammoth Cave of Kentucky has passages more than 40 miles in length and has been formed by the removal of millions of cubic yards of rock. The extent of caves is generally determined by faults and disloca- tions, and rock openings on a smaller scale are usually determined by the prevailing joint systems. The breaking in of caves near the surface produces the "sink-holes" so characteristic of certain limestone plateaus. Both caves and sink-holes have a certain importance in the origin of the class of zinc-lead deposits common to many limestone areas, and caves of dissolution in the oxidized part of ore deposits in limestone are sometimes the receptacles for a great variety of secondary minerals. The floors of caves are usually covered with red "cave earth," a residual deposit of silica, kaolin, limonite, etc., derived from the less soluble con- stituents of the limestone. Deposits of bat guano and nitrates are sometimes found in caves. Small solution cavities are often found in more resistant rocks that have been exposed to hot solutions of great solvent power. 3. By Fractures of Various Modes of Origin, (a) Contraction Joints Produced by Tensile Stress in Igneous Rocks. When mag- mas congeal to igneous rocks tensile stresses which result in fis- sures and joints are developed. This is best exemplified in effu- sive rocks, which often show regular columnar structure and which are always full of irregular joints and cracks. No doubt these open spaces may guide metal-bearing solutions. In the literature many authors attribute fissure veins in effusive rocks to contraction, but usually without sufficient reason. The tensile stresses cannot produce long fissures with regular strike and dip. According to the views of many geologists, smaller irregular veins in dikes or other -intrusive rock masses fill contraction fissures. This explanation has been advanced for the hori- zontal tin-bearing joints in the Zinnwald granite, Saxony, and for other similar "stockworks;" also for the so-called "lad- der veins," which are short transverse fissures in dikes, usually extending only from wall to wall. Well-known examples of this kind in Telemarken, Norway, 1 have been described by Vogt; in Victoria, Australia, 2 by Whitelaw; and at Beresowsk, in the 1 Zur Klassification der Erzvorkommen, Zeitschr. prakt. Geol., 1895, p. 149. 2 Mem., Geol. Survey Victoria, vol. 3, 1905, p. 11. 140 MINERAL DEPOSITS Ural Mountains, 1 by Rose, Helmhacker, Karpinsky, Posepny. and Purington. In places, however, the transverse fissures may extend over the contact into the wall rock or correspond to the general joint systems of the vicinity, a fact which throws some doubt on the correctness of the explanation given (Fig. 35). (6) Contraction Joints by Shrinking of Limestone when Changed to Dolomite. Dolomite is not uncommonly formed near certain metal deposits and it is possible that this process when carried on by rapidly moving solutions and in comparatively free space may result in openings suitable as receptacles for ore minerals. FIG. 35. Section of Morning Star dyke, Woods Point, Victoria, showing ladder veins. After 0. A. L. Whitelaw. (c) Expansion Joints Produced by Increase of Rock Volume. Peridotite upon change to serpentine near the surface and near fissures is believed to increase its volume greatly and such serpen- tine often breaks into smooth fragments. Extreme irregularity is a characteristic of all expansion joints and they are of little importance in ore deposition. 1 Guide, Seventh Int. Geol. Congress, 1897, p. 42. F. Posepny, Archiv fur prakt. Geologie, vol. 2, 1895, p. 499. C. W. Purington, Eng. and Min. Jour., June 13, 1903. OPENINGS IN ROCKS 141 (d) Fissures Produced by Torsional Stress. The celebrated ex- periment by DaubreV carried out by twisting a thick glass plate has shown that torsional stress may result in several systems of long and radiating fissures. This experiment has frequently been cited by geologists to explain divergent vein systems, but G. F. Becker has pointed out that such fissures do not follow approxi- mate planes, like fissure veins, but are decidedly curved and warped. Becker 2 regards torsional stress as a system of tensions. FIG. 36. Section through a saddle reef, Bendigo, Victoria. A, Sandstone; B, shaly sandstone; C, gold-bearing quartz. After T. A. Richard. (e) Openings Produced by Folding of Sedimentary Rocks. The bedding planes of sediments are primary structures which often serve as ducts for metal-bearing solutions. Better passageways for such solutions are provided when a series of sediments of un- equal resistance is folded. A sandstone, for instance, will accom- modate itself to bending with difficulty and will easily break at 1 Etudes synthetiques de geologic experimentale, Paris, 1879, p. 316. 2 The torsional theory of joints, Trans., Am. Inst. Min. Eng., vol. 24, 1894, pp. 130-138. 142 MINERAL DEPOSITS anticlines or synclines, whereas softer shales will bend without breaking; the same process may cause a slipping between the vari- ous members. Such tensional stresses may then easily produce open cavities. The quartz-filled so-called "saddle reefs" of the gold mines of Bendigo and other places in Victoria are believed to have been formed in this maner by tensile stresses, but they are also accompanied by irregular masses or "makes" of quartz which fill spaces of discission across the beds (Figs. 13 and 36). (f) Openings Produced by Shearing Stress under the Influence of Gravity. In many disturbed regions the rocks are broken by normal faults along which the various blocks have settled down under the influence of gravity. Such normal faulting is espe- cially characteristic of regions which do not bear evidence of strong compressive stress. Step-faulting is common and friction breccias and crushed zones frequently follow the faults; the open spaces, more or less continuous, offer good paths for the circula- tion of water if above the "level of discharge;" the fault planes are often long and regular. But in spite of all this, mineral de- posits, except some of purely surface origin, are not common along such faults. At Clifton, Arizona, for instance, faults are abundant, but the copper deposits do not ordinarily appear in them. There are exceptions, however. At Creede, Colorado, a gold-silver vein occupies an important fault fissure, and similar cases are known from the silver-lead veins of the Harz Mountains in Germany. In volcanic regions, such as Silverton and Cripple Creek (Fig. 37), in Colorado, systems of nearly vertical fissure veins contain rich deposits. They have obviously little connection with the main structural features of the country, the dislocations are usually small, and the veins were formed shortly after the close of volcanic activity. F. L. Ransome 1 believes that these fissure systems were generated by stresses resulting from slight vertical movements or settling, following an enormous transfer of volcanic material from an intratelluric to a superficial position. Vertical upthrusts of underlying magmas may have caused faulting accompanying or following vein formation. 2 (g) Openings Produced by Compressive Stress.- In contrast to recently congealed lavas, the rocks which have formerly been far 1 F. L. Ransome, Bull. 182, U. S. Geol. Survey, 1901, p. 66. 2 J. E. Spun-, Relation of ore deposition to faulting, Econ. Geol., vol. 11, 1916, pp. 601-622. OPENINGS IN ROCKS 143 below the surface of the earth but which have been exposed by erosion are usually traversed by more or less regular joint systems, persistent over large areas. While some of these joint systems may be caused by the inherent texture of the rock, they are in Scale of Feet 1000 2000 3000 FIG. 37. Plan of the principal veins of the Cripple Creek district, Colo- rado, showing a roughly radial distribution. G, granite and gneiss; V, tertiary volcanic rocks. After Lindgren and Ransome, U. S. Geol. Survey. most cases the effect of compressive stress. Closely spaced joint systems form transitions into slaty cleavage, and recrystallization of minerals takes place by preference along these planes. In ex- 144 MINERAL DEPOSITS treme cases fissility or cleavage in very thin laminae develops. Joints and cleavage present narrow paths for mineralizing solu- tions and ore deposits are often determined by their direction. There are all transitions from joints to fissures along which perceptible movement has occurred. In many districts the fissures which have received the ores are identical in strike and dip with the joint systems of the country rock. A common condition is that two sets of veins and joints occur which have the same strike, but dip in opposite directions (Fig. 38). Such vein systems are termed conjugated fractures. The explana- tion of such joints, fissures, and occasionally accompanying schistose structure is furnished by certain experiments by Dau- breV and by the mathematical deductions of G. F. Becker. 2 FIG. 38. Vertical section of a conjugated system of fractures. These show that compression develops two systems of fractures along the planes of maximum shear; these shearing planes are inclined to the direction of maximum stress. The accompanying dislocations will largely be reverse faults in which the hanging wall has relatively moved upward. In Daubree's experiment on a mass of beeswax and resin two conjugated systems of joints and fissures were formed, making an angle of about 45 with the line of pressure; similar results have been obtained by testing cubes of building stones. If the stress is not exerted hori- zontally the dip of the veins will be correspondingly affected. At Grass Valley, California, and in many other districts there are two such conjugated systems of fissures which have been filled with ore. 1 fitudes synthetiques de geologie experimentale, Paris, 1879, p. 316. 2 Finite homogeneous strain, etc., Bull., Geol. Soc. America, vol. 4, 1893, p. 13. C. K. Leith, Structural Geology, New York, 1913, p. 16. OPENINGS IN ROCKS 145 In the locality just mentioned the majority of the dislocations are small, but tangential stresses sometimes produce great dis- locations. The Mother Lode of California, a vein system nearly 100 miles in length, is believed to represent a reverse fault or system of faults with considerable throw. When rocks are recrystallized in the deeper zones of the earth's crust they may become so plastic that deformation by rupture cannot take place. The growth of crystals then probably takes place predominantly in a plane perpendicular to the stress and a close schistose structure like that in many gneisses may de- velop which offers scarcely any interstitial space available for the circulation of solutions. If the fissures were perfect planes it would be difficult to con- ceive of open spaces along them except by tensional stresses pull- ing the walls apart; but as they are not, movement along them tends to produce a series of openings, alternating with numerous touching points. As a matter of fact the mode of mineral deposition shows that open spaces existed and that they some- times were large, in exceptional cases even 20 feet or more in width. In mine workings in hard rock old stopes frequently remain open for an indefinite length of time, and it is probable that such large open spaces may exist down to a depth of at least several thousand feet. Moreover, it is to be remem- bered that at the time of deposition the fissures were filled by water under a pressure at least equal to that of the hydro- static column. The depositing solutions emanating from magmas under conditions of far stronger pressure may even have made way for themselves in the manner of an igneous dike or pegma- tite vein, actually forcing the rocks apart. Some of the phe- nomena of deep-seated veins are difficult to explain on any other assumption. Gaping fissures are not, however, necessary for the circula- tion of solutions. Water may ascend along a number of closely spaced fissures usually called a sheeted zone in which very little open space exists. But in this case mineral deposition is usually effected by replacement. The solutions are forced into the adjoining rock and transform its minerals into ore. The stresses set up in a mass consisting of various rocks are extremely complex and it may only be possible to ascertain the dominant mode of fracturing. 146 MINERAL DEPOSITS Force of Crystallization. Minerals crystallizing from solutions exert a certain pressure on the walls which confine them. 1 Many geologists have held that this force is sufficient to enlarge cavities along fractures and thus make room for mineral deposits. There is strong evidence in the structure and texture of veins which is unfavorable to such a view, except where conditions of light load prevail as near the surface or near open spaces. Some curious phenomena in regard to inclusions of rocks in veins may find their explanation by the action of this force, for in fissures filled with solutions a comparatively slight force might suffice to detach fragments from the walls. 1 G. F. Becker and A. L. Day, The linear force of growing crystals, Proc., Washington Acad. Sci., vol. 7, 1905, pp. 282-288. S. Taber, Pressure phenomena accompanying the growth of crystals, Proc. Nat. Acad. Sci., vol. 3, 1917, pp. 297-302. S. Taber, Am. Jour. Sci., 4th ser., vol. 41, 1916, p. 535. CHAPTER XI THE FORM AND STRUCTURE OF MINERAL DEPOSITS The form of ore deposits is always important, for the mining methods used for a body of irregular outline must, for instance, be very different from those for a tabular vein. In the great majority of deposits the form is rudely tabular, for they usually follow the planes of dislocations or tabular dikes or the bedding of sedimentary rocks. Great weight was formerly" attached to the form, both in empirical classification and in genetic interpre- tation. At present the tendency is to regard form as largely acci- dental, and to place more emphasis on the mineral association. A convenient and fundamental though not strictly genetic classification divides mineral deposits into syngenetic, or those formed by processes similar to those which have formed the en- closing rock and in general simultaneously with it; and epigenetic, or those introduced into a pre-existing rock. Syngenetic Deposits. The syngenetic deposits include the mag- matic segregations or accumulations of useful minerals formed by processes of differentiation in magmas, generally at a con- siderable depth below the surface. Their form may be wholly irregular or roughly spherical, but more often they are rudely tabu- lar or lenticular, and they are usually connected by transitions with the surrounding rocks. They are either wholly enclosed in the igneous mass, or lie along its margins, or, in some cases, form dikes or offshoots from a deep-seated reservoir. The last class of ores may be called epigenetic with reference to the rocks incasing the dikes. The width and thickness of these deposits may range from a few inches to several hundred feet, and in rare cases, their length may exceed one mile. Masses of chromite in peridotite or titanic iron ore in anorthosite, furnish examples of this type. The syngenetic deposits also include sedimentary beds; they have, as a rule, a tabular or sheet-like form; they are horizontal 147 148 MINERAL DEPOSITS if not disturbed, but are frequently folded and faulted. Parallel to their bedding their extent may be counted by miles, as in the case of the Clinton hematite ores of the Appalachian region, or the French and German limonite beds; nevertheless, each bed usu- ally thins out in wedge-shaped form and may be replaced by others at a slightly different horizon. In deposits of metallic ores the thickness is rarely more than 20 feet and this may include intercalated beds of barren material. Coal beds, especially those of lignite, or brown coal, may attain a thickness of 100 feet or more. Beds of rock salt, anhydrite, and gypsum are in some cases several hundred feet thick. In all sedimentary deposits displacements and folding may locally produce an appearance of great thickness. In plastic material like rock salt such deformation is especially effective. Epigenetic Deposits. The epigenetic deposits have various forms, but among those which follow fissures the tabular or sheet-like form is most common. Deposits concentrated in the zone of weathering are often extremely irregular and of limited extent, and several of them are usually found in close proximity. Some hematite ores, like those of the Mayari district in Cuba, which are developed by the weathering of serpentine, may form superficial sheets of great extent. Replacement deposits in limestone are extremely irregular, although their form as a whole is often dependent upon the bedding, the fissuring, or the contact with other rocks. They are seldom large, but in a few cases, like the galena deposits in southeastern Missouri or the zinc blende deposits in the Joplin region of the same State, they may be followed at a general horizon for several miles. The ore deposits in metamorphic rocks which have undergone strong mechanical deformation and chemical changes usually as- sume lenticular form, and the occurrence of successively overlap- ping lenses is particularly characteristic. In these deposits a steep dip is a common feature,- but the main trend of the ore-body in the plane of its strike is usually not in the direction of the dip. The strike of a tabular or lenticular deposit is the direction of a horizontal line in the plane of the deposit, measured with reference to a meridian. The dip is measured by the vertical angle between a horizontal plane and the plane of the deposit. Complementary to the dip is FORM AND STRUCTURE OF MINERAL DEPOSITS 149 the hade or underlie, which is measured by the angle between the vertical and the plane of the deposit. The plunge 1 (Fig. 39) of an ore-body is the vertical angle between a horizontal plane and the line of maximum elongation of the body. In lenticular ore-bodies in metamorphic rocks which have undergone strong mechanical deformation, the plunge is an important factor, and often it is determined by the direction of the cleavage or schistosity. In fissure veins the pitch of the ore shoot is usually defined as the angle between its axis and the strike of the vein, and it is measured on the plane of the vein 2 (p. 184). FIG. 39. Stereogram illustrating strike, dip, pitch and plunge of an ore-body. Spacial Relations of Veins. Veins are tabular or sheet- like masses of minerals occupying or following a fracture or a set of fractures in the enclosing rock; they have been formed later than the country rock and the fractures, either by filling of the open spaces or by partial or complete replacement of the adjoining rock, or most commonly by both of these processes combined. Such alteration or replacement does not ordinarily extend far from the fissure. In regions where the vein-forming solu- 1 Called "pitch" or "rake" by many authors. 2 See discussion in Trans., Am. Inst. Min. Eng., vol. 39, 1908, pp. 898-916. 150 MINERAL DEPOSITS tions have acted with unusual intensity a partial alteration may extend from the deposit over considerable areas. No sharp distinction can be drawn between the filled veins and replacement veins. If open spaces are available the metalliferous solutions which formed the veins in most cases found it easier to deposit their load in these spaces than to replace the country rock. Quartz is more likely to be deposited in the open paths, and likewise most of the heavy metals, unless the country rock is one particularly adapted for replacement, such as limestone. Gases like carbon dioxide and hydrogen sulphide penetrate the wall rocks with ease. FIG. 40. Section of Silver Crown lode, Silverton, Colorado, showing lode structure, a, Andesite; b, quartz; c, andesite and quartz stringers; d, ore. After F. L. Ransome, U. S. Geol. Survey. Many veins correspond closely to the old definition of a "true fissure vein," in which the ore occupies the once open spaces along the fracture, with some alteration spreading into the wall rocks. Of such character are the majority of the gold- quartz veins of California and many other occurrences. When the fissures are veiy small they are referred to as veinlets or seams, and all transitions to a slight mineralization of joint planes are found. The walls may be smooth and separated from the vein material by a clay gouge or the filling may closely FORM AND STRUCTURE OF MINERAL DEPOSITS 151 adhere to the country rock. In the latter case the vein is said to be frozen to the walls. Instead of a single break we may have a fracture consisting of a number of approximately parallel fissures, irregularly connected and spaced over a considerable width, which may attain 100 feet or even several hundred feet. These large fracture zones, when FIG. 41. Map showing veins of Central City, Colorado and vicinity. After E. S. Bastin, U. S. Geol. Survey. filled with ore and partially replaced country rock, are called composite veins or lodes (Fig. 40). The Comstock lode in Nevada illustrates this occurrence; its width in places amounts to several hundred feet. Lodes often contain two systems of fractures, intersecting at an acute angle, as shown roughly on Fig. 40. This is sometimes 152 MINERAL DEPOSITS referred to as hammock structure. A number of adjacent parallel veins are called a vein system. If connected by diagonal veins the term linked veins (Fig. 41) is used. When the fractures are closely spaced and parallel we speak of a sheeted zone or a shear zone (Fig. 42). Many of the Cripple Creek veins form good illustrations of this mode of occurrence. The width of a sheeted and mineralized zone is rarely over 50 feet and ordinarily much less. A mass of rock irregularly fractured in various directions by short fissures along which mineralization has spread is called a FIG. 42. Section of the Howard vein, Cripple Creek, Colorado, showing a sheeted zone. Ore follows the close sheeting in the center. Scale, 1 inch equals 13 feet. After Lindgren and Ransome, U. S. Geol. Survey. stockwork. Gold-quartz deposits sometimes assume this form; each seam in the several joint systems intersecting the rock may contain a thin but often strongly auriferous sheet of quartz; the mass may be mined as a whole, furnishing low-grade ore. In deeply weathered regions the upper parts of such deposits may be sufficiently disintegrated to be washed by the hydraulic method. In California such mines are called /'seam diggings." A shattered zone cemented by a network of small non-persist- ent veins is called a stringer lead or stringer lode. Sometimes ore deposits are wholly irregular brecciated masses, the ores filling the interstices between the fragments. Again, the FORM AND STRUCTURE OF MINERAL DEPOSITS 153 breccia may be localized at the intersection of two fractures and a pipe-like deposit will be formed, the ore cementing the frag- ments. Or again, ore deposition may have proceeded in a volcanic vent filled with fragments of rocks due to explosive action. Of such character was the celebrated Bassick deposit in Custer County, Colorado. Amphibolite Quartz Fio. 43. Vertical section of Schlegel milch quartz vein, South Carolina, showing lenticular vein structure in schist with offsets along joint-planes. After L. C. Graton, U. S. Geol. Survey. Brecciation, shattering, and mineralization often follow lines of weakness along dikes; in such cases, illustrated by the Douglas Island mines in southern Alaska, where a dike of diorite intrudes metamorphic clay slates, the mineralized dike is often referred to as a lode. 154 MINERAL DEPOSITS Ladder veins are deposits filling short transverse fissures sometimes occurring in dikes of intrusive rocks (see Fig. 35). Lenticular veins (Fig. 43) are confined mainly to metamorphic schists and their form is sometimes caused by deformation of an older deposit; or again the lenticular shape may be due to stresses FIG. 44. Section of Snowstorm bed-vein, Idaho. After F. L. Ransome, U. S. Geol. Survey. FIG. 45. Vertical section of gash veins filled with galena (black), or with pyrite, zinc blende, and galena in order of deposition. Drusy cavities in center. Lead mines of Wisconsin. After T. C. Chamberlin. causing bulging of the schistose layers. It is common to find a number of short lenses of gold-bearing quartz, for instance, scat- tered along a certain line or zone. Their ends sometimes overlap. Bed veins follow the bedding planes in sedimentary rocks (Fig. 44). Gash veins are deposits filling non-persistent openings that are of fair width but soon cease when followed along strike or dip; FORM AND STRUCTURE OF MINERAL DEPOSITS 155 156 MINERAL DEPOSITS they are particularly characteristic of deposits of galena and zinc blende in limestone and are believed to have been opened by tensional stress, often aided by solution. Where soft sedimentary beds have been folded and crushed, irregular open spaces are more likely to result than well-defined straight fissures. In such rocks ores may be found in the spaces LEGEND Quartz Porphyrjr Bike. Anaconda Vein System Blue Vein Fault System Steward Veint Fault } System Karue fault Middle Fault FIG. 47. N.-S. section across Butte district, showing structure and ore zones. After Reno Sales. opened along anticlines and synclines or in irregular fractures breaking across such folds. Veins and lodes rarely occur single but on the contrary have a tendency to cluster in vein systems such as illustrated in Figs. 37 and 41. In some places may be found several intersecting vein systems of great complexity and differing ages and differing mineralization as, for instance, is the case in the great copper dis- FORM AND STRUCTURE OF MINERAL DEPOSITS 157 trict of Butte, in Montana, illustrated in plan and section in Figs. 46 and 47. * The peculiar divergent fractures at the Leonard mine form what is sometimes called a horsetail structure. The veins at Butte are moreover in many places disrupted by later faults. Veins in Relation to the Country Rock. Veins crossing the bedding in stratified rocks are referred to as cross veins; those parallel to the stratification or schistosity are often called bedded veins or bed veins. Differences in the texture and hardness of the rocks traversed influence the form of the vein markedly. In hard dikes crossed by the vein the deposit often splits up into stringers, resuming its typical form beyond this barrier. In fractures formed under light load near the surface there is a great tendency to irregularity and brecciation, espe- cially in the hanging wall. Following G. F. Becker's proposal such may be called chambered veins. In a vein of strong dip there will also be a tendency for the hanging wall to settle ac- companied by the development of minor vertical fissures. Such conditions were found, for instance, in the Comstock lode, Nevada (Fig. 165) and in the El Oro mines of Mexico (Fig. 159); at both places the vertical hanging wall veins were exceptionally rich, the richness being possibly caused by the impeded circulation of the depositing waters. Large masses of country rock included in the vein material are called horses. Frequently the vein follows a fissure along the walls of a dike; the lamprophyric dikes which are the last phases of batholithic intrusions are especially favored places for ore deposition. Clayey and soft rocks are most resistant to the development of regular fissures; a fracture in hard rock will suddenly die out when encountering such material; many veins pinch immediately upon entering clay shales or masses of clayey gouge. One of the best examples of this is furnished by the veins of Rico, Colorado, which do not extend through the whole sedimentary series in that district, but suddenly cease at a certain stratum of yielding, plastic rocks, termed the blanket, under which almost all the ore-bodies occur. A consequence of this peculiarity of fissuring is that in some regions rich ores are often found just below cer- tain horizons of shale. In southern New Mexico a persistent Devonian shale plays this part of "indicator" (Fig. 70) 1 Reno Sales, Ore deposits of Butte, Montana, Trans., Am. Inst. Min. Eng., vol. 46, 1913, pp. 1-109. 158 MINERAL DEPOSITS The vein solutions were arrested at this horizon and there deposited their load. When a vein follows the contact between two formations, say between granite and andesite, we speak of it as a contact vein. The contact is usually caused by faulting movements in the plane of the fissure, and such veins are in no wise different from ordinary fissure fillings. They should not be confused with contact-metamorphic deposits, which belong to a separate class. Vein Walls. In a simple filled fissure vein we have well- defined foot and hanging walls, which often are smooth sur- faces and represent a single fissure opened by a small or large movement along its slightly curved plane. In a replacement vein the fissures are comparatively tight and in most cases appear to have been formed under stronger compressive stress that reduced the open spaces to a minimum. The vein-forming solutions were forced into the country rock, and the ores formed by replacement gradually merge into unaltered rock. In such cases we may find a single fissure plane with ore on both sides and not limited by any well-defined walls. The exact limits of commercial ore can be found only by assay and are often spoken of as "assay walls." In a composite vein or lode or in a sheeted zone there may be several smooth walls and if no cross-cutting is undertaken there is danger that parallel ore-bodies separated by sheets of country rock may be overlooked. Outcrops. The character of the outcrop of a vein, or in fact of any deposit, is determined by the predominant minerals and by the prevailing surface conditions. In regions of long- continued rock decomposition and inactive erosion, as, for in- stance in some of the Southern Appalachian States, even the most resistant outcrops may be reduced by weathering and nothing but fragments scattered over a wide area may be visible at the surface. Under conditions of fairly active erosion veins with predomi- nant quartz stand out prominently and can be easily traced. On the other hand, veins with carbonate gangue are likely to weather more rapidly than the surrounding rock, and the deposits may be indicated by little depressions or by notches in the ridges. Where the sulphides are abundant, their oxidation is conspicu- ously reflected in the outcrops. Deposits of mingled quartz FORM AND STRUCTURE OF MINERAL DEPOSITS 159 and sulphides then form prominent outcrops of limonite and residual quartz; this is the gossan of the Cornish, the ironstone of the Australian, the eiserner Hut of the German, and the colo- rados of the Spanish terminology. More details in regard to the weathering of ore-deposits are given in Chapter XVIII. Length and Depth of Veins. Where veins follow great dis- locations their length may be considerable. One of the more re- cent veins of Freiberg, Saxony, called the Halsbriicker Spat, has been followed for almost 5 miles. Some of the lead-bearing veins in the Harz Mountains, Germany, are traceable for 12 miles. Exceptionally long single ore-bearing fissures are found in the Silverton quadrangle, San Juan region, Colorado; some of them are 5 miles long. Some of the Mother Lode veins in California can be traced for many miles. The longest single quartz vein known appears to be that of the Pfal, in the Bavarian Forest, which is said to be traceable in a straight line practically without interruption for 140 kilometers through the pre-Cam- brian rocks. 1 The quartz is said to be barren of metals. The great majority of single ore-bearing veins are short and their outcrops can rarely be traced for more than one mile; they do not, as a rule, occupy great dislocations, but rather sub- ordinate fissures. The great dislocations are formed during moun- tain building by tangential stresses, whereas the ore-bearing veins are, as a rule, formed after epochs of igneous activity. In the Coeur d'Alene district, Idaho, for instance, the rich galena veins show little connection with the principal structural faults of the region and were probably not formed at the same time. Veins do not necessarily continue to great depths. There are all kinds of fissures, some disappearing within a short distance below the surface, others continuing down to the greatest depths attained, or about 6,000 feet (Morro Velho, Brazil). Deep tunnels have been run to intersect veins of favorable appearance on the surface and have failed to disclose their continuation in depth. There is no definite relationship between depth and length of a fissure, though it is true that fissures showing strong movement and shattering are likely to continue to great depths. The ore-body may be limited in depth, while the barren fissure continues below it as strong as ever. Bends and curves in strike and dip are common in veins, but 1 E. Suess, Das Antlitz der Erde, Leipzig, 1883, vol. 1, pp. 270-272. W. von Gumbel, Geologic von Bayern, Cassel, 1894, vol. 2, pp. 461-464. 160 MINERAL DEPOSITS as a rule a vein retains its general angle of dip with remarkable persistence. The dip may be at any angle, but veins dipping from 50 to 80 are most common. The North Star vein at Grass Valley, California, is one of the best instances of a low-dipping vein of great length; with a dip of 20 it has been followed for 5,000 feet. Still natter veins are called blanket veins and seldom are very persistent or uniform. CHAPTER XII THE TEXTURE OF MINERAL DEPOSITS FILLING AND REPLACEMENT Introduction. The ore minerals and gangue which make up an ore deposit present various types of texture. The texture of an ore is dependent upon many factors. Space available for deposition, concentration and composition of the generating solutions, time, temperature, and pressure all are of impor- tance in determining the primary texture. Many changes take place in a deposit once formed; the secondary textures, so far as they are caused by solution and redeposition, are influ- enced by the same factors, and, in addition, deformation by pressure plays a most important role. Texture of Deposits of Igneous Origin. The ores consolidated from magmas have in general the texture of igneous holocrystal- line rock. The principal minerals comprise chalcopyrite, pyrite, pyrrhotite, magnetite, chromite, and ilmenite. The texture is ordinarily coarse granular, hypidiomorphic; the chalcopyrite and pyrrhotite are rarely crystallized, but may contain phenocrysts of pyrite and magnetite, both of which are frequently developed with crystalline outlines. The ores may, of course, contain phenocrysts and anhedrons of other rock-forming minerals, par- ticularly soda-lime feldspars, olivine, and pyroxene. Eutectic texture results if the magma was a eutectic mixture from which two minerals crystallized simultaneously after the manner of graphic granite. Approximation at least to such texture is shown by some intergrowths of magnetite and apatite. If the ores have been subjected to dynamic metamorphism, granulation and metasomatic development of hornblende, garnet, biotite, and epidote in coarse or fine aggregates follow and the ore may acquire schistose structure. Texture of Pegmatite Dikes. The pegmatite dikes are believed to have been deposited by magmatic solutions of great fluidity and low temperature (about 600 C.). In many cases the pegma- tites form transitions between igneous rocks and veins deposited 161 1G2 MINERAL DEPOSITS by hot solutions. Their texture is coarsely crystalline, often drusy, and the minerals have a strong tendency to idiomorphic development. Large crystals are the rule, and sometimes they attain enormous dimensions; crystals of spodumene at the Etta mine, South Dakota, are 30 feet or more in length. Quartz crystals several feet long have been observed in these deposits. A rough tendency to crustification is often present, and the walls of the dikes are then lined with crystals of feldspar or mica. Texture of Sedimentary Deposits. Ores and minerals of sedimentary deposits are usually fine grained, and in many cases they have been deposited as colloids in which subsequent fine- grained crystallization has developed. Coarsely crystalline, allotriomorphic structure may develop in deposits consisting of calcite, salt, or gypsum. In many cases the structure is clastic with development of new-formed minerals between the grains. Newly formed quartz, if present, nearly always assumes a microcrystalline or crypto- crystalline texture. Subsequent metamorphism is likely to enlarge the crystalline grains and result in coarser-grained ores. Concretions. 1 Concretions are rounded bodies of some mineral aggregate which are often found in shale and sandstone. Calcite, silica, siderite, pyrolusite, barite, pyrite, marcasite and limonite are among the minerals which most commonly form concretions. The structure is often concentric or radial. In some cases the stratification planes pass through the concretions, while in other cases they may bend around them. These struc- tures are of some economic importance as regards ores of iron and manganese, especially, siderite, limonite and pyrolusite. They often have a center of a clastic grain or a fragment of a fossil shell or leaf. The concretions result from processes of solution and precipitation in soft or semi-consolidated sediments. Accidental precipitation, say of pyrite around decomposing or- ganic material, may start the action and the laws of mass action and preferred growth of larger crystals continue the process. Concretions generally derive their substance from the surround- ing rock. Sometimes the minerals simply fill pores and inter- stices; but in many cases the original substance may have been removed by metasomatic processes. Concretions are frequently 1 J. E. Todd, Concretions and their geological effect, Bull, Geol. Soc. Am., vol. 14, 1904, pp. 353-368. James Geikje, Structural and field geology, 1905, Chapter VIII. THE TEXTURE OF MINERAL DEPOSITS 163 altered with volume changes and development of cracks and interior cavities. When small, uniform and abundant they are called oolites. The oolities usually result from separation in colloidal solutions, often coupled with adsorption of electrolytes. 1 The oolitic texture is characteristic of many deposits of calcite, siderite, calcium phosphate, limonite and psilomelane; pyrite rarely assumes this form. The oolites are often affected by later alteration and recrystallization. Texture of Residual and Oxidized Deposits. In the residual deposits of the zone of oxidation, the ore-bodies are usually very irregular in structure and texture. In large part they were deposited as colloids, which subsequently in part have developed fine-grained crystalline texture. Earthy, clayey concretionary, mammillary, stalactitic, or piso- litic textures are common, the last being defined as a coarser development of the oolitic form. Coarser crystalline form is assumed by some minerals like calcite, barite, zinc carbonate, zinc silicate, and lead carbonate. Crustification or drusy struc- ture is common in places. Quartz, where developed, is usually fine-grained or cryptocrystalline. THE TEXTURE OF EPIGENETIC DEPOSITS Primary Texture of Filled Deposits. The epigenetic deposits are of manifold form and origin, but the majority of them result from aqueous solutions either by filling of open cavities or by replacement of surrounding rocks. Precipitation from complex solutions in open spaces takes place in a certain orderly succes- sion, and the deposits therefore readily assume a banded texture ; crystallization is facilitated by the open spaces, but the older crystals interrupt the development of the products of later crystallization. Hence a hypidiomorphic to panidiomorphic texture is most common. Banding by deposition is called crustification, a term intro- duced by Posepny. In many classes of veins, whether banded or not, a drusy texture is common. In deep-seated veins formed at a temperature but slightly lower than that of the pegmatites the texture is usually coarsely crystalline and massive; sometimes even drusy cavities are 1 For recent literature regarding the origin of oolites see Fortschritte der Min. Krist., u. Petr. Jena. 1913, p. 43. 164 . MINERAL DEPOSITS lacking. Delicate and repeated banding is absent, but a coarsely banded or comb structure recalling that of the pegmatite veins is sometimes encountered. It is usually expressed by quartz crystals developing from the sides or by metallic minerals like tourmaline, wolframite, or cassiterite attached to the walls of the fissure. In veins formed at intermediate temperatures a coarsely crys- talline 'massive texture is most common; combs and rough FIG. 48. Thin section showing normal texture of quartz filling. Black, arsenopyrite ; remainder, quartz with fluid inclusions. Magnified 52 diam- eters. Gold quartz vein, Grass Valley, California. banding by deposition are by no means unknown, especially where the deposit contains calcite or barite. In quartz veins the filling appears to have taken place rapidly and completely, so that the resulting ore consists of an irregular massive mix- ture of quartz and sulphides. That here too the deposition began from the walls is indicated by some occurrences of par- tially filled veins which form a loose aggregate of prisms. Any THE TEXTURE OF MINERAL DEPOSITS 165 thin section of such quartz will usually show long crystals of earlier growth around which the later quartz has been deposited in large individuals (Fig. 48). Lines of inclusions often pene- trate from one grain into another. These inclusions consist of aqueous solutions, often with small cubes or grains of trans- parent salts suspended in the liquid. Inclusions of carbon dioxide have been reported, but are extremely scarce. The optical ^continuity of the crystals or grains is often disturbed by a peculiar divergent "flamboyant" structure which appears to be of primary origin, and not caused by internal strains. The sulphides are coarsely crystalline and sometimes roughly banded, parallel to the walls. Inclusions of country rock may be surrounded by concentric rings of sulphides, and a primary brecciated vein structure may result. Pyrite and arsenopyrite, both among the earliest minerals, have a strong tendency to crystal development, while galena, zinc blende, chalcopyrite, and tetrahedrite are much less commonly found with crystal faces. A banded structure sometimes results from the filling of several closely spaced fissures. In quartz veins in fissile rocks a peculiar book structure may result from numerous parallel sheets of slate, alternating with quartz. It has been thought that this and other features difficult to explain by the assumption of open cavities are due to the opening of spaces by the force of crystallization. Such views have been expressed by E. Suess, W. O. Crosby, E. J. Dunn, S. Taber, and others. It is improb- able that crystallization could have opened the cavities. More likely they were supported by the strong pressure of magmatic waters. But within such spaces a slight force exerted by crys- tallization could readily detach fragments of shale from the walls. Stalactites are unknown in deposits formed at high or inter- mediate temperature. In veins formed at lower temperatures and comparatively shallow depths crustified and drusy forms and fine granular texture pre- dominate. The quartz filling is usually fine-grained, ranging to cryptocrystalline and microcrystalline near the surface. The sulphides are found in small crystals or small anhedrons; large crystals of pyrite, so common elsewhere, are rarely found in these veins. On the other hand, where calcite, dolomitic carbonates, rhodochrosite, fluorite, or barite are gangue min- erals the crystals may be much larger than those found in other deposits. An example is furnished by the magnificent crystals 166 MINERAL DEPOSITS of calcite at Joplin, Missouri, and here galena also appears in unusually large, well-developed individuals. Symmetrical and delicate crustification is often associated with large drusy cavities. Brecciated structure of primary origin is common. Secondary Textures and Structures of Filled Deposits. Crush- ing and brecciation of the early minerals are extremely common; indeed, few veins are entirely free from it. Repeated opening of fissures (Fig. 51) and the deposition of new generations of vein material often take place and the cementing ore may be enriched at the expense of the older generations. FIG. 49. Specimen of quartz from Nevada City, California, showing ribbon structure by sheeting. Two-thirds natural size. A banded or sheeted structure often results from the develop- ment of shear planes in the old filling; examples of this are seen in many gold-quartz veins of California (Figs. 49 and 50). Along these shear planes the quartz is deformed and granulated, and gold may be deposited along them by processes which may be called secondary, though, as a rule, they take place shortly after the vein formation. The shearing stress exerted either before or after the filling may affect the walls of the vein and render them close-jointed or even distinctly schistose. THE TEXTURE OF MINERAL DEPOSITS 167 FIG. 50. Thin section of vein quartz from Nevada City, California, showing crushing and incipient ribbon structure. Magnified 15 diameters. Crossed nicols. FIG. 51. Cross section of Japan vein, Silverton, Colorado, showing structure produced by repeated opening of original fissure, a, Country rock; 6, quartz; c, ore. After F. L. Ransome, U. S. Geol. Survey. 168 MINERAL DEPOSITS In some deposits, especially those containing zeolites, calcite, or barite, secondary replacement processes play an exten- sive part. A vein filled by calcite may be replaced by quartz, which then plainly shows its secondary nature by its hackly or lamellar texture, casts of cleavage, pieces of calcite, or imprints of cleavage lines. Such pseudomorphic textures are sometimes accompanied by a marked enrichment of the metallic content of the deposit. Metasomatism in Mineral Deposits. The nature of meta- somatism or replacement has already been described on pages 26 and 69. Many deposits have been formed by solutions con- taining various salts and gases and capable of attacking certain kinds of rocks. Guided by fissures or other open ducts the solutions deposit part of their load in the open supercapillary spaces whenever supersaturation takes place; thus is produced the filling of fissures. As almost all rocks are porous and as the solutions are frequently under heavy pressure they will be forced into the rocks and will produce chemical and miner- alogical changes in them. At the same time the porous rock acts undoubtedly as a semi-permeable membrane through which various substances will diffuse at differing rates electrolytes and gases most easily, colloids and difficultly ionized compounds very slowly. Thus any vein will usually be accompanied by a strip of altered country rock in which the solutions have effected certain metasomatic changes. The minerals in the open fissures will ordinarily differ from those formed in the metasomatic zone. We may find, for instance, a quartz filling with various sulphides and gold, while the minerals developed in the country rock con- sist of pyrite, sericite and calcite with little if any gold. In some cases no perceptible alteration may be observed in the country rock. The only difference between a filled vein accom- panied by metasomatism and a so-called replacement deposit is that in the latter the filling of the narrow open spaces is negli- gible and the bulk of the ore has been formed by metasomatic processes. Metasomatic Processes. In a solid rock replacement may be caused by many kinds of solutions the only requirement being that some or all of the rock minerals must be unstable in the penetrat- ing fluids. The usual substances, most active in aqueous solu- tions, are oxygen, carbon dioxide, sulphuric acid, ferric sulphate, hydrogen sulphide, alkaline sulphides, and alkaline carbonates. THE TEXTURE OF MINERAL DEPOSITS 169 Replacement may occur at all temperatures above the freezing point of the solution and below the melting point of the rock; it is naturally most effective in hot solutions. Replacement may proceed at any pressure. It may be effected by the ordinary surface waters, by sea water, by hot ascending waters and by magmatic emanations whether gaseous, fluid or above the critical temperature. There is no rock that is proof against replacing natural solu- tions of some kind. Limestone and dolomite are most easily replaced and even at ordinary temperatures, for instance, by iron carbonate (siderite) or by zinc carbonate (smithsonite) . Granite, diorite, and in fact all igneous rocks are also subject to replace- ment. Even quartzite, slate and aluminous shale may be re- placed by other minerals though they are more resistant than others. Replacement by sulphides may to some extent take place at ordinary temperatures (for instance chalcocite replacing pyrite) but large deposits of sulphide ore are usually formed by hot solutions. Mode of Replacement. As pointed out in previous chapters replacement is effected by concentrated solutions filling capillary openings of extremely small size (sheet openings larger than 0.0001 mm., p. 30), which are just above or below the limit of micro- scopic visibility. Cases have been noted when replacement begins from a crack doubtless filled with a film of solution and connecting a series of just visible fluid inclusions. Solution and precipitation go on practically simultaneously dependent upon the constantly changing equilibrium, the supply of solvent and the facility of escape for the dissolved material. Two or several minerals may be dissolved at the same time to make room for the new as in the replacement of shale by a pyrite crystal. The volume of the rock remains constant, held by pressure. The moment a place is available some mineral will separate out from the concentrated solution. This law fails to apply in free crystals or when rock pressure can be overcome by the force of crystalliza- tion, or when a solid is replaced by a gel, or when the solutions circulate so rapidly that there is a strong balance in favor of solution. As crystal grains develop they will exert a different amount of pressure in various directions thus facilitating solution in the direction of greatest pressure. The development of crystals in the host mineral is a result of this action. The power of crystallization of the different minerals varies 170 MINERAL DEPOSITS greatly, for some are found only as anhedrons in metasomatic rocks, while others always assume their crystal form. The following list gives the relative power of crystallization in solid rocks of some minerals, as beginning with those of strongly emphasized individuality: Rutile, tourmaline, staurolite, arseno- pyrite, pyrite, magnetite, barite, fluorite, epidote, pyroxene, amphibole, siderite, dolomite, albite, mica, galena, zinc blende, calcite, quartz, orthoclase. When a crystal has ceased to grow solution may still continue parallel to its faces. As no more material for the crystal is at hand the voids are immediately filled by the next precipitate available. Thus are explained the thin films of quartz or calcite which so frequently surround metasomatic pyrite crystals. When taking place under the law of constant volume replace- ment cannot ordinarily be expressed by the simple chemical formulas 1 usually given. The reactions are likely to be more complicated. 2 In metasomatic processes gangue minerals like sericite, calcite, siderite, barite and fluorite replace all silicates. Ferromagnesian silicates will be attacked first, then the soda-lime feldspars, lastly orthoclase and albite. The degree of attack on quartz depends probably on the amount of alkaline carbonates in the solution. All sulphides replace all silicates as well as quartz. (Fig. 52). Sulphides and sulphosalts readily replace other sulphides. A succession common in many ores is (1) pyrite (oldest), (2) chalcopyrite, (3) galena and zinc blende, (4) sulphosalts (like tetrahedrite). Any of the later minerals may replace any of the earlier products (Fig. 53). Our knowledge of these manifold replacements have been greatly increased by the use of the study of polished sections in reflected light. Sulphides also easily replace gangue minerals but the latter X W. Lindgren, Volume changes in metamorphism, Jour. Geol. vol. 26, 1918, pp. 542-554. 2 Smithsonite often replaces calcite with preservation of structures in- dicating constant volume. The reaction is supposed to follow the formula CaCO 3 + ZnSO 4 = CaS0 4 + ZnCO 3 , both ZnSO 4 and CaSO 4 being water soluble salts. One cubic centimeter of calcite contains 1.192 milligrams CO 2 and 1.518 milligrams CaO while one cubic centimeter of the resulting smithsonite contains 1.514 milligrams CO 2 and 2.787 milligrams ZnO. It is clear then that the principle of equal volumes requires more C0 2 than is available in the calcite. If the process follows the formula, shrinkage of volume will necessarily result. THE TEXTURE OF MINERAL DEPOSITS 171 including sericite, chlorite, calcite, quartz, fluorite and barite very rarely replace sulphides. Texture of Metasomatic Rocks. In metasomatism new minerals develop at countless points in the old rock, some grow- ing with crystal form (metacrystic or crystalloblastic series, p. 170) while others grow into irregular grains. Each new grain may be called a metasome, each new crystal a metacryst (pseudo- phenocryst). 1 The resulting textures will be holocrystalline ; FIG. 52. Replacement veinlets of galena (white) in cryptocrystalline quartz (dark gray) with vugs (black). Tintic, Utah. Magnified 11 diameters. the new minerals frequently contain inclusions of the old (sieve texture) and if the replacement is incomplete, as often is the case, enough of the old texture may be preserved to indicate the original rock (relict texture). It- is characteristic of some replacements that even if the process has been carried to completion the original texture may be preserved as in silicified oolitic limestone and in silicified dolomites (Fig. 56). In many cases, however, the original texture is wholly destroyed. 1 This term was first introduced by A. G. Lane, Bull., Geol. Soc . Am., vol. 14, 1903, p. 369. Grubenmann and Becke use the terms xenoblast and idioblast. N. Grubenmann, Die kristallinen schiefer, Berlin, 1910, p. 91. 172 MINERAL DEPOSITS FIG. 53. Feathery geocronite (5PbS-Sb 2 S 3 ) (white) replacing galena. Tintic, Utah. Magnified 24 diameters. - y FIG. 54. Galena (light gray) replaced by pearceite (9Ag 2 S-AsjS 3 ) (dark gray), in cryptocrystalline quartz. Tintic, Utah. Magnified 227 diameters. THE TEXTURE OF MINERAL DEPOSITS 173 The structure of a rock may be faithfully preserved even when metasomatic action has destroyed its texture. Such preserved structures are, for instance, stratification, joints, breccias, folds and vesicules in lavas. Preservation of texture of limestone which has been completely replaced by sulphides are mentioned by S. F. Emmons 1 and J. M. Boutwell. 2 It is held by some petrographers that metamorphic rocks show no recognizable succession in order of crystallization but this is certainly not always correct. In many replacements gangue FIG. 55. Same replacement magnified 690 diame^rs. Note that earlier barite plates (black) are not replaced by galena but by the later pearceite. minerals like quartz and barite may crystallize first, while pyrite comes next and other sulphides later. Irving 3 has pointed out that in some cases replacement begins from a great number of points in the rock where metasomes or metacrysts may develop (Fig. 57) and by continuation of the same process (Fig. 58) the remainder of the rock is finally re- replaced; the contact is then indefinite. In other cases the complete change occurs rapidly, advancing like a wave over the 1 S. F. Emmons, Trans., Am. Inst. Min. Eng., vol. 23, 1893, p. 602. 2 J. M. Boutwell, Prof. Paper 38, U. S. Geol. Survey, 1905, p. 193. 3 J. D. Irving, Jour. Canadian Min. Inst., vol. 14, 1911, pp. 395-471. 174 MINERAL DEPOSITS country rock; the contacts are then sharp and the process prob- ably consisted in replacement of the original rock by colloid silica. The replaced rock is usually dense and compact; in places, however, drusy cavities occur in it. Under the influence of the same solution different results may be produced in different rocks. Limestone may be silicified while diorite may be transformed to sericite. Replacements at High Temperature. Complete recrystalliza- tion, development of silicate minerals with little or no water, FIG. 56. Thin section of dolomite completely silicified, but retaining texture and crystal form. After J. D. Irving, U. S. Geol. Survey. and coarse texture are typical of deposits formed by replacement at high temperatures. Mineralizers like fluorine, boron or phos- phorus are frequently introduced. The best examples of such textures are found in the replace- ment of limestones in contact-metamorphic deposits (Figs. 250 and 251) resulting in coarse aggregates of metacrysts of andradite garnet with metasomes of quartz, calcite, epidote and pyroxene. The limestone may be recrystallized in part to coarse calcite. Magnetite and sulphides develop in large grains. THE TEXTURE OF MINERAL DEPOSITS 175 Adjoining tin-bearing veins the rocks are recrystallized to greisen, consisting of coarse metasomes of muscovite, topaz, quartz, tourmaline, fluorite and cassiterite (Fig. 226). Cal- careous rocks or greenstones containing much lime are recrystal- lized to aggregates of axinite, actinolite, garnet, etc. Replacement by apatite (containing phosphorus), scapolite (containing chlorine) and pyroxene occur adjoining certain high temperature veins. Along many deep-seated gold quartz vein much albite, biotite, and zoisite develop in the wall rock. Replacements at Intermediate Temperature. Replacements at more moderate temperature are likely to result in fine-grained textures, and hydrous silicates like chlorite and sericite are abundant (Fig. 61). There are exceptions to this where barite or fluorite replace limestone for both of these minerals easily develop as perfect metacrysts (Figs. 62 and 64). Silicification of limestone, argillaceous shale and rhyolite is a very common process taking place frequently with preservation of texture. The quartz will usually be fine-grained. Silicified limestones are called jasperoids (Figs. 57 and 58). Limestone may be replaced by massive sulphides (Fig. 63). Alteration by hot waters of granular and porphyritic igneous rocks as well as of schists of similar composition results in sericiti- zation of the femic and salic minerals, sometimes also of the quartz, with development of fibrous aggregates. Pyrite, sec- ondary quartz, rutile, albite and adularia are sometimes found in these rocks. In some classes of such metasomatic rocks carbonates of calcium, magnesium and iron also occur. Serpentine is altered near some gold quartz veins to coarse aggregates of ankerite, quartz and mariposite (chromiferous sericite). Quart zite and quartzitic slates in some lead deposits may be extensively replaced by siderite (Fig. 60). In deposits which have been formed by hot waters hear the surface where the rocks are permeable the incipient alteration of igneous rocks is often wide spread with alteration of the femic minerals to chlorite, calcite or epidote (propylitization). Replacement of effusive rocks by alunite, pyrite and kaolinite is characteristic of some deposits formed near the surface. Replacement at Low Temperature. Under the influence of cool solutions the intensity of replacement is diminished. The minerals formed are strongly hydrated, the texture fine-grained. 176 MINERAL DEPOSITS FIG. 57. Incipient silicification of limestone. Aspen, Colo. White areas represent quartz crystals with small inclusions of limestone. Magni- fied 30 diameters. FIG. 58. Silicified limestone ("jasperoid"). Aspen, Colo. Crossed nicols All quartz. Small inclusions of calcite in some of the grains. Magnified 30 diameters. THE TEXTURE OF MINERAL DEPOSITS 177 In igneous rocks chlorite, quartz and kaolin, possibly also sericite, may form by the action of cool weak solutions. Replacement FIG. 59. Replacement veinlet of tourmaline in fresh andesine grain. Keystone mine, Meadow Lake, Nevada County, Cal. t, Tourmaline; /, andesine; e, epidote; s, sericite. Magnified 50 diameters. FIG. 60. Siderite with pyrite and galena, replacing quartzite. Helena and Frisco mine, Coeur d'Alene, Idaho, q, Quartz grains; s, sericite; si, siderite; black, galena and pyrite. Magnified 100 diameters. by sulphides such as pyrite, galena and zincblende may take place Limestone may be silicified to fine-grained jasperoids. 178 MINERAL DEPOSITS To a limited extent sulphides may replace other sulphides. Chalcocite for instance replaces pyrite, chalcopyrite and bornite. In acid descending waters kaolin replaces sericite and other silicates. FIG. 61. Andesine crystal in granodiorite, replaced by sericite and calcite. Pinetree vein, Ophir, Placer County, Cal. q, Quartz; m, musco- vite; c, calcite; s, sericite. Magnified 80 diameters. FIG. 62. Barite (B), replacing gray, fine-grained limestone (L), Ouray, Colo. After J. D. Irving, U. S. Geol. Survey. Heated alkaline waters are not believed to be capable of de- veloping kaolin from the aluminum silicates of the rocks; alkaline silicates like sericite will result. On the other hand the ordinary dilute ground waters will develop kaolin in the rocks. THE TEXTURE OF MINERAL DEPOSITS 179 In other words kaolin is confined to the uppermost metamor- phic zone and rarely ventures far below the zone of weathering. 1 FIG. 63. Galena, replacing crystalline dolomite. Elkhorn mine, Mon- tana, g, Galena; p, pyrite; c, calcite grains of limestone; q, secondary quartz. Magnified 15 diameters. FIG. 64. Fluorite replacing limestone. Florence mine, Judith Moun- tains, Montana. /, Fluorite: I, limestone; q, secondary quartz. Magnified 7 diameters. Criteria of Replacement. F. Posepny first established replace- ment as a mode of origin of mineral deposits. Shortly after- 1 W. Lindgren, The origin of kaolin, Econ. Geol, vol. 10, 1915, pp. 89-93. 180 MINERAL DEPOSITS ward S. F. Emmons 1 demonstrated it to be a common mode of origin and illustrated it by the description of many ore-bodies in Colorado and elsewhere. About 1900 W. Lindgren described the principal modes of metasomatism. 2 In 1911 J. D. Irving 3 published a paper of great value in which the criteria of replacement ore-bodies were summarized. Some of these criteria in favor of replacement have already been mentioned but they may be briefly recalled here: 1. Form of ore-body, more or less irregular. Gradually fading limits. Not always conclusive. 2. Presence of unsupported residual rock masses. Sometimes the orientation of bedding may be proved parallel with the surrounding rocks. 3. Absence of crustification. A banding may be observed in places due to preservation of bedding or shearing planes. 4. Absence of concave contacts; in limestone, for instance, solution of cavities tends to produce flat concave depressions; a filled cave would show this whereas replacement proceeds with convex outlines toward the unaltered rock. 5. Preservation of textures and structures of original rock. The last- named criterion is the most conclusive. The criteria for the determination of replacement are some- times difficult to establish; many mistakes have been made along this line. Replacement veinlets crossing the older minerals and dependence of the replacing mineral on minute fissures and cracks constitute good evidence. The projecting of crystals of. one mineral into another is not always a safe proof of replace- ment. The apparent host may possibly be a later mineral molded about the crystals. In many cases adjoining minerals may have developed practically simultaneously. A peculiar type of replacement results in pseudo-eutectic texture simulat- ing an intergrowth (Figs. 54 and 55). 1 S. F. Emmons, The genesis of certain ore deposits, Trans., Am. Inst. Min. Eng., vol. 15, 1887, pp. 125-147. S. F. Emmons, Structural relations of ore deposits, idem, vol. 16, 1888, pp. 804-839. S. F. Emmons, On the origin of fissure veins, Proc., Colorado Sci. Soc., vol. 2, 1888, pp. 189-208. 2 W. Lindgren, Metasomatic processes in fissure veins, Trans., Am. Inst. Min. Eng., vol. 30, 1901, pp. 578-692. 3 J. D. Irving, Some features of replacement ore-bodies and the criteria by means of which they may be recognized, Jour. Canadian Min. Inst., vol. 14, 1911, pp. 395-471; Econ. GeoL, vol. 6, 1911, pp. 527-561. THE TEXTURE OF MINERAL DEPOSITS 181 Role of Colloids in Filling and Replacement. It is well known that colloid deposits, for instance, of silica, iron hydroxide and aluminum hydroxide play an important part in mineral deposits formed at or near the surface. Colloid minerals are also often deposited during the oxidation of ore deposits. In the discussion in this chapter the colloids have not thus far been considered. There is, however, an increasing mass of evidence that colloid silica or silica gel is of considerable importance in the origin of deposits formed relatively near the surface by ascending waters. Some of the quartz filling in such veins is extremely fine-grained and bears evidence of having been deposited as a stiff jelly which soon afterward was crystallized in chalcedonic or cryptocrystal- line form. 1 Clear evidence of this is seen in some filled veins from the Tintic district, Utah, 2 where the original delicate band- ing by deposition is still seen though the substance is now micro- crystalline quartz. There is also good evidence presented in the last-named paper on the Tintic district to show that in some deposits formed at moderate depths and not very high temperatures, limestone and dolomite may be replaced by silica gel which afterward crystal- lized to chalcedony. This type of replacement appears to be characterized by sharp contacts with the unaltered rock; it does not proceed from crystal nuclei of quartz starting at numerous points but advances like wave and stops with sharp contacts (see p. 173). Later metalliferous solutions penetrated this gel and deposited sulphides in it. Sometimes a banding has been produced which strongly recalls the so-called Liesegang rings 3 in artificial gels and indicate a sort of rhythmical precipitation of sulphides. 1 W. Lindgren, Geology and mineral deposits of the National District, Nevada, Bull. 601, U. S. Geol. Survey, 1915. 2 W. Lindgren, Processes of mineralization and enrichment in the Tintic mining district, Econ. Geol., vol. 10, 1915, pp. 225-240. 3 R. E. Liesegang, Geologische Diffusionen, 1913, p. 180. Reviewed by' A. Knopf in Econ. Geol, vol. 8, 1913, p. 803. CHAPTER XIII ORE-SHOOTS 1 Form of Primary Ore -shoots. Commercial ore or mineral does not ordinarily occupy the whole volume of a deposit. The ore is in most cases surrounded by material of poorer grades, sometimes fading into the country rock, or again sharply separated from it. In replacement deposits the disseminated grains of galena, for instance, or zinc blende, may gradually become so few that the mass can no longer be treated with profit. In veins, only certain parts of the sheet-like body can be extracted, while the remainder of the vein material may consist of gangue minerals only, or of clayey attrition masses or breccias. Those parts of a deposit in which the valuable minerals are so concentrated that their utilization becomes possible are called ore-shoots. Their occurrence and form are exceedingly variable, and it is often most difficult to ascertain the causes which have guided their development. A full discussion of this subject is scarcely possibly here, for it involves the whole question of gene- sis of mineral deposits. In. deposits of sedimentary origin the ore-shoots have, of course, the general tabular form, but admixture with gangue materials or valueless matter may so dilute the ore that only certain parts of the body can be extracted. Various assort- 1 C. R. Van Hise, Some principles controlling the deposition of ores, Trans., Am. Inst. Min. Eng., vol. 30, 1900, pp. 27-177. T. A. Rickard, The formation of bonanzas in gold veins, Trans., Am. Inst. Min. Eng., vol. 31, 1902, pp. 198-220. The localization of values in ore-bodies, etc. Discussion by J. D. Irving, F. C. Smith, Reno Sales, F. L. Ransome, H. V. Winchell, H. Sjogren, and W.^Lindgren, Econ. Geol, vol. 3, 1908, pp. 143-154, 224-229, 326-330, 331-336, 425-427, 637-642; vol. 4, 1909, pp. 56-61. C. W. Purington, Ore horizons in the San Juan Mountains, Econ. Geol., vol. 1, 1905-06, pp. 129-133. H. C. Hoover, The valuation of gold mines, Eng. and Min. Jour., May 19, 1904. R. A. F. Penrose, Jr., Some causes of ore-shoots, Econ. Geol., vol. 5, 1910, pp. 97-133. 182 ORE-SHOOTS 183 ments of detritus and complex conditions of precipitation from waters of seas, lakes, and rivers have influenced the concentra- tion of the richer ore masses. In addition, alterations by meteoric waters are common; in the case of phosphate deposits and beds of siderite they have resulted in enrichment. In deposits of igneous origin the general form of the deposit is also that of the ore-shoots. In some deposits, such as the magnetite deposits of northern Sweden and the dike-like de- posits of ilmenite at Iron Mountain, Wyoming, there is prac- tically no waste material and the whole igneous body constitutes ore. More commonly the irregular lenticular or tabular masses of igneous rocks in which ore minerals have developed by mag- matic segregation (for instance, gabbro containing chalcopyrite) have nuclei of richer material gradually fading into more normal rock. In the epigenetic deposits the outlines of the ore-shoots are exceedingly variable. In those deposits which are formed by replacement this is particularly true, and few rules can be laid down for their occurrence; the form is determined by the fissures giving access to the solutions, by the presence of impermeable rocks, and by the varying susceptibility to replacement of the original rocks. Most attention has been given to the shoots in fissure veins. Although the ore in the main follows the fissure and therefore has a tabular or sheet-like form, it rarely occupies the whole space along this fissure, but is concentrated in bodies of vary- ing size, shape, and continuity; smaller bodies are known as bunches, pockets, or kidneys; in gold-quartz veins these may be exceedingly rich. Narrow ore-shoots, greatly elongated in the vertical direction, whether occurring in fissure veins or independ- ently of them (for instance, in volcanic necks), are called chim- neys, pipes, or necks (Fig. 68). Ore-shoots may be entirely irregular, but commonly have a more or less well-defined columnar, steeply pitching shape, best shown in projection upon the plane of the vein. Fig. 65 shows the terminology proposed 1 for various dimensions of an ore-shoot in a vein. The pitch length, or axial length, is the distance be- tween the two extreme ends of the shoot. The pitch is the angle 1 W. Lindgren and F. L. Ransome, Prof. Paper 54, U. S. Geol. Survey, 1906,'p. 206. 184 MINERAL DEPOSITS which the pitch length makes with the strike of the vein, and is measured on the plane of the vein. The stope length is the hori- zontal length of the ore-shoot on any particular level. The thick- ness or width is measured perpendicularly to the plane of the vein. The breadth of the ore-shoot is the stope length, multiplied by the sine of the pitch. Fig. 66 shows the ore-shoots of a gold-quartz vein at Nevada City, California. Flat-dipping shoots are not so common. Fig. 67 shows an excellent example of a flat shoot in the celebrated Eureka-Idaho vein at Grass Valley, California. FIG. 65. Diagram illustrating the terms used to describe the dimensions of ore-shoots. After W. Lindgren and F. L. Ransome, U. S. Geol. Survey. In parallel veins the shoots are often, roughly speaking, coextensive. Sometimes the shoots in a series of parallel veins persistently recur where the veins cross a certain stratum or dike, as, for instance, where the gold-quartz veins of Gympie, Queensland, intersect certain carbonaceous strata, or as at Thames, New Zealand, where the veins intersect certain soft- ened and altered andesites. Many shoots follow intersections of veins or of veins with fissures. Shoots, however large, do not continue indefinitely, but end in depth, usually with gradual deterioration. Small masses or kidneys are likly to be found below the termination of a large ore-shoot. Exploration may find another shoot below the first, either on the same fissure or imbricating on a parallel ORE-SHOOTS 185 FIG. 66. Ore-shoots of veins at Nevada City, California. Idaho -Maryland Shaft Ma.slin Shaft FIG. 67. Approximate outline of the Eureka-Idaho ore-shoot, Grass Valley, California, in projection on the plane of the vein. 186 MINERAL DEPOSITS vein. When great depth is attained the grade of the ore usually decreases in the deeper levels, but this rule is not without ex- ceptions. Many shoots are lenticular, that is, they contain a rich LEVEL :'.'( :' : ! BURNS SHAFT _ 1 -1 FIG. 68. Stereogram of Anna Lee ore chimney, Cripple Creek, Colorado. Shoot probably determined by intersection of the basic dike with a fissure. After V. G. Hills. nucleus, outward from which the ore gradually decreases in tenor. H. C. Hoover, from an examination of 70 mines, concluded ORE-SHOOTS 187 that ore-shoots are generally lenticular and that the probable minimum extension of an ore-shoot below any given level would be a factor of not less than a radius of one-half of its breadth. At Cripple Creek Lindgren and Ransome found that the shoots which begin distinctly below the surface have a marked elon- gated form, the ratio between pitch length and breadth varying from 13^ : 1 to 5 : 1. Primary ore-shoots rarely continue for more than 2,000 feet along the strike, or for more than 2,000 feet along the pitch length. In a given district the pitch of the ore-shoot is often pre- dominantly in one direction; thus at Nevada City and Grass Valley the shoots pitch to the right of an observer who looks down the dip of the vein. In another district the opposite may be true. In some places the tenor varies directly, in others inversely with the swelling of the vein. According to a rule often quoted, the shoots follow the directions of the striations on the vein walls, but this again by no means has universal application. Shoots of Successive Mineralizations. While in some veins the whole width consists of uniform ore, it is exceedingly common, especially in thick veins, to find that there are certain streaks which are far richer than the rest. They may follow foot-wall or hanging-wall, or the center of the vein, or may switch from one side to another. Such phenomena indicate re-opening of the vein or brecciation after the first period of vein-filling and enrichment. Superficial or Secondary Shoots. Descending surface waters decompose and often enrich the upper part of veins or other de- posits. Such enriched superficial portions of an ore deposit are dependent upon the ground-water level and, when projected upon the plane of the vein, follow the surface of the ground and terminate below along an irregular and jagged line. Oxidized ores, as well as sulphides due to enrichment, are found in them, usually at different levels. The surface shoots are in fact char- acterized by horizontal extension, in contradistinction to the predominance of the vertical direction in the primary shoots. The mineralogical characteristics of superficial shoots will be discussed in detail in a later chapter. Their tendency is to spread along the strike of the vein, often also out into the wall rock. Thus pay ore may be found for a long distance along the 188 MINERAL DEPOSITS trend of the vein and its appearance will be that of the oxi- dized croppings of a long primary shoot, when in fact deeper explorations may prove the existence of only a few narrow primary ore-bodies underneath the continuous surface ore. Sometimes, as in Calico, San Bernardino County, California, and numerous other places, oxidized silver ores will be found in croppings along a vein which are simply concentrations of a primary vein fill- ing that contains no workable shoots. To this class belong also the horizontal or flat shoots of secondary copper sulphides (chalcocite and covellite) formed by descending solutions in copper deposits at or near the water level. The primary material may or may not constitute commercial ore. If spread over wide mineralized areas such shoots are often called chalcocite blankets. Lateral spreading is often characteristic of shoots of oxidized ores. Descending metal solutions may wander out in the country rock and here form new bodies. Causes of Primary Ore-Shoots. Ore-shoots are due to the abundant precipitation of valuable minerals from their solutions. The causes are in part chemical and in part mechanical: 1. Decrease of pressure and temperature. 2. Favorable chemical character of wall rock. 3. Favorable physical character of wall rock. 4. Intersections. Decrease of Pressure and Temperature. The fundamental reason for the occurrence of ores in veins and allied epigenetic deposits in the upper crust is probably that the metals were in solution in hot waters which were ascending and gradually encountered conditions favorable for precipitation. First among these conditions is decreasing temperature. If this is true the deposits should gradually become poorer or barren in depth. 1 In a general way this is doubtless true, but for many substances the vertical space through which deposition can take place is very 1 T. A. Rickard, Persistence of ore in depth, Trans. Inst. Min. and Met., vol. 24, 1915, pp. 3-46, with discussion. W. Lindgren, Ore deposition and deep mining, Econ. Geol, vol. 1, 1905, pp. 34-46. F. L. Garrison, Decrease of value in ore-shoots with depth, Trans. Cana- dian Min. Inst., vol. 15, 1912, pp. 192-209. J. F. Kemp, The influence of depth on the character of metalliferous deposits, Compte Rendu, 12e Session, Canada, Congres geologique internat., 1914, pp. 253-260. Malcolm Maclaren, Idem, pp. 295-304. ORE-SHOOTS 189 large. We know that gold-bearing quartz was deposited in Cali- fornia over a vertical distance of 4,000 feet, while in southeastern Alaska and at Bendigo, Australia, the interval is not less than 5,000 feet. This deposition took place at considerable depth below the surface, probably several thousand feet below it, and as it is known that gold-bearing quartz may also be deposited within the upper zone, we have thus a total vertical range of at least 9,000 feet. At the lowest levels at the places mentioned the ore is of low grade, but in Alaska at least there is a large quantity available. The richest ore was doubtless deposited MYSORE GM. McTaggart> Hancock's G\eu Riddlesdale'8 Kowse's Tayl Vertical FIG. 69. Pitching ore shoots in gold quartz veins, Kolar, India. After T. A. Richard. close to the surface, where we find the bonanzas of the Tertiary gold and silver veins; but below this bonanza zone the decrease in tenor of the ore is very slow and rich shoots and pockets may be found at great depth below the original surface. The most per- sistent gold-bearing ore shoots known are those in veins formed at intermediate or high temperatures. Such are, for instance, the North Star vein at Grass Valley, California, which with very slight impoverishment has been followed for 6,400 feet on a dip of 20 (p. 569). The Kolar veins in India have been mined to a vertical depth of 4,000 feet in shoots of considerable regularity (Fig. 69), and with little change in tenor of ore. 190 MINERAL DEPOSITS The most persistent ore body known is that of Morro Velho mine in Brazil, where a pitching ore shoot has been worked to a vertical depth of 6,200 feet and a pitch length of 9,000 feet (Fig. 236). For copper ores the vertical range of deposition is likewise great, though unlike gold and silver they seem to be deposited in greatest quantity at lower levels and high temperatures. Lead, on the other hand, appears to be precipi- tated nearer the surface and at lower temperatures; while zinc in this respect stands between copper and lead. The relations set forth explain why so little decisive evidence of vertical succession in deposition is available from observa- tions at any one mine. In the Cornwall veins tin and tungsten prevail in the lower levels in granitic country rock, while copper was deposited in the cooler region of the slates covering the granite batholiths; the lead ores are found some distance away from the intrusive granite. In many lead mines it has been noted that within a distance of 700 to 3,000 feet from the surface the lead minerals give .way to pyrite and zine blende. In quicksilver mines the ore often becomes impoverished within 1,000 feet below the surface. The dependence of the deposition of various metals upon temperature and therefore also upon the vertical and hori- zontal distance from the place of origin of the mineralizing solutions has been emphasized lately by several investigators. l Character of Wall Rock. The character of the wall rock has sometimes a decided influence on the ore-shoots, but it is not always easy to decide whether it is due to chemical or mechanical causes. In replacement deposits limestone and lime shale are usually favorable, but in the Coeur d'Alene district of lead- bearing veins a quartzitic schist is the rock best adapted for replacement by siderite and galena. At Freiberg, Saxony, the gray gneiss is the favorable rock, while the veins split or become unproductive in the red gneiss or in the mica schists. Carbonaceous rocks are believed to influence deposition favor- ably by their reducing action; the gold-quartz shoots of Gym pie, 1 J. E. Spurr, A theory of ore deposition, Econ. Geol., vol. 2, 1907, p. 790. L. De Launay, La metallogenie de 1'Italie, Congres geologique internal., Mexique, vol. 1, 1906, p. 571. Also in Gltes Mineraux, vol. 1, Paris, 1913. W. Lindgren, Processes of mineralization and enrichment in the Tintic mining district, Econ. Geol. vol. 10, 1915, p. 228. ORE-SHOOTS 191 Queensland, are often quoted, as well as the supposedly car- bonaceous "indicator" at Ballarat, Victoria. The well-known replacement of fossil wood by chalcocite in a certain class of copper deposits may be added to these examples, as well as the supposed influence of certain oil shales on the deposition of lead ores in Wisconsin. The importance of precipitation by carbonaceous material has been overestimated, but in many cases the hydrocarbons have certainly favorably influenced the deposition of ores. 1 Rocks containing pyrite or other sulphides often enrich trav- ersing veins. Examples of this are known from Kongsberg, Norway, where the silver veins are productive when crossing certain schists with disseminated sulphides. At Ophir, Cali- fornia, gold-quartz veins are enriched when crossing "iron belts" of pyritic amphibolites. Where a vein cuts through a thick series of sedimentary rocks it often widens and contains rich ore in the limestones, while poor or barren in shale or sandstone. Similarly, where a thick series of igneous rocks, as in the San Juan region, Colorado, is intersected by veins ore horizons will develop in rocks which by their physical and chemical character are most favorable to con- tinuous fissures or to replacement. Rhyolites are generally unfavorable because fissures often tend to splitting in such rocks; tuffs likewise because the solu- tions tend to disperse through great masses of rock. On the other hand, rocks like andesites and latites are usually favorable. Purington (op. tit.} has shown that in the San Juan Mountains the andesitic breccias which contain abundant ferro- magnesian silicates are most favorable to ore deposition. Impermeable Barriers. The conditions outlined above would tend to produce more or less horizontal ore-bodies. Such ore-bodies are most conspicuous where impervious rocks inter- pose barriers to the solutions. The occurrence of ore in hori- zontal extension below such barriers is in fact one of the best indications that the solutions have been ascending in the main. Fig. 70 shows the occurrence of oxidized silver ores below the Devonian shale at Chloride Flat (Silver City), New Mexico, and similar occurrences are not uncommon in other mining districts of New Mexico. The blanket veins of Rico, Colorado (Fig. 71), 1 W. P. Jenney, The chemistry of ore deposition. Trans., Am. Inst. Min. Eng., vol. 33. 1903, pp. 445-498. 192 MINERAL DEPOSITS FIG. 70. Sketch section showing occurrence of ore-shoots in limestone at contact of overlying Devonian shale at the Bremen mine near Silver City, New Mexico, a, Limestone; 6, shale; c, ore. After R. A. F. Penrose, Jr. FIG. 71. Diagrammatic section across a lode, and ore-body formed beneath an impervious stratum (blanket) of black shale, Rico, Colorado. After F. L. Ransome, U. S. Geol. Survey. ORE'S HOOTS 193 present another good illustration of this principle, as do also the ores of the American Nettie mine near Ouray, Colorado, and the siliceous gold ores replacing dolomite in the Black Hills. FIG. 72. Longitudinal section along the Neu Hoffnung vein, Freiberg, Germany, showing ore-shoots along intersection with several other veins. After R. Beck. The impermeable stratum is not necessarily shale; it may be a gouge in a fissure, or a sheet of volcanic rock which, for some reason, the fissures failed to penetrate. The same principle of 194 MINERAL DEPOSITS impermeable barriers serves to explain why the vein material is often confined between the clay seams of hanging and foot wall without entering the adjacent country rock by replacement. Where one fissure is faulted by another, deposition may occur because the circulation becomes impeded at the fault. It is not entirely clear why deposition of rich ores should take place when the solutions are impeded and partial stagnation follows, but the conditions observed bear sufficient testimony to the fact. Where the solutions have moved downward, as in the concen- tration of hematite ore from poorer "iron formations," it is often observed that ores occur on impervious basements and in troughs caused by shales, clayey fissures, or dikes. Intersections. Enrichment and ore-shoots along intersec- tions of two veins or of a vein and a fissure are very common phenomena, well exemplified at Freiberg, Saxony (Fig. 72), and at Cripple Creek, Colorado. Van Hise attributes the shoots at such intersections to the mingling of two solutions and conse- quent precipitation of some constituents. In part they may be due to the shattering of the rocks at the intersection, and Penrose notes that shoots are more likely to occur where the intersection takes place at acute angles, forming wedge-shaped blocks that are easily broken along their edges. Though enrichment at intersections is common it is by no means a universal rule, and indeed sometimes a vein is impover- ished at the intersection with a barren fissure. The occurrence of the large shoots such as those in the gold- quartz veins of California, at Cripple Creek, and in the Coeur d'Alene lead mines cannot be fully explained by intersections or by the influence of the wall rock. Such shoots are generally considered as the result of decrease in temperature of ascending solutions in channels of circulation. CHAPTER XIV THE CLASSIFICATION OF MINERAL DEPOSITS Classification by Form and Substance. A genetic classifica- tion of deposits of useful minerals is really equivalent to the classification of "geological bodies" as definied in Chapter I and is therefore naturally beset with all the difficulties connected with an imperfect knowledge of geological processes. The early attempts in the way of systematic treatment, however, avoided this troublesome path by the simple expedient of classifying by substance or uses, or by form. These schemes are followed in many text-books, even among those of recent date; undoubtedly they have some advantages, especially for the miner, the indus- trial chemist, or the metallurgist, who are principally interested in the form of the deposit or in the study of ores of certain metals. By substance and uses mineral deposits may be classified as follows : 1 . Structural materials Stone, glass sand, cement rock, clay, asphaltum. 2. Fuels Coal, petroleum, natural gas, peat. 3. Abrasives Corundum, garnet. 4. Fertilizers Potash salts, phosphates, green-sands. 5. Precious stones Diamond, opal, tourmaline. 6. Various industrial uses. . .' Graphite, barytes, borax, asbestos. sulphur. 7. Metallic ores Iron ores, copper ores, gold and silver ores, tin ores, aluminum ores, etc. However convenient, it is evident that this classification cannot lead to a thorough appreciation of the manifold processes by which mineral deposits are formed in nature. The first attempts at a classification of the deposits themselves were made by the miners and thus the early and not yet entirely abandoned schemes refer to the form of the geological bodies. But form is closely connected with genesis and even in one of the earliest classifications on this basis, that of Bernhard von Cotta, 1 1 Die Lehre von den Lagerstatten, Freiberg, 1859. 195 196 MINERAL DEPOSITS the difficulty of avoiding genetic conceptions is felt in his defini- tion of a vein as a "filled fissure." He divides ore deposits as follows : I. Regular deposits. A. Beds. B. Veins. a. Ordinary fissure veins (true fissure veins). b. Bedded veins. c. Contact veins. d. Lenticular veins. II. Irregular deposits. C. Stocks. (Irregular masses with distinct limits.) a. Recumbent. 6. Vertical. D. Impregnations. (Irregular masses, fading into coun- try rock.) With variations this plan of classification is followed in many of the older text-books. Not unlike it is a classification by J. A. Phillips in his treatise on ore deposits, revised in 1896 by H. Louis. Lately James Park has adopted the same plan with some modifications in a useful and practical text-book on mining geology. 1 His classification is as follows: I. Superficial deposits. a. Fragmentary. 6. Massive. II. Stratified deposits. a. Constituting beds. 6. Disseminated through a bed. III. Unstratified deposits. a. Deposits of volcanic origin. 6. Stockwork deposits. c. Contact or replacement deposits. d. Fahlbands. e. Impregnations. /. Segregated veins. {/. Gash veins. h. True fissure veins. 1 A text-book of mining geology, London, 1907, p. 219. CLASSIFICATION OF MINERAL DEPOSITS 197 Park states that this classification is only an empirical arrange- ment to facilitate the study of ore deposits, and a provisional classification by origin also is given. L. De Launay 1 arranges the deposits according to the prin- cipal elements contained. This logical, though not genetic, plan has been followed in part in the index appended to this book. Genetic Classifications. A genetic classification is the most desirable both theoretically and practically. In exploring and exploiting ore deposits, the miner is almost forced to form an idea of its origin in order to follow up the ore-bodies to best advantage. Von Groddeck and Stelzner were really the first mining geologists who appreciated and applied the genetic prin- ciple in classification. 2 Of course, the time was hardly ripe for its introduction until the conceptions of genesis had crystallized into fairly definite form. Stelzner remarks, with good reason, that it is only by standing upon the ground of a genetic theory that the miner finds courage to sink deep shafts or drive long tunnels. We are still in doubt as to the true mode of origin for many deposits. But, as von Groddeck and Stelzner have pointed out, this applies to any classification and this very uncertainty is a stimulus to further investigations. The different classifications proposed will not be given here in detail. An excellent account is found in Kemp's "Ore deposits of the United States and Canada," Appendix I. Von Groddeck and Stelzner, Posepny, Wadsworth, Monroe, Kemp, Crosby, Hoefer, Spurr, Van Hise, Weed, and several others have more or less successfully attacked the problem of a consistent genetic classification. Von Groddeck, followed by Stelzner and Beck, makes the primary distinction 'whether the useful minerals were originally formed in or with the rock in which they now occur or whether they were introduced into pre-existing rocks. Stelzner called the former syngenetic, the latter epigenetic. (Author's lecture notes, Freiberg, 1881.) 1 L. De Launay, Gltes Mineraux et Metalliferes, 3 vols., Paris, 1913. 2 The former says: "I must confess that I have never been able to under- stand the satisfaction which many people feel when they are informed that a certain deposit, for instance, is a 'stock.' This information has, on the contrary, always produced in me a feeling of deep dissatisfaction." Quoted in Stelzner and Bergeat, Erzlagerstatten, pt. 1, 1904, p. 10. 198 MINERAL DEPOSITS J. F. Kemp divides the deposits into (I) those of igneous origin, (II) those precipitated from solutions, and (III) those deposited from suspension, or residues after the decomposition of rocks. Difficulties appear here too, for what are igneous mag- mas but solutions? j^j*jj Beck's classification is in part based on that of Stelzner. In the first edition of his hand-book " Die Lehre von den Erzlager- statten" the syngenetic or epigenetic origin was made the principal basis of classification. In the edition of 1909 this is changed and the deposits are classified as follows, on the basis of the various phases of their genetic history : 1. Magma tic segregations. 2. Contact-metamorphic ore. deposits. 3. Fissure veins. 1 ,, , , . . . . . , ,. Morphologic facies of a single genetic J 6. Secondary alterations. 7. Sedimentary ore deposits. 8. Detrital deposits. While this is a decided improvement upon the first classifica- tion adopted by Beck, the description of the various deposits shows that many genetically different types are forced into one and the same subdivision. Weed 1 goes further and gives the origin of the ore-forming solutions. His first class includes igneous deposits, segregated in a magma; his second, igneous emanations, including contact deposits, and tin veins; his third, gas-aqueous or pneumato- hydato-genetic deposits formed by magmatic waters mingled with ground waters. His fourth and smallest division includes those mineral masses formed by surface waters. This classification has not been generally accepted because it brings up the admit- tedly difficult separation of meteoric and magmatic water. The best genetic classification of mineral deposits would seem to be that according to geological processes. Mineral deposits must have been formed by igneous processes, alteration, cementa- tion, deformation, erosion, or sedimentation. Recognizing this, Van Hise 2 classifies ores as follows: Those produced (1) by proc- 1 W. H. Weed, In "Ore deposits," a discussion republished from the Eng. and Min. Jour., New York, 1903, pp. 20-23. 2 C. R. Van Hise, A treatise on metamorphism, Mon. 47, TJ. S. Geol. Survey, 1904. CLASSIFICATION OF MINERAL DEPOSITS 199 esses of sedimentation; (2) by igneous processes; (3) by meta- morphic processes, including under this heading practically all veins and allied geological bodies, conceiving them to be de- posited by the circulating ground water. It is probably impossible to produce a classification which will win the approval of all. In the ultimate analysis by far the larger number of mineral deposits have been formed by physico- chemical reactions in solutions, whether these were aqueous, igneous, or gaseous. According to this view the only con- sistent division that can be made is that between deposits formed by mechanical concentration of pre-existing minerals and those formed by reactions in solutions. A genetic classification should not be confined to a general indication of the relative time of ore deposition whether at the same time or later than the country rock. Nor should it confine itself to a statement of the agents of ore deposition whether aqueous, igneous, or gaseous solutions, or whether sedimentary, igneous, or metamorphic processes. The state- ment of the place of ore deposition at the surface or below it; in shallow waters or in deep seas is important but not sufficient. Some authors have attempted a classification by mode of deposition whether by replacement or by filling of open cavities but all such attempts have been failures, for the two processes are so closely associated that separation is impossible. The genetic classification should ultimately determine the limits of ore deposition in each class by temperature and pressure. Each deposit should be considered as a problem in physical chemistry, and the solution of this problem, with the necessary- geological data, will suffice to fix the mode of formation of the deposit. We are far from having the complete material for such a classi- fication, but we have at least a few starting points. It is neces- sary to determine, by experiment or by observation in nature, the limits of existence of each mineral species. Some will be found to be "persistent" under widely differing conditions of temperature and pressure like fluorite, quartz, or gold. For others a far more limited range will be established. By col- lecting the data of mineral association, sequence of deposition, and stability range of the component parts of the deposit it will be possible to ascertain the conditions prevailing at the time of ore deposition. 200 MINERAL DEPOSITS An absolutely consistent genetic classification is at present impracticable for its forces the geologist to take a definite stand on problems which, as yet, have not been solved. 1 Perhaps it is well not to expect too much from physical chemistry, magnificent as its services have been. The com- plications, even in simple systems, become great when, besides temperature and pressure, concentration, mass action, and time must be considered. In multicomponent systems the difficulty increases enormously. At the same time it is believed that the direction indicated is the only safe one to take in classifying the complex phenomena of ore deposition. OUTLINE OF PROPOSED CLASSIFICATION Detrital and Sedimentary Deposits. In the scheme followed in this book there are two major divisions. The first includes deposits formed by mechanical processes of concentration. This includes the detrital deposits such as placers and quartz sand formed at moderate temperature and pressure. The second division contains the great majority of mineral deposits which have been produced by chemical processes of concentration. Many important processes, such as those pro- ductive of iron ores and phosphates, for instance, take place by interactions of solutions in bodies of surface waters. These proc- esses may be of inorganic origin or they may take place through the medium of living bodies, almost always at moderate tem- peratures. The products are usually mingled with detrital matter. They may be enriched by secondary processes in the unconsolidated strata or by processes of weathering after their exposure to air. Another class of deposits is formed in bodies of surface waters by their evaporation and consequent precipitation of the salts dissolved in them; these are frequently termed the "saline resi- dues." Common salt, gypsum, and borates are among the sub- stances found in these deposits. Concentration of Substances Contained in the Rocks. Instead of at the surface or in bodies of surface waters the processes of concentration of useful substances may go on in the rocks themselves. We may distinguish two cases: the sub- 1 T. Crook, The genetic classification of rocks and ore deposits, Mineralog. Mag., London, vol. 17, 1914, pp. 55-85. CLASSIFICATION OF MINERAL DEPOSITS 201 stances were originally contained in the same geological body in which the deposit is found, or they may have been intro- duced from the outside. The apparent objection to this basis of subdivision, namely, the difficulty of deciding the source of the mineral or metal, is met in many cases by the knowledge acquired during late years. There may be deposits for which the qestion cannot be decided, but I believe that in the near future we shall in most cases have sufficiently good evidence. No one seriously maintains that the gold in the quartz veins of California, for instance, has been leached from the surrounding country rock, and surely no one denies that the oxidized nickel silicate ores of certain peridotites were originally contained in minute distribution in these rocks. In the case of substances contained in the geological body itself, the concentration may be effected by (1) rock decay and residual weathering that is, by oxygenated surface waters; (2) by the ground water of the deeper circulation; (3) by processes of dynamic and regional metamorphism, and (4) by zeolitization of surface lavas. Residual Weathering. Rock decay tends to destroy the rocks as units; to break them down, mechanically and chemically, and to re-assort their constituents in new combinations. In the decaying mass certain constituents are concentrated or precipi- tated; its detritus is swept away and deposited in rivers, lakes, and oceans; its soluble constituents are carried into the larger reservoirs of water and there perhaps precipitated in various forms. It is true that not quite all the sedimentary deposits are derived from the decaying rocks; the fossil coals are indirectly made from the carbon of the atmosphere; volcanic ashes con- tribute a share to the sediments; the exhalations of eruptive magmas, as well as ascending waters, contribute some dissolved matter from the lower part of the earth's crust. Processes of sedimentation and rock decay take place at moderate temperatures and pressures and the new minerals formed are, as a rule, characterized by high hydration. Below C. mineral deposits do not form, except in so far as freezing of water is retarded by rapid motion or dissolved salts. Few of the deposits have been formed at temperatures above 50, and this only exceptionally during eruption, evaporation in shallow desert lakes, or oxidation of pyritic rocks. The pressure is in 202 MINERAL DEPOSITS general little different from that of the normal atmosphere, but in deposits of deep seas or lakes considerably higher pressures prevailed. This increased pressure, at low temperature, appears to have had little influence on the mineral associations formed. Deep Circulating Waters. Under the influence of the ground water of the deeper circulation many ore deposits are formed, concerning some of which there may be room for differing opin- ions. Copper may be leached from greenstones and the ores of the metal may be deposited in veins in the same rock. Hema- tite, like that of the Lake Superior region, may be concentrated from the surrounding low-grade "iron formation." Barite, magnesite, and sulphur are other instances. Regional Metamorphism. Again, the agency may be meta- morphism under stress or regional metamorphism; in such case the change takes place with very little water and it is not con- sidered probable that a great concentration, say of the metals contained, can be effected. Other materials may form, such as slate from shales, or useful minerals like garnets or graphite may develop in the rock. During static metamorphism, temperature and pressure are likely to be somewhat higher than at the surface. Regional metamorphism takes place under heavy pressure and at fairly high temperatures at great depth. It may merge into igneous metamorphism. Zeolitization. The processes of zeolitization take place shortly after the consolidation of an igneous rock by the aid of residual magmatic water or of surface water. Under certain circum- stances a concentration of metals can be effected by this process, of which the copper deposits of Lake Superior offer an excel- lent instance. Introduced Ores not Connected with Igneous Rocks. Much more common is the case where the valuable minerals have been introduced into the rock from without, and to this class belong the majority of the metal deposits. Deposits of this kind occur along fissures or form replacements along fissures or are found in general where opportunity is offered for vigorous circulation of the depositing waters. For a long time it was held by many that the metallic contents of fissure veins were derived from the sur- rounding rock, but it is now generally admitted that such a view in most cases is erroneous. Certain metallic ores occur entirely independent of igneous rocks; the mineral associations in these indicate a deposition at CLASSIFICATION OF MINERAL DEPOSITS 203 moderate pressure and temperature, the latter probably rarely reaching 100 C. Of this kind are certain lead-zinc deposits in limestone or the copper deposits in sandstone which are so com- mon in various parts of the world. Most geologists agree that such deposits have been formed by surface waters, at moderate depths,' and that the metals have been leached from neighboring strata and, after a comparatively short wandering, deposited in fractured rocks in their present resting places. These deposits are generally poor in gold and silver. Deposits Genetically Connected with Igneous Rocks. There is also another and larger class which appears only in or near igneous rock and whose epoch of formation usually can be shown to have followed closely after the eruption. This class has been clearly recognized by almost all geologists. There is also general agreement that these deposits have been laid down by heated, ascending waters, although there is no unan- imity L as to the source of the water or the source of the metal. To some the water and the dissolved metals are simply igneous emanations from a cooling magma; to others the waters are of atmospheric origin and, heated by their passage through still warm igneous rocks, have dissolved the metals contained in them. Nearly all metal deposits of the American Cordilleran region belong to this division. It is subdivided into several groups, according to the evidence of mineral association and geological relations. The first group includes ores deposited at slight depth below the surface; the temperature is here relatively low, perhaps from 50 to 150 C., and the pressure will scarcely ex- ceed 100 atmospheres. Examples of this group are found in the gold and silver veins, of Tonopah, Nevada, the Cripple Creek gold telluride veins, and the California quicksilver veins. A second group comprises the deposits formed by hot ascending waters at moderate depths, say from 5,000 feet to 10,000 feet be- low the surface, at temperatures of perhaps from 150 to 250 C. and correspondingly increased pressure. The present outcrops are exposed by deep erosion and they almost always appear in or close to intrusive bodies. As examples may serve the gold- quartz veins of California and the metasomatic pyritic deposits of Leadville. A third, deep-seated group includes veins and contact-meta- morphic deposits. During the genesis of these the temperature 204 MINERAL DEPOSITS was high, but in most cases below 575 C., the crystallographic inversion point for quartz. The pressure was probably very high. The cassiterite veins, some gold-quartz veins of the Appalachian type, and the tourmaline-copper veins belong in this group, which with great confidence may be ascribed to emanations from magmas. The deposits unquestionably 'formed by direct igneous emanations are the contact-metamorphic ores appearing in limestone along igneous contacts. They contain oxide ores, such as magnetite and specularite, together with sulphides of copper, zinc and iron, and present an association of other minerals characteristic of contact metamorphism The emanations from effusive bodies are deposited as subli- mates of little economic importance. Products of Magmatic Differentiation. The last class is that of the deposits formed by concentration in igneous magmas; of all types these have formed at the highest temperature and pressure. They include oxides or sulphides segregated in the magmas, like the iron ores of Kiruna in northern Sweden, the titanic iron ores of the Adirondacks, or the copper-nickel ores of Sudbury. They also include the pegmatite dikes, which contain many gems and rare metals and which are regarded as segrega- tions from cooling granitic magmas. The pegmatites were formed at comparatively low temperatures probably from 500 to 800 C. but during the differentiation of the other deposits mentioned considerably higher temperatures probably prevailed. The pressure must, of course, have been very high. Metamorphism and Surface Enrichment of Deposits. Tn the proposed classification the mineral deposits are supposed to have suffered no change from their original condition. This is of course rarely strictly true, for chemical changes as a rule begin soon after the cessation of the agency which caused the deposition. In sedi- mentary beds this is particularly the case, for cementation and hardening and various chemical actions begin almost from the time of deposition. It is, however, not the custom to refer to these changes as metamorphism. Many mineral deposits have undergone great changes from their original conditions. They may have been reached by igneous metamorphism and thus a coal bed transformed into anthracite or a bed of limonite into magnetite. Or they may have, been sheared or crushed during regional metamorphism Or, most common of all cases, they may have been altered by CLASSIFICATION OF MINERAL DEPOSITS 205 surface waters. Such oxidizing surface waters, as well as similar waters at somewhat greater depth, when they have parted with their free oxygen, produce peculiar modifications and often most important enrichments. A CLASSIFICATION OF MINERAL DEPOSITS 1 I. Deposits produced by mechanical processes of concentration. (Tempera- ture and pressure moderate.) II. Deposits produced by chemical processes of concentration. (Tempera- ture and pressure vary between wide limits.) A. In bodies of surface waters. 1. By interaction of solutions. a. Inorganic reactions. b. Organic reactions. 2. By evaporation of solvents. B. In bodies of rocks. 1. By concentration of substances contained in the geological body itself. a. Concentration by rock decay and residual weathering near surface. b. Concentration by ground water of deeper circulation. c. Concentration by dynamic and regional metamorphism. d. Zeolitization of surface lavas. Temperature, to 70 C. Pressure, moderate to strong. C. + Temperature, 0-100 C Pressure, moderate. Temperature, 0-100 C. Pressure, moderate. Temperature up to 400 C. + Pressure, high. Temperature, 50-300 C. + Pressure, moderate. 2. Concentration effected by introduction of substances foreign to the rock. a. Origin independent of igneous activity. By circulating atmospheric waters at moderate or slight depth. b. Origin dependent upon the eruption of igneous rocks. a. By hot ascending waters of uncertain origin, but charged with igneous emanations. 1. Deposition and concen- tration at slight depth. 2. Deposition and concen- tration at intermediate depths. 3. Deposition and concen- tration at great depth or at high tempera- ture and pressure. Presented before the Geological Society of Washington, May 10, 1911. Temperature, to 100 C. Pressure, moderate. Temperature, 50-150 C. Pressure, moderate. Temperature, 150-300 C. Pressure, high. Temperature, 300-500 C. Pressure, very high. 206 MINERAL DEPOSITS b. By direct igneous emanations. 1. From intrusive bodies. Con- f Temperature, probably tact metamorphic deposits j 300-800 C. and allied veins; [ Pressure, very high. 2. From effusive bodies. Sub- limates, fumaroles. Temperature, 100-400 C. Pressure, atmospheric to moderate. C. In magmas, by processes of differentiation. a. Magmatic deposits proper. Temperature, 700-1500 C. + Pressuse, very high. 6. Pegmatites. Temperature, about 575 C. . Pressure, very high. CHAPTER XV DEPOSITS FORMED BY MECHANICAL PROCESSES OF TRANSPORTATION AND CONCENTRATION; DETRITAL DEPOSITS , INTRODUCTION Weathering tends to destroy rocks and mineral deposits by disintegration and chemical decomposition. In part, new minerals, like kaolin and limonite, form; in part, the more resist- ant minerals, like quartz, gold, platinum, magnetite, cassiterite, and garnet, are set free in individual grains. Erosion now steps in and the detritus is swept down the slopes and into the water channels. Mechanical separation in running water or along sea or lake beaches sorts the detritus according to specific gravity and size of grains. The heaviest particles, as those of gold, magnetite, and garnet, tend to collect in the lower part of the assorted detritus; the minute and easily moved scales of clayey substance are carried far but ultimately deposited as sedimentary beds; the colloids are coagulated by the electrolytes in the sea water. DETRITAL QUARTZ DEPOSITS The quartz grains are often accumulated as beds of almost pure quartz sands. These are used extensively as ingredients in pottery and glass, also for abrasive purposes in sawing soft rocks, such as marble. Such sands should contain 99 per cent, silica. Somewhat argillaceous quartz sands without carbonates and carrying 80 to 90 per cent, silica are used as molding sands and are mined on a large scale, though occurring in thin beds. 1 When compacted by pressure and by cementation the quartz sands are transformed into siliceous sandstones and quartzites and these are used for millstones, whetstones, and grindstones. Comparatively few Realities furnish good material. 1 E. F. Burchard, Requirements of sand and limestone for glass-making, Bull. 285, U. S. Gteol. Survey, 1906, pp. 473-475. L. Heber Cole, The occurrence and testing of foundry molding sands, Trans., Canadian Min. Inst., vol. 20, 1917, pp. 265-291. 207 208 MINERAL DEPOSITS With the development of modern methods of grinding the im- portance of millstones has greatly decreased. Technical and statistical information on these subjects is contained in Mineral Resources of the United States, 1916, part 2, under "Abrasive Materials" and "Sand and Gravel" and for 1916, p. 634, where a list of literature may also be found. In case of very fine-grained whetstones a doubt may exist whether the material is of detrital origin or formed by chemical agencies. The so-called novaculite of Arkansas, the best whet- stone known, is a good example of this. It occurs in the Silurian beds of Garland and Saline counties in that State, and is used for what are known, according to color and quality, as Washita and Arkansas stones. The latter are snow-white and are the harder. The rock is much jointed and only small pieces are obtainable. Branner considers this material a metamorphosed chert, while Griswold 1 believes it to be a fine-grained sediment. DETRITAL CLAY DEPOSITS 2 The fine material resulting from the decay of rocks is carried away, suspended in water, and deposited in river beds, lakes, and seas as sedimentary clay. The nature of clays is a much discussed subject. Perhaps the best definition is given by G. P. Merrill, who says 3 that the clays are widely diverse in origin and in mineral and chemical composition but have the common prop- erty of plasticity when wet and that of induration when dried. Clays are finely comminuted aggregates of hydrous aluminous silicates, detrital quartz and other mineral fragments, often also, iron hydroxide and calcic and magnesic carbonates. The sedi- mentary clay is therefore to be regarded rather as a rock than as a mineral and its principal use is for structural purposes; The detailed description of these deposits, therefore, does not fall within the scope of this book. The larger part of the clays are derived from decomposition 1 L. S. Griswold, Whetstones and the novaculites of Arkansas, Ann. Rept. Arkansas Geol. Survey, vol. 3, 1890. 2 For more details in regard to the important clay industry the reader is referred to H. Ries, Clays, New York, 1908. Information as to pro- duction, etc., is given in Mineral Resources of the United States, published annually by the U. S. Geol. Survey. Further notes regarding residual kaolin deposits may be found on pp. 325-328. 3 G. P. Merrill, Rocks, rock-weathering, and soils, New York, 1897, p. 135. TRANSPORTATION AND CONCENTRATION 209 and hydration of feldspathic minerals; other silicates, however, contribute their share. It has been supposed that the mineral kaolinite (H 4 Al 2 Si 2 O 9 ). is one of the principal constituents of clay. Probably it is present because the formation of kaolinite from feldspars can easily be traced in decomposing rocks at the surface, but in the clays the mineral is so comminuted that it cannot be readily identified. It is known that colloid hydrous silicates of aluminum exist and there are also a number of more or less indefinite compounds of this kind in nature, such as halloy- site, smectite, and pholerite. The sedimentary clays rarely approach kaolinite in composition. Kaolinite should contain 46.5 per cent. Si0 2 , 39.5 per cent. A1 2 O 3 , and 14 per cent. H 2 O; but by reason of admixture of quartz and undecomposed silicates, the sedimentary clays usually contain much more silica than the amount indicated. Clays without carbonates generally contain more magnesium than calcium, and potassium exceeds sodium. Titanium often exceeds one per cent. Much of the titanium and potassium is probably present in colloid state. Traces of copper, nickel, lead, zinc and vanadium are sometimes found. Regarding residual clays derived from the decomposition of rocks in place see p. 325. The clays formed by the action of sulphuric acid on silicates in the oxidized part of ore deposits are described on p. 480. FULLER'S EARTH ' Fuller's earth is the name given to certain sediments of clay- like material, originally used in England by fullers for cleansing cloth of grease. At present this substance is extensively used for deodorizing, decolorizing and clarifying fats and oils; much of it is employed in the refining of petroleum. Its_value thus depends upon its adsorbent qualities. The material occurs in sedimentary beds of Mesozoic, Cenozoic, 1 J. T. Porter, Properties and tests of fuller's earth, Bull. 315, U. S. Geol. Survey, 1907, pp. 268-290. T. W. Vaughan, Fuller's earth of Florida and Georgia, Bull. 213, U. S. Geol. Survey, 1903, pp. 392-399. Mineral Resources, U. S. Geol. Survey. Annual publication. Articles by F. B. Van Horn and J. Middleton. Charles L. Parsons, Fuller's earth, Bureau of mines, Bull. 71, 1913. E. H. Sellards and H. Gunter, Second Annual Report, Florida Geol. Survey, 1908-09, pp. 255-290. 210 MINERAL DEPOSITS and Quaternary age, but a similar material is also derived from the weathering of basic igneous rocks. Microscopic examina- tion gives little evidence of its origin; in color it ranges from gray to dark green; it possesses little or no plasticity. The chemical analysis also has little value in determining its quality. J. T. Porter believes, and probably justly, that the material owes its quality to the adsorbent power of colloid hydrous aluminium silicates. The analyses show that the silica varies between 47 and 75 per cent., alumina from 10 to 19 per cent., lime from 1 to 4 per cent., magnesia from 2 to 4 per cent., ferric oxide from 2 to 10 per cent., and combined water from 5 to 21 per cent. In Gadsden County, Florida and Decatur County, Georgia, it occurs in Tertiary strata and is mined in open pits; in Arkansas it is obtained from weathered basic dikes. The further prepara- tion includes drying, grinding and bolting to sizes from 30 to 100 mesh per inch. Very fine material clogs the filter presses. The domestic production amounts to about 70,000 short tons and about 17,000 tons were imported. Florida yields most of the total domestic production. The price of the Florida material is about $10 per ton. PtACER DEPOSITS Origin and Distribution. The heavier and less abundant minerals in the rocks are the most resistant to decomposition and when the weathered rock is eroded and sorted by water they usually become concentrated in the lower parts of the sand and gravel beds. The gold-bearing gravels, which form an important source of supply of this metal, were called placers 1 by the early Spanish miners of this continent, and this name is probably the best that can be adopted for deposits of this class. Instead of gold the valuable mineral may be cassiterite, magnet- ite, monazite, diamonds, or other precious stones. Other terms have been employed, as "gravel deposits" or " gold-bearing gravels," or "alluvial deposits"- all equally objectionable, for the material may be sand instead of gravel, and it may be deposited along the ocean beach instead of in watercourses. 1 Derivation uncertain: Placer, pleasure; Plaza, place. Stelzner (Die Erzlagerstatten, p. 1261) says placer is a local Spanish term for sand bank. The Germans use "Seife," meaning washings. In French the word "allu- vions" is often used. TRANSPORTATION AND CONCENTRATION 211 The processes of erosion and concentration have been active since earliest geologic time, and so we may have detrital depos- its or placers of differing ages. Land deposits are, however, usually thin and easily removed and thus placers of pre-Tertiary age are comparatively rare. In the formation of placers nature simply employs in her own leisurely way the processes of crushing and concentration which we use in ore dressing. The rocks are broken and com- minuted by the expansion due to alternating heat and cold; by the growth of plants; or by the impact of sliding and water- carried rocks; or by the grinding action of ice; or finally by chemical decomposition and hydration. The products are con- centrated in water courses or along shores by running water or in ocean currents by motion similar to that on tables and jigs. Spherical particles of different substances fall in water at a rate proportional to their weight divided by the resistance. As the resistance is proportional to the area exposed, a fragment of" quartz the size of a pea will fall much more slowly than a piece of gold of the same size. It will in fact be carried along easily in a current of water in which a piece of gold of the same size will sink instantly. Thus the specific gravities of the valuable minerals play a prominent part in the formation of placers. The specific gravity of the more important substances is as follows: Quartz, 2.64; feldspar, 2.55 to 2.75; ferromagnesian silicates, 2.9 to 3.4; garnet, 3.14 to 4.13; diamond, 3.54; corun- dum, 4.0; monazite, 5.0; magnetite, 5.0; cassiterite, 6.4 to 7.1; gold, 15.6 to 19.33; platinum, 14.0 to 19.0 (21 to 22 when chemically pure). The shape of the particles is also of importance. Flaky minerals, like molybdenite, scaly gold, or specularite, are difficult to concentrate in spite of their high specific gravity. GOLD PLACERS Introduction. Gold is the most important placer mineral. Roughly speaking, about $70,000,000 out of a world's pro- duction of about $450,000,000 are derived from Tertiary or Quaternary placer deposits; discoveries in Alaska and the North- west Territory have lately increased the output. Gold placers as a rule are easily discovered and worked; the supplies of old and long-settled countries were generally long ago exhausted. Bo- 212 MINERAL DEPOSITS hernia, Italy, Spain, and Hungary, now almost barren of placers, once furnished their share. New deposits are usually discovered on the outskirts of civilization, as in Brazil in the eighteenth century, in Australia and California during the middle of the last century, and in Alaska and Siberia to-day. The production of placer gold in the United States, including Alaska, in 1897 was $7,800,000; in 1916 it was $22,882,000, the increase being due to the recently discovered placers of Alaska and to the development of the dredging fields in California. Practically all this gold comes from Quaternary and Tertiary placers, some dating back as far as the Eocene. A small quantity is obtained from Cretaceous conglomerates in Oregon and north- ern California. Permian gold-bearing conglomerates occur in Bohemia, according to Posepny. 1 Permo-Carboniferous con- glomerates containing detrital gold have been described by Wilkinson from New South Wales. 2 In most cases the gold content of these older conglomerates is small and they can rarely be profitably worked. Probably the best example of ancient placers is furnished by the Cambrian basal conglomerate of the Black Hills, South Dakota, which unconformably covers the pre-Cambrian schists and gold-bearing quartz veins. It was first described by W. B. Devereux 3 and later by J. D. Irving. 4 This conglomerate, which is from 2 to 30 feet thick and is overlain by quartzite, carries in places much gold of unquestionably detrital origin, as indicated by the rounded grains, and has been profitably worked in several mines. The gold was derived from the erosion of auriferous lodes in the pre-Cambrian rocks and was deposited in depressions along the old shore line. In part the gold-bearing conglomerate is cemented by pyrite, which probably also contains some gold. Maclaren 5 believes that the scarcity of economically important deposits of detrital gold in older forma- tions is due to its solution, in depth, by alkaline solutions. There is little evidence in support of this view. Origin of Placer Gold. In primary deposits gold is mainly contained in veins, lodes, or shear zones and these appear in 1 Genesis of ore deposits, 1902, p. 163. 2 Idem, p. 162. 3 W. B. Devereux,' Trans., Am. Inst. Min. Eng., vol. 10, 1882, pp. 465-475. 4 J. D. Irving, Economic resources of the northern Black Hills, Prof. Paper 26, U. S. Geol. Survey, 1904, pp. 98-111. 6 J. M. Maclaren, Gold, London, 1908, p. 90. TRANSPORTATION AND CONCENTRATION 213 rocks of many different kinds. It is often stated that gold is distributed as fine particles in schists and massive rocks and that placer gold in certain districts is derived from this source. Most of these statements are not supported by evidence, though it is not denied that gold may in rare instances be distributed in this manner. Even in the Yukon region, concerning which such statements have often been made, the origin of the gold from veins, lodes, and shear zones is beginning to be recognized. 1 The great majority of gold placers have been derived from the weathering and disintegration of auriferous veins, lodes, shear zones, or more irregular replacement deposits. These primary deposits were not necessarily rich and may not be profitable to work. In many regions the rocks contain abundant joints, seams, or small veins in which the gold has been deposited with quartz. Eluvial Deposits. Gold placers may be formed by rapid erosion of hard rocks, but such placers are not often rich and highly concentrated. In the great placer regions the concen- tration has generally been preceded by an epoch of deep secular decay of the surface. It has been supposed by many that this deep rock decay is peculiar to the tropics, but this is not correct. The process has been active in the southern Appalachian States, in California, and even in Alaska, as well as in countries like the Guianas and Madagascar. When the outcrops of gold-bearing veins are decomposed a gradual concentration of the gold follows, either directly over the primary deposits or on the gentle slopes immediately below. The vein when located on a hillside bends over (Fig. 73) and disintegration breaks up the rocks and the quartz, the latter as a rule yielding much more slowly than the rocks; the less resistant minerals weather into limonite, kaolin, and soluble salts. The volume is greatly reduced, with accompanying gold concentration. The auriferous sulphides yield native gold, hydroxide of iron, and soluble salts. Some solution and redeposition of gold doubtless take place whenever the solutions contain free chlorine. The final result is a loose, ferruginous detritus, easily washed and containing easily recovered gold. This gold consists of grains of rough and irregular form and has a fineness but slightly greater than that of the gold in the primary vein. Stelzner has applied to such residual concentrations, which may be worked like ordinary 1 A. H. Brooks, The gold placers of parts of Seward Peninsula, Alaska, Bull. 328, U. S. Geol. Survey, 1908, pp. 108 et seq. 214 MINERAL DEPOSITS placers, the term eluvial gold deposits. Their occurrence is illustrated in Fig. 73. In the gold region of the southern Appalachian States the decomposition of the country rock, which generally is a schist, may reach a depth of 100 feet or more. 1 The decomposed material of the auriferous veins slides downhill, mixing with the weathered rock, and during this process the gold in part sinks deeper into the detritus. This has given rise to a peculiar system of mining by which the whole mass is washed by the hydraulic method and the more resistant quartz boulders crushed in a stamp mill with coarse mesh. This has been practiced at Dahlonega and is often called the Dahlonega system. Similar FIG. 73. Diagram showing development of eluvial and stream placers. deposits were worked in California, particularly in Eldorado county, and are here called "seam diggings" from the fact that the gold occurs disseminated in quartz seams traversing a certain belt of schists. Such deposits frequently occasion legal contests owing to the uncertainty whether they should be considered as placers or as mineral-bearing veins. In certain regions of Brazil 2 the schists and gneisses are covered by auriferous detritus accumulated in place. Another example is the "Tapanhoancanga" of the same country. This is a bed of residual or lateritic iron ore up to 10 feet thick covering the 1 G. F. Becker, Reconnaissance of the gold fields of the southern Appala- chians, Twenty-sixth Ann. Rept., U. S. Geol. Survey, pt. 2, 1895. 2 O. Derby, Peculiar modes of occurrence of gold in Brazil, Am. Jour. Sci., 3d ser., vol. 28, 1884, p. 440. O. Derby, Notes on Brazilian gold ores, Trans., Am. Inst. Min. Eng., vol. 33, 1892, pp. 282-283. TRANSPORTATION AND CONCENTRATION 215 underlying hematite schist and containing gold throughout. The gold probably occurs in veinlets in the schists and the gold-bearing detrital material has been concentrated from a considerable thickness of schist weathering in place. Excellent examples of eluvial deposits are reported from Dutch, British, and French Guiana, 1 though ordinary stream placers are the most common deposits in these countries. Over a great part of this gold-bearing territory secular decay of crystalline rocks has resulted in a deep mantle of ferruginous clayey earth laterite and in places the gold has been concen- trated in this material below outcrops of gold-bearing veins. Many of the stream beds are also worked for placer gold, the detritus usually resting on the clayey surface of the compact laterite. It is stated that many rocks in the Guianas centain gold and that the placer gold is derived from such material; particularly are the basic rocks, diabases and amphibolites, said to be aurif- erous. This conclusion should probably be accepted with some reserve. It seems more probable that the gold contained in the greenstones is of secondary origin and that here, as elsewhere, granitic intrusions have caused the formation of a series of gold-bearing veins in the surrounding rocks. Processes of Concentration. In most cases the cycle has been carried further and the material is not only decomposed, but eroded, transported, and redeposited. This can be effected by wind, by streams, or by the surf of the sea. Eolian Deposits. Deposits concentrated by eolian agencies can, of course, be formed only in dry countries where long sub- aerial decay has paved the way for the work of the dust' storms; 1 C. G. Dubois, Geologisch-bergmannische Skizzen aus Surinam, Freiberg i. S., 1901, pp. 112. C. G. Dubois, Beitrage zur Kenntnisa der surinamischen Latent, etc., Tsch. Min. u. petr. Mitt., 22, 1903. E. D. de Levat, The gold fields of French Guiana, Mineral Industry, vol. 7, 1899. E. D. de Levat, Guide pratique etc. de 1'or en Guyane francaise, Paris, 1898. A. Bordeaux, Trans., Am. Inst. Min. Eng., vol. 41, 1910, pp. 567-593. J. B. Harrison, The geology of the gold fields of British Guiana, London, 1908. J. B. Harrison, in the Reports of the Instit. of Mines, British Guiana. E. E. Lungwitz, Die Goldseifen von British Guiana, Zeitschr. prdkt. GeoL, 1900, pp. 203-218. 216 MINERAL DEPOSITS from the decomposed and crumbled outcrops of the lodes the winds blow away the lighter sand, leaving a mass of coarser detritus which contains the gold. Such wind-born placers have been noted by H. C. Hoover 1 and T. A. Rickard 2 near the crop- pings of the West Australian gold veins. No examples of this kind are known from the Cordilleran States of America. Stream Deposits. Running water is by far the most important agency in the formation of gold placers. First of all, attention must be directed to the high specific gravity of gold, which explains many of the puzzling features of the placers. Placer gold is six or seven times as heavy as the most common accom- panying minerals feldspar and quartz and it settles to the bottom in flowing water with surprising rapidity. It is almost impossible to lose a particle of gold, of the value of one cent, in a miner's pan; it sinks immediately to the bottom of the gravel and sand after one or two preliminary shakes in water. Once lodged at the bottom it stays there, in spite of shaking and rotating. This illustrates the fundamental fact that (pie) gold is mainly on the bed-rock. The rapid settling of the gold accounts for the partial failure of some devices for placer mining, particularly the clam-shell and the suction dredges. The ease with which some concentration, according to the specific gravity, is effected is shown by the well-known fact that in powdered samples of ore, as well as in dumps at the mine, a settling of the heavier ore particles toward the bottom can often be observed. Suppose we have a gold-bearing quartz vein deeply altered by rock decay; now let the region be raised say 500 feet by one of these slow oscillations which so commonly affect the crust. A river has excavated a valley to the corresponding depth in this elevated plateau, and this valley under the influence of a pause in the elevating movement becomes filled with gravels to a width of about 100 feet. Let a tributary gulch with steep grade be cut back into the plateau to the gold deposit (Fig. 74) ; when the gulch reaches it the eluvial deposit will be carried down by sliding and washing; the clay and limonite are rapidly removed in sus- pension; the angular gravel of quartz and rock, grinding the 1 H. C. Hoover, The superficial alteration of Western Australian ore deposits, Trans., Am. Inst. Min. Eng., vol. 28, 1899, pp. 762-763. 2 T. A. Rickard, The alluvial deposits of Western Australia, idem, pp. 480-537. TRANSPORTATION AND CONCENTRATION 217 fragments of gold between them and on the bed-rock, will be moved downward, the fine grains in suspension, the coarser ones dragging and rolling on the bottom. There is little deposition; the transporting power is great and in flood time the whole gravel mass, of no great depth, will probably be in motion. Heavy gold nuggets may lodge in the lee of little ridges. The gold settles rapidly; most of it, " continually hammered ''and slowly shaping itself in flat, smooth grains, will be dragged down stream and finally reach the edge of the flood-plain in the river. 100 200 300 FEET CONTOUR INTERVAL 50 FEET Strongly Aurilerous Gravel (on Bedrock) FIG. 74. Plan of quartz vein and placers below it, illustrating the develop- ment of pay streaks. At this place the larger part of the gold stops. It is not washed out with the sand and gravel but stays on the bed-rock near the margin. The finer particles will, of course, be carried out a little distance, but they soon sink into the water-filled gravel after the manner of grains of heavy ores in concentrating jigs. Just as in the gulch the whole mass of detritus is transported, so it is thought that in larger streams the body of water-soaked gravel and sand works downstream very slowly. During this process the lighter gold contained in the detrital material also works for- 218 MINERAL DEPOSITS ward and downward, gradually joining the nuggets or coarser pieces, which have already reached their final resting ground. This mode of operation contains the key to the genesis of the placers. It is not to be expected that the coarse and ordinary fine gold will be carried out into the middle of wide flood-plains. As the flood-plain widens it will cover the accessions of gold along its margin, and the final result will be a streak of rich gold-bearing gravel, resting on the bed-rock and extending downstream deep underneath the surface. When this is traced upstream the primary deposit, the vein, will be found. The actual occurrences of course show infinite variation. Let us assume that, as happens in the Creswick district in Victoria, Australia, a broad stream with moderate grade crosses a deeply decomposed belt of soft slate containing an abundance of small veins or stringers of quartz with native gold, and that in addition a fair balance between transportation and deposition persists for a long time. The result will be a gravel deposit, only a few feet deep, but with an abundance of gold concentrated on the bed-rock over the whole width of the stream. Each freshet is sufficient to churn up and move forward the whole mass of gravel, continually adding to the concentrated gold on the clayey bed-rock. Again, we may assume extremely active erosion, as is the case in the Sierra Nevada of California. Canyons several thousand feet in depth have been cut in an uplifted plateau, veritable trenches or sluice boxes, the grade of which is from 60 to 150 feet per mile. Stretches of wild gorges with polished bottoms alternate with stretches of less grade where shallow gravel ac- cumulates. These canyons receive for long distances an abun- dant supply of gold of all sizes from older hill gravels or from decaying quartz veins. The result will be that but little gold will lodge in the gorges, while extremely rich shallow gravel bars will accumulate in the convex stream curves (Fig. 75). Gradient, volume, and load usually vary in the same stream so that deposition may be going on in one part of its valley and erosion in another. Continued corrasion of the stream-bed results in deepening the canyon and leaving the bars as elevated benches. The miners of 1849 first found these bars and worked them. In searching for the source of the gold they soon found a trail of metal leading up the gulches to great masses of older gravels on the hills, 2,000 to 3,000 feet above. These gravels TRANSPORTATION AND CONCENTRATION 219 were washed by the hydraulic method; immense masses of tailings with a little gold were carried down to the rivers, totally overloading them. After the prohibition of hydraulic mining the streams gradually resumed active transportation. The whole gravel mass moved slowly downstream and a gradual recon- centration on the bed-rock took place. The tailings deposited became enriched and will ultimately be reworked. 1 The torrential floods of the canyons scarcely permitted the lodgment of fine gold. This was swept out through the narrow FIG. 75. Low gravel bars, American River, California, showing placer de- posits on inner side of bends. After R. L. Dunn. portals into the Sacramento Valley, where the grade of the streams suddenly diminishes. The most minute particles may have been carried as far as San Francisco Bay, but the bulk of the fine gold lodged in the flood-plains within a few miles of the mouth of the canyons. Easily caught upon the clayey "false bed-rock" of a volcanic tuff, this gold, the average particles of which are about 0.3 millimeter in diameter, formed meandering pay streaks at the base of a sandy gravel bed from 10 to 60 feet in depth. Such deposits are now worked by dredging. By an odd paradox, gold is at the same time the easiest and the most difficult mineral to recover. It is divisible to a high 1 G. K. Gilbert, Hydraulic-mining debris in the Sierra Nevada, Prof. Paper 105, U. S. Geol. Survey, 1917. 220 MINERAL DEPOSITS degree and owing to its insolubility the finest particles are pre- served. A piece of gold worth one cent is without trouble divisible into 2,000 parts, and one of these minute particles can readily be recognized in a pan. In extreme subdivision the gold acquires a scaly, flat form, being known as flour gold or flake gold, is carried away very readily by water, and does not sink easily in sand or gravel. In part the flour gold is suspended by air films, and it can be carried away in rivers of moderate grade for hun- dreds of miles. The gold occurring in the sand bars of Snake River, Idaho, is a good example of this. 1 It will settle in thin pay streaks at bars and other favorable places, but the next freshet will probably destroy the sand bars and sweep the gold away. This accounts also for the distribution of fine gold in great masses of gravel beds for example, in the wash 600 feet thick deposited by glacial streams at Tacoma and other places on Puget Sound. Almost every pan of this gravel will show a "color," but the material contains only a fraction of a cent per cubic yard. The fine colors along the Columbia River in northeastern Washington range in value from less than 0.0005 to 0.02 cent, the average being about 0.002 cent. 2 The much-discussed concentration of gold on the bed-rock seems, then, to be due partly to the natural jig-like movement in moderately deep gravels, 3 during long-continued conditions of fair balance between loading and erosive power; partly to slow forward and downward motion of heavier gravel masses, 4 of which exact measurement as yet is lacking, and last and largely to the fact that heavier gold will not be carried out into the gravel flats of rivers of gentle grade the only ones that have extensive flood-plains but is immediately deposited on the marginal bed-rock of the gradually deepening and widening gravel The best conditions for the concentration of gold are found injnoderately hilly countries where deep secular decay of rocks has been followed by slight uplifts. Subsequent slight elevations would easily produce re-sorting and enrichment of the gravels. In regions of gold placers the richest material is usually pro- 1 J. M. Hill, Gold of the Snake River, Bull. 620, U. S. Geol. Survey, 1916, pp. 271-294. 2 A. J. Collier, Bull. 315, U. S. Geol. Survey, 1907, p. 61. 3 F. Posepny, Genesis of ore deposits, New York, 1902, p. 154. 4 T. A. Rickard, Min. and Sci. Press, Aug. 15, 1908. TRANSPORTATION AND CONCENTRATION 221 duced by repeated reworking of gold-bearing gravels by nature. Each reworking increases the richness of the gravels, eliminates easily decomposed pebbles, and finally results in a gravel of the hardest, most resistant rock quartzite or quartz. Quartz is the common gangue mineral in gold regions; hence the preva- lence of "white gravels" or "white channels," almost exclusively composed of white quartz pebbles. CLASSIFICATION OF FLUVIAT1LE AND MARINE PLACERS According to their occurrence the placers may be conveniently divided as follows. 1 PLACERS CLASSIFIED Present topographic cycle Past cycles, elevated Past cycles, depressed 1. Gulch and creek i 1. High creek gravels. 1. Deep creek gravels, gravels. 2. River and b gravels. 3. Gravel plains. 4. Beaches. Bench gravels. 2. \ Hill gravels or high river gravels. 3. Elevated gravel plains. 4. Elevated beaches. 2. Deep river gravels. 3. Depressed gravel plains. 4. Depressed beaches. -^Examples of present gulch, creek, and river gravels are not difficult to find; they occur in all gold-bearing regions where erosion is active and where precipitation is abundant enough to cause the sorting and carrying forward of the gravels in the stream beds. In the upper parts of the stream courses the gravel will be coarse and semiangular; in the lower parts the sands increase and the pebbles are smoother. Where the rivers emerge from their narrow valleys and spread with gentle grade over flood-plains, more extensive sand and gravel beds will accumulate, generally, however, with less gold than in the more confined part of the course. Some of the fine gold may reach the sea and is concentrated by the surf and the oblique shore currents into thin pay streaks on the sandy beach. 1 See also A. H. Brooks, The gold placers of parts of Seward Peninsula, Alaska, Bull. 328, U. S. Geol. Survey, 1908, p. 115. 222 MINERAL DEPOSITS Marine Placers. Beach placers occur along many shores and arc often produced by concentration from a sea bluff or elevated gravel plain. The beach at Nome, Alaska (Fig. 76), is a narrow strip about 200 feet wide, from which over $2,000,000 in fine gold has been washed ; the flaky gold averaged 70 or 80 colors to the FIG. 76. Diagrammatic section illustrating development of beach placer. After A. J. Collier and F. L. Hess, U. S. Geol Survey. cent. 1 ^Two older elevated beach lines are found farther inland. The beach gold of the Oregon and California coasts is much finer, the colors ranging from 100 to 600 to the cent. FIG. 77. Gold dredging on the Solomon River, Alaska. After P. S. Smith, U. S. Geol. Survey. Buried Placers. Subsidence or overloading may cause the placers to be deeply covered by barren detritus. Many of the streams of Alaska, particularly in their lower reaches, are thus covered; the process of concentration is stopped, the present 1 A. J. Collier and F. L. Hess, Bull. 328, U. S. Geol. Survey, 1908, pp. 140-228. TRANSPORTATION AND CONCENTRATION 223 watercourses having insufficient grade to effect the transporta- tion of detritus. Fig. 77 shows the dredging operations on the Solomon River, Alaska. The depth of the gravel in the river bottom is about 20 feet. Fig. 78 shows a diagrammatic section of the Oroville dredging ground, Butte County, California. The depth of the gravel is about 30 feet. At Fairbanks, Alaska, according to Prindle, 1 the placers occur in tributaries of moderate length, which flow in open valleys; some of the deposits are as much as 300 feet deep. The pay gravels, in part subangular, lie on the bed-rock and are from a few inches to 12 feet in thickness; these are covered by 10 to 60 feet of angular wash, evidently accumulated rapidly without opportunity for concentration, and above this rests a thick deposit of muck over which the sluggish N S FEATHER RIVER Upper L DREDGING GROUND s ^ gEZ ~^ r ^^ e tionl6 " w^s^ -About 4 Miles- Fio. 78. Diagrammatic section across Feather River below Oroville, California, a, Bed-rock; b, lone formation; c, tuffs of Oroville. streams pursue their way. The richest gravel worked in 1905, containing from $5 to $10 per cubic yard, occupies pay streaks on the bed-rock 150 to 200 feet wide, considerably less than the average width of the valley bottom. All the gravel on the bed-rock is, however, more or less auriferous. The gold is moderately coarse. Near the head of the stream deposition closely follows cutting and there the deeply buried, more or less permanently frozen pay streaks of the lower valleys merge into the deposits of the present stream activity. On a large scale similar conditions prevailed in Victoria, Australia. 2 Here there existed in Pliocene time an extensive river system with shallow, well washed, and locally extremely rich gravels which were formed during a prolonged time of nice balance between erosion and deposition. The region then 1 L. M. Prindle, The Fairbanks and Rampart quadrangles, Bull. 337, U. S. Geol. Survey, 1908. 2 W. Lindgren, Min. Mag., 2, 1905, p. 33. E^W. Lindgren, Eng. and Min. Jour., Feb. 16, 1905. i H. L. Wilkinson, Trans., Inst. Min. and Met., 1907, p. 9. Stanley Hunter, Mem. 7, Geol. Survey Victoria, 1909. 224 MINERAL DEPOSITS became depressed and covered by thick beds of sand and clay. Above this were poured out basalt flows, in places several hun- dred feet thick (Figs. 79 and 80). The broad valleys remain on the whole as before, but the present streams are weak and have little power of transportation and concentration. The dis- coveries of gold were made near the sources of the old rivers, Surface SCALE (= 100 200 300 FIG. 79. Diagram illustrating buried gravel channels (deep leads) of Victoria, Australia, and method of mining these deposits. where their gravels are near the surface; they were followed up- ward into the gullies of the slate hills, and downward below the level of the basalt flows. Such were the conditions, for instance, at Ballarat. South of Ballarat certain of the Pliocene stream gravels merge into coastal gravel plains, soon becoming marine in character. Such coastal gravel beds are opened in the Pitsfield FIG. 80. Longitudinal section of the Chiltern Valley and Rutherglen deep leads, Victoria, Australia, showing steeper grade of Tertiary river beds. mines, where the pay streaks of fine gold, resting on an almost level bed-rock, are worked beneath several hundred feet of sands and gravels. The Sierra Nevada of California, 1 on the other hand, offers an excellent instance of the result of elevation on gravel deposits. In the early Tertiary the surface of this range was comparatively gentle, and during long periods of rock decay and well-balanced 1 W. Lindgren, The Tertiary gravels of the Sierra Nevada, Prof. Paper 73, U. S. Geol. Survey, 1911. TRANSPORTATION AND CONCENTRATION 225 conditions gold from the quartz veins had become strongly concentrated on the bed-rock of the streams. The deeper gravels were then covered by a considerable thickness of more rapidly accumulated and poorer, but well-washed material, and this in turn by heavy masses of rhyolitic tuffs and andesite breccias so that the old channels were sealed in places by as much as 1,500 feet of superincumbent barren material. The range was elevated by mountain-building disturbances; new rivers were laid out and rapidly eroded canons to a depth of 2,000 or 3,000 feet. Even- tually the old gravels were exposed and now rest as more or less connected remnants on the summits of the ridges between the modern canons; the heavy gravel masses are worked by the hydraulic method, or the pay streak on the bed-rock is extracted by tunneling operations in the "drift mines" (Figs. 81, 82, 83). FIG. 81. Schematic representation of the four principal epochs of Tertiary gravels in the Sierra Nevada, a, Deep gravels (Eocene); 6, bench gravels (Miocene); c, rhyolitic tuffs and inter-rhyolitic channel; d, andesitic tuffs and intervolcanic channel. The gold from the destroyed portions of the old channels, together with more set free from the quartz veins during the erosion, accumulated in the modern canons. Along their slopes benches remain in places, indicating transient accumulations of gravel during the process of canon cutting. Somewhat similar conditions exist in some parts of Alaska. Near Nome on the ridges surrounding Anvil Creek are "high gravels" 600 to 700 feet above the present rivers. These gravels, some of which are rich, are the remnants of an old, now almost wholly eroded system of drainage. In the Klondike also high gravels occur a few hundred feet above the present creeks, the most conspicuous instance being the "White channel," described by McConnell 1 (Fig. 84). 1 R. G. McConnell, Klondike gold fields, Ann. Rept., Geol. Survey Canada, 14, 1901. R. G. McConnell, Report on gold values in the Klondike high-level gravels, Ann. Rept., Geol. Survey Canada, 1907. 226 MINERAL DEPOSITS Elevated beaches have been mined, for in- stance, at Nome, where there are two old beach lines 37 and 70 feet above the present level of the ocean. In Santa Cruz County, California, a similar elevated beach was mined for some time. Gold-bearing beach sand occurs)all along the Pacific coast from San Diego to Alaska, and in many other parts of the world. Size and Mineral Association of Placer Gold. Gold occurs in placers in all sizes, from masses weighing 200 pounds to the most minute flakes. Large nuggets are recorded from California; still larger specimens, weigh- ing as much as 2,184 ounces, were obtained in Victoria, Australia. It is often stated that heavier masses occur in placers than in quartz veins. This is decidedly erroneous. A mass of native gold found in the Monumental mine of Sierra County, California, weighed 1,146 troy ounces, and a quartz vein at Hill End, New South Wales, yielded a specimen which contained about 3,000 ounces. Every one who has had much experience in gold mining has noted the occurrence of thick sheets and masses of gold in deposits of certain kinds for instance, in the pockety quartz veins of Alleghany, California. Almost all the so-called placer nuggets of unusual size have been obtained from superficial deposits at or just below the croppings of rich veins. This applies to the Ballarat nuggets, weigh- ing from 80 to 160 pounds, which occurred in small steep gulches underneath the basalt flows, but immediately below the extremely rich outcrops of the quartz veins. It also applies to the nuggets of Carson Hill, Cali- fornia, the Poseidon nugget of Victoria (found in 1906 and weighing 953 troy ounces), and other occurrences. Some very rich placer deposits for instance, those of the Klondike, Yukon Territory, and the Berry mines in TRANSPORTATION AND CONCENTRATION 227 Victoria, Australia contain no specially large pieces of gold. The heaviest nugget found in the Klondike is said to have weighed 85 ounces. The angularity of the gold is proportional to the distance traveled; the final product is usually a flat, rounded grain from a fraction up to 1 millimeter in diameter. Occasionally crys- tallized gold is found in placers, but this is unusual and indicates close proximity of the primary deposit. There is probably no authenticated case of crystallized gold occurring in the gravels of larger water courses where there has been long transportation, and this is assuredly a strong argument against the assumption that such crystals are formed by^second- ary processes in the gravels. Fragments of quartz often adhere to the gold or form part of jsSu^. 1 ^^___ Volcanic Capping /. "~~-~- . = r~jr~~233| .r.iu-': x sV:i 3000 . ] >MK%; 2 FIG. 83. Longitudinal section of "Blue gravel channel," at Breece & Wheeler mine, Forest Hill divide, Placer County, California. After R. E. Browne. the rounded nugget. While the quartz pebbles so abundantly found in gold-bearing gravels do not ordinarily contain visible gold, there are many instances of such occurrences for ex- ample, at Elk City and Idaho City, Idaho, and at Dutch Flat and Nevada City, California. Some placer gold, more fre- quently the scaly variety, is covered by a thin film of silica, manganese dioxide, or limonite, and does not amalgamate easily. The most abundant minerals associated with the gold in placers are magnetite and ilmenite ("black sand"), garnet, zircon ("white sand"); and monazite ("yellow sand"), as well as many others of the heavy minerals occurring in the rocks which contain the primary gold deposits. Cassiterite is common in placers, and some deep gold placers in Victoria contain enough to make it a 228 MINERAL DEPOSITS valuable by-product. Gray platinum and silvery foils of iridos- mine are present in small quantities in many California placers adjacent to areas of serpentine. None of the minerals mentioned are ordinarily derived from the gold-bearing veins, but from the surrounding rocks. Pyrite or marcasite may form in the gravels; sometimes this pyrite contains a little gold, but contamination of the assay sam- Muck Stream gravels Terrace gravels White gravels / e Yellow gravels f High level River gravels Klondike schist y White Channel gravels FIG. 84. Sections across Bonanza Valley, Yukon Territory, showing several types of gravel deposits. After R. G. McConnell, Geol. Survey Canada. pies by the placer gold itself is always a possibility. Again the pyrite may be clastic and derived from the surrounding rocks, for pyrite does not seem to oxidize readily in running water; or, as near Nevada City, in the Harmony channel, the gravel may contain undecomposed pyrite, rich in gold, and derived directly from the primary veins over which the water course flowed. Other occasional associates of gold, probably derived from its primary deposits, are silver in nuggets (Alaska), native bismuth TRANSPORTATION AND CONCENTRATION 229 (Queensland and Alaska), native amalgam, palladium-gold, native copper, and cinnabar. The presence of native lead has usually been explained by accidental admixture of hunter's shot, but J. Park 1 asserts that there is an instance of its undoubted presence in gravel, the lead containing a skeleton of native gold. Fineness and Relation to Vein Gold. The fineness of placer gold (or parts of gold per thousand) varies from about 500 up to 999. Silver is always alloyed with the gold, but other metals are rarely prominent; copper is occasionally present. While vein gold may have a fineness of 997 to 999, this is exceptional; far more commonly its fineness ranges from 500, which corre- sponds to electrum, to about 800 or 850. This is assumed to be determined on specimens of native gold, for it is obviously not fair to take the usually lower figure of the retorted bars, which become admixed with impurities during amalgamation. The placer gold in any district will usually be of higher grade than the vein gold, and its fineness increases with the distance transported and with the decreasing size of the grains. Thus, while in Cali- fornia the vein gold averages 850 fine, the transported placer gold in the Tertiary channels averages 930 to 950. It has been shown that this increase in fineness is due to the solution of the sil- ver in the alloy in the outer layer of the grains by the action of surface waters. McConnell has proved that in the nuggets from the Klondike the outside actually has a greater fineness than the inside. The loss of silver in the outer part was from 5 to 7 per cent. This interesting result well illustrates the relative insolubility of gold. 2 Gold in Relation to Bed -Rock. While the bulk of the gold usually rests on the bed-rock or within a foot or two of it, this is not an invariable rule. In some gravels the coarser gold is oc- casionally scattered through the lower 4 to 20 feet. But it is never, except in minute quantities, distributed equally through a 1 J. Park, Mining geology, London, 1907, p. 18. 2 W. B. Devereux, The occurrence of gold in the Potsdam formation, Trans., Am. Inst. Min. Eng., 1882, pp. 465-475. Ross E. Browne, Colorado placer gold, Eng. and Min. Jour., vol. 59, 1895, pp. 101-102. W. Lindgren, The gold belt of the Blue Mountains of Oregon, Twenty- second Ann. Rept., U. S. Geol. Survey, pt. 2, 1901, p. 637. See also Prof. Paper 73, U. S. Geol. Survey, 1911. R. G. McConnell, Report on gold values in the Klondike high-level gravels, Geol. Survey Canada, 1907, p. 979. 230 MINERAL DEPOSITS great thickness of gravels. An excellent instance is McConnell's section of the White channel deposit in the Klondike. (See Fig. 84.) The washed gravel is here 150 feet thick. The gold content of the gravel is as follows: 0- 6 feet above bed-rock, $4. 13 per cubic yard. 6-12 feet above bed-rock, $0. 18 per cubic yard. 12-18 feet above bed-rock, $0 . 047 per cubic yard. 18-24 feet above bed-rock, $0 . 04 per cubic yard. 24-30 feet above bed-rock, $0.034 per cubic yard. 30-36 feet_above bed-rock, $0.032 per cubic yard. 36-42 f eet'above bed-rock, $0 . 032 per cubic yard. 42-48 feet'above bed-rock, $0.045 per cubic yard. 48-54 feet above bed-rock, $0.025 per cubic yard. From 54 feet above bed-rock the quantity of gold contained per cubic yard gradually and steadily diminished to $0.006 at the top. There is only fine gold in the upper gravels. A local enrichment has taken place on false bed-rock, a clayey stratum above the real bed-rock; here the gold is much coarser than directly above or below, but finer than on the bed-rock. Oc- casionally rich gravel may be found a few feet above bed-rock while it is less rich immediately on it. Coarse and moderately coarse gold moves very slowly. Mc- Connell found, for instance, that the White channel, where inter- sected by gulches, has left almost the whole amount of its gold in these immediately below the place where the trenching has occurred. In some cases the horizontal movement scarcely equaled the vertical. Smooth, hard bed-rock is poorly adapted to retain the gold; when it is somewhat clayey and decomposed much better results are obtained. Schists and slates make good bed-rock when decomposed, especially when they strike parallel to the channel. Serpentine forms a smooth and unsatisfactory bed-rock. Gold works down into bed-rock in a most 'surprising way. In hard rock it settles into the most minute crevices. In soft rock it burrows to a depth of 1 to 5 feet, so that^it is always necessary to mine this amount of the bed-rock. In limestone, irregular solution cavities contain the detrital gold, and these sometimes descend to a depth of 50 feet or more. Compact clay is good bed-rock, also clayey sandstone and clayey volcanic tuffs, the occurrence of the latter being exemplified in the Oro- ville_dredging grounds, in California. TRANSPORTATION AND CONCENTRATION 231 In glacial till and moraines there has been little opportunity for concentration, and unless the primary vein deposits were unusually rich, these gravels are of little value; the gold contained in them may, of course, be concentrated by glacial streams work- ing over the morainal detritus. Grade of Auriferous Watercourses. -All kinds of grade occur in watercourses containing gold-bearing gravels. In steep creeks the grade may be many hundred feet per mile, but the placers in these are usually poor. California rivers, in the Sierra Nevada, have grades of 50 to 100 feet or more per mile. Many of these have been extremely rich where gravel bars have had an oppor- tunity to accumulate. The White channel, in the Klondike, has a grade of about 30 feet per mile. Many of the present Alaskan streams have a grade of 100 to 150 feet per mile. In the prin- cipal Tertiary channels of Victoria, Australia, low grades down to 20 feet prevailed. In depressed or elevated channels of past epochs, as in Cali- fornia, Victoria, and the Klondike, changes of original grade must be considered. This is best established in the California channels, which now have grades of 100 to 150 feet, whereas the original streams had much less, the increase being due to the westward tilting of the Sierra Nevada. The best results of gold concentration are probably obtained in rivers of moderate grades, perhaps 30 feet per mile, under nicely balanced conditions of cor- rasion and deposition. Whenever overloading and active de- position take place concentration of coarse gold ceases. On the other hand, where erosion is rapid conditions for rich placers are less favorable, unless, as in the present streams of the Sierra Nevada, the gold supply is unusually abundant. The Pay Streak or "Run of Gold." 1 Except in smaller creeks the distribution of the gold in a gravel bed is far from regular. There is usually gold on the bed-rock over the whole area of the stream bed, but the richer part makes a narrower streak which follows a devious course, distinctly affected by the character of the bed-rock, sometimes splitting and re-forming, following first one side, then crossing diagonally to the other side. It is not necessarily in the deepest depression or gutter. Fig. 85 shows this devious course of the pay streak in comparatively shal- low gravels at Maryboro, Victoria. It is clearly independent 1 J. B. Tyrrell, The law of the pay streak in placer deposits, Trans. Inst. of Min. and Met., London, May 16, 1912, Min. and Sci. Press, June 1, 1912 232 MINERAL DEPOSITS of the course of the present small stream. In broader gravel plains, of which the Homebush and Pitsfield Tertiary placers of Victoria are examples, the "run of gold" follows a distinct and \yell-defined course on an almost level country rock. All this shows clearly enough the impossibility of the view that the gold was first uniformly distributed through the gravels and then gradually settled to the bottom under the influence of gravity. n Goidbearing Gravels covered by 20-50 It. of Alluvium SCALE FIG. 85. Map showing position of pay streak in alluvial gravels of Maryboro, Victoria. After S. B. Hunter. These pay streaks assuredly indicate epochs of well-balanced and long-maintained conditions during which the gravels could accu- mulate to only moderate depths and were at all times water- soaked and in a condition of slow movement. With more abundant loading of detrital material the gold-transporting power of the stream diminishes at a very rapid rate. Solution and Precipitation of Gold. Many of the earlier ob- servers, such as Genth, Lieber, Selwyn, Laur, Egleston, C. New- TRANSPORTATION AND CONCENTRATION 233 bery, and Daintree, concluded from observations in various parts of the world that placer gold, and particularly the large nuggets, has been deposited by circulating solutions. At present the mechanical derivation of the gold seems established beyond all doubt, although under exceptional circumstances some solution and redeposition may have taken place. 1 Even now, however, some writers, like J. M. Maclaren, 2 are inclined to place much em- phasis on this secondary deposition. It is probable, nevertheless, that this process is absolutely insignificant from an economic point of view. Nuggets, when cut and polished, almost always show a granular structure perfectly in accordance with vein gold. Liversidge, in a long series of experiments, found only two speci- mens (both from New Guinea) which showed a concentric struc- ture indicative of concretionary deposition. Very rare instances are quoted of quartz pebbles with dendritic films of gold 3 or of nuggets with minute gold crystals on their surface. 4 The col- lection of J. Edman, of San Francisco, contained a small crystal of magnetite coated with a thin film of gold. This came from the Tertiary deposits at Providence Hill, Plumas County, Cali- fornia, and Mr. Edman stated that he had never seen similar occurrences in the modern gravels. It seems to be well estab- lished that pyrite reduced by organic material in the gravels may contain some gold and also that the metal is occasionally found at the roots of trees or in the grass roots. The gold crystallized in minute octahedrons in the clay of Kanowna, Western Australia, is, as Maitland 5 pointed out, im- mediately above or adjacent to the decomposed croppings of the veins and the occurrence can scarcely be called a placer. The gold which works down into the soft bed-rock of the placers is in all cases, where I have observed it, of clearly detrital origin. It is stated that the ashes of trees in the gold-bearing region of the Guianas contain an appreciable quantity of gold. Origi- nally asserted by Lungwitz, this has been denied by Dubois and Kollbeck and then reasserted by Harrison on the basis of 1 Regarding the older literature, see the text-books of Stelzner, Bergeat, and Beck. In more detail, see Liversidge, Jour., Roy. Soc. N. S. W., vol. 27, 1893, p. 343; vol. 31, 1897, p. 79; vol. 40, 1906, p. 161. 2 J. M. Maclaren, Gold, London, 1908, pp. 80-86. 3 R. G. McConnell, Ann. Rept., Geol. Survey Canada, vol. 14, 1901, p. 64-B. 4 Gordon, Trans., Am. Inst. Min. Eng., vol. 25, 1895, p. 294. 5 J. M. Maclaren, op. cit., p. 83. 234 MINERAL DEPOSITS careful investigations. 1 His statement must be accepted, although it certainly taxes the imagination to believe that gold- bearing solutions can exist in a soil together with organic matter. From widely separate parts of the world gold has been reported in the ash of coal, but in this case it may be detrital and con- tained in admixed sand and clay. Gold is easily brought into the colloid state and as such it may be transported in solutions of colloid silica. It is very readily precipitated by electrolytes but this mode of solution may account for some cases of secondary gold in placers. According to the most recent investigations 2 gold is soluble in superficial waters only when free chlorine becomes liberated by the inter- action of sulphuric acid, sodium chloride, and manganese dioxide, a combination that must sometimes occur in ore deposits subject to oxidation; in the presence of oxidizing pyrite some gold may therefore be taken into solution, as chloride but it would probably not remain long before encountering reducing substances. While gold is slightly soluble in sodium carbonate, sodium sulphide, and other similar compounds, these would not ordinarily be encoun- tered in the waters of the zone of oxidation. Relation to Primary Deposits. That placer gold is directly derived by mechanical processes from vein deposits or analogous occurrences is absolutely certain, and examples of convincing character are present everywhere. This does not imply that the primary deposit can be worked at a profit. In most cases the gold is traceable up to the deposit. On this principle the pocket hunter proceeds, panning the detritus and working up hill until the source of the scattered gold has been found. The area in which the detritus occurs has the shape of a triangle, the apex of which is the pocket. It is a common experience that rivers or creeks crossing a vein or a mineral belt are enriched immediately below it, the coarseness of the gold increasing upstream to the place where the outcrops are crossed. As examples may serve the great accumulations of placer gold in the Neocene gravels of Eldorado County, California, where the Mother Lode crosses them, and the rich channels in upper Nevada County, just below the belt of 1 J. B. Harrison, Geology of the gold fields of British Guiana, London, 1908, p. 209. 2 W. H. Emmons, Enrichment of ore deposits, Bull. 625, U. S. Geol. Survey, 1917, p. 305. TRANSPORTATION AND CONCENTRATION 235 quartz veins at Washington and Graniteville. There are fine examples in Victoria, where the gravels are rich only where they cross or follow systems of veins or "reef lines." The White channel of the Forest Hill divide, California, follows a belt of quartz stringers in clay slate. The Idaho Basin 1 presents an excellent instance of large gravel bodies the gold content of which is traceable up to certain auriferous vein systems. Economic Notes. The world's annual production of placer gold is about $70,000,000. To this the Alaska and Yukon dis- tricts contribute $20,000,000, California $9,000,000, Victoria $2,000,000, and Siberia $18,000,000. While placers are found in almost all gold- and silver-producing regions, Brazil, the Ural Mountains, Siberia, California, Alaska, and Victoria have had by far the greatest total production. Gold-bearing gravel is often measured by the ton, but more commonly by the cubic yard. Still another measure is by sur- face area, sometimes by the square foot, in Australia commonly by the square fathom; this is especially applicable to deep mining when only the richest bottom layer is mined; at least 2 feet of gravel and 1 foot of soft bed-rock are extracted, making one square fathom equivalent to a minimum of 4 cubic yards. . In river bars gravels are worked by wing dams and pits kept dry by simple pumping devices. On a large scale they may be ground-sluiced or washed by the hydraulic method, with the aid of elevators when the natural fall is insufficient. The elevated gravels of earlier periods are worked in California by tunnels and drifting operations on the bed-rock. The mini- mum cost of working under the most favorable conditions is 50 cents per cubic yard, but is commonly $1 to $2 per cubic yard; most of the gravels actually worked contain at least $1.50, and often much more. The whole gravel body may be washed by the hydraulic method, when the expense may be reduced to 2 to 5 cents per cubic yard; of course the cost of preliminary work like ditches, etc., is often great. Some creek gravels in the Seward Peninsula contain from $2 to $6 per cubic yard; the width of the deposit may be about 50 feet, the depth 3 to 6 feet. The depressed gravels of earlier periods are worked by drifting from shafts, as in Victoria, where, however, the preliminary pumping, to permit access, is an extremely heavy expense, often 1 W. Lindgren, The mining districts of Idaho Basin and the Boise Ridge, Eighteenth Ann. Rept., U. S. Geol. Survey, pt. 3, 1898, pp, 617-744. 236 MINERAL DEPOSITS indeed prohibitory. Some of these Australian channels have been extremely rich, the workable portions ranging from $2 to $15 per cubic yard. Some of the channels are in places several hundred feet in width. Of late, gravels have been extensively worked in California, Alaska, the Klondike, and elsewhere by the dredging process. In California, where this method has reached its highest develop- ment, $7,769,000 was obtained from 58 dredges in 1916 from the flood-plains of the rivers at the foot of the Sierra Nevada, and the cost has been reduced from about 10 cents to 3 or 4 cents per cubic yard handled. In Alaska the cost is of course much higher and gravels containing less than 50 cents per cubic yard are rarely worked. The dredge will probably prove to be the most efficient placer-mining machine of the future, replacing the hydraulic method, which offers difficulties in the disposition of the tailings. In 1917 Alaska yielded $2,500,000 from dredging. Certain gravels in the dry regions of Arizona and northern Mexico are treated by pneumatic concentration in so-called dry washers, but the output of these placers is insignificant. 1 Yields of placer deposits are often calculated in dollars per lineal foot of channel. Good channels for drifting may produce from $70 to $500 per foot. The richest drift mine worked was probably "Madame Berry" in Victoria, with average width of 450 feet, yielding $1,293 per foot along channel. The two claims below this produced, respectively, $843 and $443 per foot, the last- named channel being mined 1,000 feet wide. The White Channel in the Klondike gave $380 per foot; the Red Point channel in Placer County, California, $72, the width being 120 feet; the American Hill hydraulic mine, Nevada County, 1,000 feet wide, $414; the Nome creeks, Alaska, 50 feet wide, about $100. By drifting operations alone, only a part of the gold will be extracted, say one-fifth to one-half, dependent upon the thickness of overlying gravels. THE GOLD-BEARING CONGLOMERATES OF SOUTH AFRICA Of the extensive literature the following principal papers are quoted : G. F. Becker, The Witwatersrand Banket, etc., Eighteenth Ann. Rept., U. S. Geol. Survey, pt. 5. 1896, 1 T. A. Rickard, Trans., Am. Inst. Min. Eng., vol. 28, 1899, p. 480. F. J. H. Merrill, Min. and Sci. Press, July 13 : 1912. TRANSPORTATION AND CONCENTRATION 237 G. A. F. Molengraaff, Die Reihenfolge der geol. Form, in Siid Afrika, Neues Jahrbuch, 1900, B. 1, pp. 113-119. F. H. Hatch and G. S. Corstorphine, Petrography of the Witwatersrand conglomerates, etc., Proc., Geol. Soc. S. A., vol. 7, pt. 3, 1904, pp. 140-145. F. H. Hatch and G. S. Corstorphine, The geology of South Africa, London. 1905. J. W. Gregory, The origin of the gold in the Rand Banket, Trans., Inst. Min. and Met., London, vol. 17, October, 1907. Also Econ. Geol., vol. 4, 1909, pp. 118-129. Discussion in Econ. Geol, vol. 4, 1909, by G. F. Becker and G. A. Denny R. B. Young, The Rand Banket, London, 1917, pp. 125. See also description by R. Beck, Erzlagerstatten, 2, 1909, pp. 183-200. F. H. Hatch, The conglomerates of the Witwatersrand hi types of ore deposits, San Francisco, 1911, pp. 202-219. C. B. Horwood, The Rand Banket, M in. and Sci. Press, Oct. to Dec., 1913. E. T. Mellor, The upper Witwatersrand System; the East Rand, Trans., Geol. Soc. S. A., vol. 18, 1915, pp. 11-71. E. T. Mellor, The conglomerates of the Witwatersrand with discussion, Trans., Inst. Min. and Met., London, vol. 25, 1916, pp. 226-348. Hugh F. Marriott, Mining on the Rand, Trans. Inst. Min. and Met., London, 1918, Min. and Sd. Press, July 20, 1918. The development of the gold-bearing conglomerates of the Witwatersrand district, in the Transvaal, is one of the most wonderful chapters in the history of mining. From an in- conspicuous beginning in 1887, the production of these unique deposits has steadily increased. In 1917 the ore production amounted to about 28,000,000 tons, with a yield of $180,000,000. The total production to the end of 1917 exceeds $2,500,000,000, which is more than the total gold production of California, Colo- rado and Alaska. The average content of the ore has decreased, probably mostly on account of reduction in mining and metal- lurgical costs, from $12 to $6 or $7 per ton, and, according to Hatch, it is probable that in the future ore of $5 grade will be utilized. The increase in production continued to 1916 but it is probable that the flood tide of output has been reached. A depth of over 5,000 feet has now been reached, and, 1 owing to a favorable geothermic gradient (p. 84), it will be possible to go considerably deeper. The ore is reduced by a combination of amalgamation and the cyanide process; stamp mills and tube mills are the grinding machinery most commonly employed. South Africa is in the main a plateau of thick sedimentary beds which are poor in fossils and in part of sub-aerial origin. 1 In 1918 the Village Deep had attained a vertical depth of 5,350 feet and a depth along the dip of 9,800 feet. 238 MINERAL DEPOSITS The oldest rocks known are the Swaziland crystalline schists , and the granites intruded in them. On their eroded surface rest the upper and lower Witwatersrand sys- tem of slates, quartzites, and con- glomerates, aggregating 19,000 feet in thickness, and on top of these in turn a thick series of volcanic flows, called the Ventersdorp system (Fig. 86). H The age of the Witwatersrand sys- &3 tern is not definitely known; it is tcj probably Cambrian or pre-Cambrian. fei Next higher in the succession of rocks j is the Potchefstrom system, includ- ^ ing the Black Reef (oldest) , Dolomite, TO and Pretoria series. This again is cov- j J ered by the Devonian Waterberg sys- ^oo tern (Table Mountain sandstone of I the Cape) and the most recent J2 "f Karroo system, which is coal-bearing j|'f and considered to be of Permo-Car- | boniferous age Each system is sep- j.g arated by an unconformity from the g f next. The Witwatersrand system is folded in a syncline extending about J 120 miles east to west and 45 miles f north to south. At Johannesburg, at the north side of the syncline, the "> dip is to the south, steep near the 2 surface, but flattening in depth to about 30. Faulting is common and there are a number of intrusive diabase dikes, thought to belong to the overlying Ventersdorp volcanic system. Auriferous conglomerates occur at .several horizons in the Witwaters- rand system and also in the Black Reef series. The productive beds % TRANSPORTATION AND CONCENTRATION 239 are, however, in the upper part of the Witwatersrand, including a thickness of about 7,000 feet of quartzites and conglomerates, among which the following are distinguished, beginning from the top: Kimberley group, Bird Reef group, Livingstone Reef group and Main Reef group. The first two are each about 500 feet thick but the conglomerates contained are of low grade, rarely exceeding $3 per ton in gold. The Main Reef group, about 90 feet thick, includes several conglomerate beds more or less persistent. FIG. 87. Gold-bearing conglomerate, Johannesburg, South Africa. Peb- bles of quartz, crushed in places. Cement of sericite, quartz, and a little chlorite. Black areas are concretions of pyrite, replacing groundmass and quartz. B, prisms of chloritoid. Drawn by J. D. MacKenzie. The usual subdivision of the Main Reef group includes from top to bottom: South Reef (3 feet). Bastard Reef (scattered pebbles) and quartzite (20 to 40 feet). Main Reef Leader (2 feet). Quartzite (2 to 20 feet). Main Reef (4 feet). Of these the Main Reef Leader is the most productive; the pebbles in the conglomerate are small, averaging 2 inches in diameter, and consist of well-rolled fragments of glassy quartz with fewer pebbles of more angular quartzite, banded chert, and slate. The pebbles lie in a matrix of sandy material, which has 240 MINERAL DEPOSITS become hardened by infiltration of silica. Pyrite occurs in abundance in the cement, averaging about 3 per cent, of the rock and being present both in crystalline form and as rounded replacements after quartz probably representing two generations, both subsequent to the sedimentation. Chloritoid, 1 sericite, calcite, and graphite are other authigenetic minerals. The gold is not contained in pebbles, but only in the cement, and forms minute angular crystalline aggregates, very seldom rounded particles. It is usually closely connected with pyrite, either enclosed by it or covering the surface of pyrite aggregates. As a rule it is not visible to the naked eye. FIG. 88. Section through Village Deep No. 3 shaft. After H. F. Marriott. In spite of a long-continued discussion there is no unanimity among geologists as to the genesis of these remarkable deposits. It is evidently necessary, for a satisfactory discussion of the question, to go beyond the limits of the Johannesburg occurrences and consider the geological relations of the Transvaal and South Africa as a whole. The first suggestion that the conglomerate may be simply an alluvial or littoral placer is refuted by the character of the gold and its close association with the pyrite. Detrital pyrite may, of 1 A prismatic colorless mineral usually described as chloritoid is common (Fig. 7), but its idejitification appears questionable. TRANSPORTATION AND CONCENTRATION 241 course, occur in gravels, but there should always be some mag- netite and ilmenite present. Their absence is a strong argument against the theory of direct placer deposition. It is clear that if this is a placer deposit there has been extensive recrystallization and some migration. Equally untenable is the hypothesis of F. W. Voit 1 that the gold has been brought to the surface by hot springs which discharged into the ocean. The advocates of the placer theory, among whom are G. F. Becker, J. W. Gregory, G. A. Denny, R. B. Young and E. T. Mellor, are compelled to admit a recrystallization of the gold and a transformation of magnetite and ilmenite into pyrite. Many geologists and engineers, impressed with the difficulties confronting the placer theory, hold that the deposits are epigen- etic that the gold and pyrite have been introduced by a post- sedimentary infiltration, perhaps after the intrusion of the diabase dikes. Such views are held by H, Louis, J. H. Ham- mond, R. Beck, F. H. Hatch, G. S. Corstorphine and R. B. Horwood. Small and irregular quartz veins which in some places contain a little gold and sulphides intersect the Witwatersrand series. Whether there is any enrichment along the few diabase dikes is a disputed question. No doubt these quartz veins are related to the dikes. E. T. Mellor has recently approached the subject from the wider geological viewpoint and his papers contain very strong arguments in favor of the original deposition of the gold in alluvial gravels. He considers the quartzite and conglomerate series as large delta deposit rather than shore gravels and shows the existence of many horizons of gold-bearing conglomerates. Against the infiltration theory stands a long array of strong arguments: 1. The absence of channels followed by the solutions; 2. the regular distribution of the gold in the conglomerate; often it is concentrated in its lower layers; 3. the practical confinement of the gold to the conglomerates, though the quartzites are equally permeable; 4. the conglomerates were deposited in an alluvial plain skirting the deeply eroded Swaziland schists with their lenticular gold quartz veins and would thus certainly contain some gold. The fine flake gold would be recrystallized and pressed between 1 F. W. Voit, Der Ursprung des Goldes in den Randconglomeraten, Mon- atsber. Deutsch. geol. Gesell., vol. 60, Nog. 5 and 7, 1908. 242 MINERAL DEPOSITS secondary growths of quartz grains. The original black sand would be recrystallized to pyrite by action of meteoric waters in sediments which contained sulphates and organic matter. The difficulty not yet fully explained lies in the abnormal rich- ness and extent of the conglomerates. It is pointed out, how- ever, that large areas of the conglomerate are practically barren. The rich beach sands of Nome, Alaska, have been cited as an analogous case but the analogy is by no means perfect. Very fine detrital gold would, of course, be expected in a delta deposit near the shore line. Similar conglomerates of considerable geological antiquity are found in West Africa at Tarkwa and Abosso 1 and these have been worked on a fairly large scale.- Instead of pyrite these con- tain magnetite, and ilmenite with chloritoid. PLATINUM PLACERS 2 It is known that platinum occurs as a primary constituent of peridotites, and specimens showing its intergrowth with olivine and chromite have been described. Almost the entire world's production is obtained from placers and 95 per cent, of it is extracted from the placers on the eastern slope of the Ural Mountains, where detrital platinum occurs in the gravels of the stream courses, which head in certain areas of peridotite and pyroxenite. It is associated with iridosmine, iridium, chromite, and often also with gold. The crude platinum forms small rounded grains, very rarely nuggets up to 20 pounds in weight, and its fineness (per thousand) ranges from 750 to 850, the remainder being iron, copper, and various metals of the platinum group, particularly iridium. Platinum-bearing gravels occur also in Colombia, South America, in river beds and Tertiary conglomerates, and the pro- duction from them is increasing. In the United States the metal occurs in small quantities together with gold in almost all the gold-bearing districts in northern and central California 1 Edward Halse, Trans., Fed. Inst. Min. Eng., vol. 2, 1891, p. 69. R, Beck, Erzlagerstatten, 2, 1909, p. 200. 2 J. F. Kemp, Geological relations and distribution of platinum and associated metals, Butt. 193, U. S. Geol. Survey, 1902. Louis Duparc, Le platine et les g'ites platiniferes de 1'Oural. Geneve, 1911. D. T. Day, W. Lindgren and J. M. Hill, in successive issues of Mineral Resources, U. S. Geol. Survey. TRANSPORTATION AND CONCENTRATION 243 and in southwestern Oregon, where serpentine or peridotite is found. From 400 to 700 ounces are annually recovered, chiefly from the black sands of the dredges. 1 Platinum also occurs in the beach sands of southern Oregon, together with more or less gold; a small quantity of this is usually also recovered. The Tulameen district, British Columbia, formerly yielded some production. The normal world's production of platinum before the war was about 300,000 troy ounces, but it is much less at the present time. /Regarding platinum and palladium in vein de- posits and in peridotite see p. 790, where the uses of the platinum metals are also described?) For a long time the price of platinum was less than that of gold; a gradual rise increased the value to $20 per ounce, and in 1911 it had reached $45. Increasing scarcity forced the price up to $105 in 1917. Crude platinum with 70 per cent, to 85 per cent. Pt is sold from $30 to $60 per ounce. CASSITERITE PLACERS 2 The original home of cassiterite (or oxide of tin) is either in the granites, in pegmatite dikes, or in quartz veins. From any of these sources it may be set free by weathering and disintegra- tion, and, on account of its high specific gravity, it easily becomes concentrated in gravel deposits of different types. Among the accompanying minerals tourmaline, topaz, and wolframite are the most common. Grains of metallic tin are reported to occur with cassiterite in Nigeria and Australia. Eluvial deposits imme- diately below the croppings are numerous and are worked on a large scale at Mount Bischoff, in Tasmania. A small deposit of this kind resting in a shallow gully immediately below a pegmatite dike was mined near Gaffney, South Carolina, in 1905. The earliest production of stream tin came from gravels below the tin- 1 J. M. Hill, Platinum, Mineral Resources, U. S. Geol. Survey, pt. 1, 1917. James W. Neil, Recovery of platinum in gold dredging, Min. and Sci. Press, Dec. 8, 1917. 2 Sydney Fawns, Tin deposits of the world, London, 1907. H. W. Kayser and R. Provis, The Mt. Bischoff tin mine, Proc., Inst. C. E. (London), vol. 123, 1896, pp. 377-387. O. H. Van der Wyck, The occurrence of tin ore in the islands of Banca and Billiton, Seventeenth Ann. Rept., U. S. Geol. Survey, 1896, pt. 3, pp. 227-242. Mineral Resources, U. S. Geol. Survey, 1895. L. C. Graton, Reconnaissance of some gold and tin deposits of the southern Appalachians, Bvll. 293, U. S. Geol. Survey, 1906. 244 MINERAL DEPOSITS bearing lodes of the Erzgebirge, in Saxony, and of Cornwall, both sources now practically exhausted. Considerably more than one-half of the world's production of about 123,000 short tons of tin is still obtained from placers, mainly in the Malay Peninsula and the islands of Banca and Billiton, near Sumatra. New South Wales and Victoria furnish minor amounts and in the latter state some cassiterite is saved in working Pliocene auri- ferous stream channels. In this case the tin ore appears to be sparsely disseminated in granite and is liberated after its disin- tegration. At the Briseis mine, in Tasmania, the deposit worked consists of 14 to 45 feet of river gravel, covered by 20 to 40 feet of decomposed basalt and containing from 2 to 4 pounds of cassi- terite per cubic yard. In the United States small amounts of stream tin are recov- ered in Alaska near the extreme western point of the American continent, in the Black Hills of South Dakota, and in North and South Carolina. As tin is worth from 25 to 90 cents per pound and the easily reduced cassiterite contains 78.6 per cent, of the metal, it is clear that a small quantity, say 2 pounds per cubic yard of gravel, might suffice for profitable working. Some gravels in Tasmania now worked average only 0.6 pound per cubic yard. MONAZITE PLACERS 1 Monazite, an anhydrous phosphate of cerium, lanthanum, and other cerium metals, usually contains also from 3 to 8 per cent, of thoria, making it valuable for the production of nitrate of thorium, which is utilized in the manufacture of incandescent gas mantles. The mineral has a specific gravity of 5.203, a resinous luster, and a yellow to brown color; when occurring in placers it is found together with gold, zircon, magnetite, ilmenite, garnet, etc., after concentration in sluices. From its associated minerals it is cleaned in electromagnetic separators, the final product being about 90 per cent. pure. The source of the monazite is in the granites, gneisses, and pegmatites, where it occurs as a primary mineral. As its value (changing with the percentage of thoria) 1 J. H. Pratt and D. B. Sterrett, Monazite and monazite mining in the Carolinas, Trans., Am. Inst. Min. Eng., 40, 1909, pp. 488-511. D. B. Sterrett, Mineral Resources, U. S. Geol. Survey, 1906, pp. 1195- 1209. W. T. Schaller, idem, pt. 2, 1916, pp. 223-237. TRANSPORTATION AND CONCENTRATION 245 is about 8 cents per pound, monazite gravels may in places form workable deposits, especially where, as often happens, gold is present. Monazite is now obtained from marine and fluviatile placers in Brazil and India, but it is also obtained from similar deposits in North and South Carolina and has lately been found in Idaho, where a large intrusive batholith of granite or quartz monzonite evidently carries the mineral sparsely distributed throughout. The principal occurrence in Idaho is at the old placer district of the Idaho Basin. In 1910 about 100,000 pounds of monazite was mined in the United States, chiefly from placer deposits in the Carolinas. The total value is stated as $12,000. Since 1910 there has been no production in the United States, the supply being obtained from the extensive deposits in Brazil. OTHER PLACERS Magnetite, or "black sand," has been frequently mentioned above as a product of concentration in gravels and sands and is usually derived from the disintegration of igneous rocks. Along the beaches and the bars of some rivers it may accumulate in considerable masses for instance, on the lower St. Lawrence BIG MOUNTAIN . i!nn -S 1187 1200 ^S^SSS*! 1130 1001 3?*>>viiaL-^. .Shaft No. 1 Clay / Sandy Limestone 1200 -1100 1000 - 900 800 1 ' /vv ^/'C*Jv*' 700 Conglomerate Calcareous and 600 Ore Gritty Clay 500 400 500 -100 300 200 100 Feet FIG. 89. Section of Iron Mountain, Missouri, showing mining of detrital ore underneath limestone and sandstone, and of hematite ore in the por- phyry. After G. W. Crane. River, Canada, and along the Columbia River, Oregon but it is exceptional that such deposits have been utilized. 1 More or less ilmenite is usually mixed with the magnetite. There are several examples of eluvial deposits of iron ore (magnetite, hematite, or limonite), formed below croppings of iron deposits, and also of such detrital masses in the debris slopes 1 The magnetite sands of Japan appear to have been rather extensively utilized; also those occurring along the coast of New Zealand. 246 MINERAL DEPOSITS of older formations. At Iron Mountain, Missouri, 1 Paleozoic rocks rest upon a deposit of boulders of iron ore and porphyry, which in turn lie upon pre-Cambrian porphyry. The porphyry itself also contains deposits of hematite (Fig. 89). Similar eluvial masses of copper and lead ores are found in places. We may recall the great debris mass of chalcocite below the croppings of the Bonanza mine 2 in the Copper River region, Alaska, and galena beds on the slopes below the Elkhorn mine, Wood River, Idaho. 3 Some placers yield precious stones. Diamonds are believed to occur as a primary mineral of some peridotites, possibly also in other rocks. Diamond placers have been worked in Brazil, India, and South Africa. In the last-named region fine stones are found in the gravel of the Vaal River, and small diamonds have lately been washed from the beach sands of Liideritz Bay, German West Africa. In a few places sapphires, more rarely rubies (both aluminum oxide), are recovered from gravels. Along the Missouri River near Helena, Montana, 4 a bar was worked for several years for the sapphires it contained. They were plentiful, but a large proportion were of yellowish or pale blue color. 1 F. L. Nason, Report on the Iron Mt. sheet, Geol. Surv. Mo., vol. 9, 1912. 2 F. H. Moffit and S. R. Capps, Bull. 448, U. S. Geol. Survey, 1911, p. 89. 8 W. Liudgren, Twentieth Ann. Rept., U. S. Geol. Survey, pt. 3, 1900, p. 210. 4 D. B. Sterrett, Mineral Resources, U. S. Geol. Survey, pt. 2, 1910, p. 877. CHAPTER XVI DEPOSITS PRODUCED BY CHEMICAL PROCESSES OF CONCENTRATION IN BODIES OF SURFACE WATER BY REACTIONS BETWEEN SOLUTIONS LIMESTONE Definition and Origin. The limestones are sedimentary rocks, composed of carbonate of calcium, usually calcite, but in recent deposits also aragonite; 1 they contain minor amounts of magnesium and iron, also varying amounts of alumina and silica, and by the increase of these constituents transitions to shale or sandstone result. Phosphate of calcium and organic matter also enter into the composition of most limestones. The rocks are always crystalline, for there is no such mineral in nature as amorphous carbonate of calcium, but the grain varies between the widest limits. When water containing bicarbonate of calcium is discharged into the ocean or bodies of fresh water the calcium carbonate is often precipitated because of changing equilibrium in the solu- tions. 2 This is exemplified by deposits along the shore of the Great Salt Lake in Utah. Such limestones often form "oolitic" beds of small, rounded concretions. Generally, organic life plays a most important part in the de- position of calcite either indirectly by precipitation by ammonium carbonate generated by decaying organisms, or directly by life processes. Bacteria 3 may be the agent, very often also algae, the latter both in the sea and in fresh water deposits. Mollusks, corals, Crustacea and echinoderms segregate calcite and aragonite 4 in their shells which accumulate on the bottom at moderate depths. 1 J. Johnston, E. M. Merwin and E. D. Williamson, The several forms of calcium carbonate, Am. Jour. Sci., 4th ser., vol. 41, 1916, pp. 473-512. 2 J. Johnston and E. D. Williamson, The role of inorganic agencies in the deposition of calcium carbonate, Jour. Geol, vol. 24, 1916, pp. 729-750. 8 T. W. Vaughn, Chemical and organic deposits of the sea, Bull. Geol. Soc. Am., vol. 28, 1917, pp. 933-944. G. H. Drew, Publ. 182, Carnegie Inst. Washington, 1914, pp. 1-78. 4 Aragonite is the unstable form of calcium carbonate and always tends to change to calcite. It is most common in recent or Tertiary deposits. 247 248 MINERAL DEPOSITS Many organisms, such as sponges, secrete silica from the sea water and thus cherty deposits may be admixed with the limestones. Many limestones are almost wholly made up of shell remains but in others no trace of organic structure may be visible. Metamorphism tends to increase the grain and destroy the fossils. Evaporation of ordinary surface waters in dry climates may produce thick beds of porous limestone. This is known in Mexico as "caliche." Calcite and aragonite are often deposited in large masses by hot springs containing bicarbonate of calcium, and such deposits may closely simulate limestones. Certain beautiful banded and translucent spring deposits are called onyx and are used for ornamental stones. Among the many varieties of limestone the following may be mentioned : Chalk. This is a white, fine-grained, loosely coherent lime- stone of comminuted shells of mollusks and also of foraminifera. Its occurrence in the Cretaceous along the English coast is well known. Extensive beds are reported from Texas, New Mexico, Arkansas, and Kansas. Chalk is used as fertilizer, for whiting, for marking, for polishing powder, and for many other purposes. "Paris white" is a pigment made by grinding "cliffstone," a hard variety of chalk. Much of this is imported. Lithographic Stone. The variety of limestone used for engrav- ing and the reproduction of colored plates is a fine-grained rock with imperfect conchoidal fracture, gray or yellowish in color, and uniform in texture. It must be porous, to absorb the grease in the printer's ink, and soft enough to work readily under the engraver's tool. Lithographic stone of good quality is difficult to find. The product from the Solenhofen quarries in Bavaria is a Jurassic limestone of unusual excellence. The material is variable in composition and its value is ascertained only by trial. 1 Lithographic stone is reported to occur in several States of the Union, but none of it appears to be as good as Solenhofen rock. The plates used are 22 or 28 by 40 inches, and 3 inches thick. The better grades are expensive, selling at about 22 cents per pound. The best grade quarried in the United States is said to come from Brandenburg, Kentucky. 2 The demand is limited. 1 S. J. Ktibel, Mineral Resources, U. S. Geol. Survey, 1900, pp. 869-873. 2 E. O. Ulrich, Eng. and Min. Jour., vol. 73, 1902, p. 895. CHEMICAL PROCESSES IN SURFACE WATERS 249 Hydraulic Limestone. Certain argillaceous limestones or dolomitic limestones are used for the manufacture of natural cement. Such rock, crushed and burned, hardens or "sets" when mixed with water, owing to hydration and crystallization of the silicate and aluminate of calcium formed during the burning. " Portland 'C cements are mixtures of limestone and argillaceous rocks, subjected to a similar process of grinding and burning. 1 Lime. Pure limestones are changed by burning and conse- quent decarbonation to quicklime which is usually shipped as "lumplime;" this "slakes" to a calcium hydroxide when mixed with water and is used, with the addition of sand, as mortar in brick constructions. The "slaking" is retarded by the presence of magnesia and argillaceous impurities. 2 Slaked lime, finely powdered is a recent product used extensively for water proofing concrete. Uses. While limestone is mainly a structural material it is also used as a flux in smelting operations. Only the pure varieties are acceptable, though magnesian limestones are also used in iron furnaces. Burnt lime is probably employed for more purposes than any other natural product. It is used for the manufacture of bleach- ing powder, ammonia, calcium carbide, fertilizers, wood alcohol, soap, glycerine, glue, glass, pottery, paints, paper and sugar. Also for tanneries and as insecticide and fungicide. DOLOMITE 3 Pure dolomite contains 54.35 per cent. CaC0 3 and 45.65 per cent. MgCOs. Beds of dolomite and dolomitic limestone are common in sedimentary deposits. They may often be distin- guished by a fine-grained sugary texture, due to a development 1 E. F. Burchard, Mineral Resources, U. S. Geol. Survey, 1916, with list of literature. 2 E. F. Burchard and W. E. Emley, The source, manufacture and use of lime, Mineral Resources, U. S. Geol. Survey, pt. 2, 1913, pp. 1509-1593. 3 F. M. Van Tuyl, New points on the origin of dolomite, Am. Jour. Sci., 4th ser., vol. 42, 1916, pp. 249-260. F. M. Van Tuyl, The origin of dolomite, Ann. Rept., Iowa Geol. Survey, vol. 25, 1914, pp. 251-422. For an extended discussion of dolomite see F. W. Clarke, Geochemistry, Butt. 616, U. S. Geol. Survey, 1916, pp. 559-571. E. Steidtmann, Origin of dolomite, etc., Butt. Geol. Soc. Am., vol.[28, 1917, pp. 431-450. 250 MINERAL DEPOSITS of uniform rhombohedral crystals. Dolomite is somewhat harder than limestone and is insoluble in dilute hydrochloric acid. Magnesium carbonate is much less soluble in water than calcium carbonate, as shown by the fact that stalactites in magnesian limestone caves are almost wholly CaCO 3 . Some travertines from mineral springs are rich in MgCO 3 and may contain up to 29 per cent, of this compound. Dolomite is doubtless deposited by direct precipitation in sea water, but much of the dolomite has been formed by alteration of the limestone by sea water, or by subsequent dolomitization by surface waters. Deep borings in coral reefs have shown that the limestone, somewhat magnesian at the surface, passes into dolomite in depth. Certain algae deposit much MgC0 3 with CaC0 3 ; some shells also contain magnesium carbonate but seldom more than 7 per cent. In warm waters the percentage of MgCO 3 in shells tends to increase. Instances are known of the deposition of thin beds of pure magnesite in bodies of water. IMPORTANCE OF CARBONATE ROCKS AS RELATED TO ORE DEPOSITS Within the zone of oxidation the carbonate rocks are often dissolved, residual clays being then developed. Accessory con- stituents such as barium (and strontium), probably present in most limestones but rarely determined, or zinc and lead in the form of sulphides, or admixed phosphates may then become concentrated and acquire economic importance. Limestone is easily silicified by waters containing silica; the silica usually appears as irregular masses of fine-grained quartz or chert. It is quite as easily dolomitized by dilute waters con- taining some magnesia, and this is often observed near ore deposits formed at slight or moderate depth. Limestone and dolomite, under the influence of heated waters, are subject to replacement by quartz, dolomite, barite, and fluorite or by metallic ores such as pyrite, blende, and galena. At high tem- perature and pressure pure limestones recrystallize to marble. Silicates, such as garnet, diopside, or wollastonite, form in argillaceous or siliceous limestone from the impurities contained or from the introduction of solutions rich in silica and iron. Lastly, the limestones are easily soluble and caves develop along fractures, forming receptacles for the deposition of ores. CHEMICAL PROCESSES IN S URFA CE WA TERS 25 1 CHERTS AND DIATOMACEOUS EARTH The silica accumulated by detrital processes as sandstone and quartzite has already been mentioned. Silica may, however, also be extracted from water and deposited as a sediment by means of organisms, such as radiolarians, diatoms, and sponges. In part this silica forms cherty masses included in limestone; part is deposited as distinct beds. Diatomaceous earth 1 is a deposit formed in lakes and swamps, as well as in the sea, and, when pure, consists of the delicate tests of diatoms, a class of algae (see Arnold and Anderson, Bull 315, U. S. Geol. Survey, p. 438). Such deposits accumulate abundantly where siliceous volcanic tuffs were deposited in lakes, as occurred at many places in the Cordilleran region during the Tertiary period. Thick beds are found in the Miocene of Santa Barbara County, California. The diatomaceous earth is frequently more or less admixed with rhyolitic glass and other detritus; the tests consist of hydrated opaline silica. The earth forms light-colored beds of extremely fine texture and it finds extensive use as a polishing powder, a steam-pipe packing, and an absorbent for various liquids. It contains up to 87 per cent. Si0 2 and 5 to 9 per cent. H 2 O. SEDIMENTARY SULPHIDE DEPOSITS As the sedimentary rocks largely consist of the detritus of the continents, it is self-evident that they may contain the metals of the rocks and ore deposits of the land areas. Iron is, of course, abundant, also in a lesser degree manganese; concretions of hydrous oxides of manganese are found in the deep sea deposits and analysis shows that these contain notable amounts of nickel, cobalt, copper, zinc, lead, and molybdenum. Many limestones have been shown to contain minute amounts of zinc, lead, and copper. River sands and gravels and even littoral ocean sands may, locally, contain some detrital pyrite, but it is extremely unlikely that a sufficient quantity of these sulphides would escape oxidation to form important deposits. Two very important analyses of composite samples have been published recently by the U. S. Geological Survey and are given 1 W. C. Phalen, and F. J. Katz, Mineral Resources, U. S. Geol. Survey, 1908-1916, under heading "Abrasives." For the preparation and uses, see Percy A. Boeck, Met. and Chem. Eng., vol 12, 1914, pp. 109-113. 252 MINERAL DEPOSITS below. 1 The rarer elements determined are of special interest. There can be no doubt now that copper, lead and zinc as well as nickel and cobalt are contained in marine and fluviatile deposits but the quantities seem to average smaller than these determined in paleozoic carbonate rocks, and considerably smaller than those obtained from pre-Cambrian igneous rocks (p. 10). ANALYSES OF SILT AND DELTA DEPOSITS 1 2 SiOj 46.64 69.96 AliOs 14.08 10.52 Fe 2 O s 4.14 3.47 FeO 1.88 MgO 1.95 1.41 CaO 7.20 2.17 NasO 2.98 1.51 KO , 1.84 2.30 H 2 O- 4.73 3.78 H 2 O+ 5.86 1.96 Ti0 2 1.84 0.59 ZrOz None 0.05 COi 4.05 1.40 POi 0.17 0.18 SO. 0.32 0.03 Cl 2.25 0.30 F ' 0.07 S 0.11 0.07 BaO 0.05 0.08 SrO 0.025 Trace MnO 0.10 0.06 V,Oi 0.028 0.02 CnOt 0.044 0.01 MoOi None AsiO . Trace 0.0004 (Ni, Co)0 0.080 0.017 CuO 0.009 0.0043 PbO 0.0004 0.0002 ZnO 0.0087 0.001 C 1.38 0.66 101.775 100.6229 LessO. 0.56 0.12 101.215 100.5029 George Steiger, 1. Terrigenous blue mud; composite of 52 samples analyst. 2. Mississippi delta mud; composite of 235 samples. George Jsteiger, analyst. Pyrite, and infrequently other sulphides, may be precipitated by chemical reactions in sediments. Beds of oolitic pyrite are known (p. 269); iron disulphide is formed in places in bogs 1 C. E. Siebenthal, Zinc and lead deposits of the Joplin region, Bull. 606, U. S. Geol. Survey, 1915, p. 72. With extensive discussion of results. CHEMICAL PROCESSES IN SURFACE WATERS 253 and streams or in oceanic sediments where hydrogen sulphide developed by decaying organic matter reacts on the sulphates of iron. If these sediments are -brought to the surface by oro- genic movements and slightly metamorphosed, the sulphide, originally in fine dissemination, may recrystallize in more prominent form. As a matter of fact, the deep sea muds thus far analyzed contain little or no pyrite. In the special case of the Black Sea, often quoted of late from N. And- roussof s description, 1 microorganisms assist in liberating hydro- gen sulphide, part of which, by reaction with iron from the sediments, develops pyrite. P Fia. 90. Vertical section through the pyritic deposit at Meggen, Ger- many. P, Pyrite; B, barite; k, limestone; bs,cs,ls, Devonian slates. After Strauss. Although pyritic clays are abundant in the unmetamorphosed sedimentary formations, there is little evidence of extensive sedimentary beds of pyrite. The deposit at Meggen, in Westphalia, is often referred to as tending to prove the existence of sedimentary beds of pyrite. Oolitic pyrite with barite and zinc blende occurs here in De- vonian beds and is worked on a fairly extensive scale the bed being from 12 to 20 feet thick (Fig. 90). Alfred Bergeat 2 ap- pears to have demonstrated the sedimentary origin of the oolitic pyrite. The barite and the zinc blende are somewhat later than the pyrite and their origin is still doubtful. Bergeat holds that all these minerals form a local marine deposit. Barite when appearing in sandstone is probably deposited 1 N. Androussof, Guide des excurs. du VII Cong. ge~ol. int., 29, 1907, p. 6. 2 A. Bergeat, Zeitschr. prakt. Geol, vol. 22, 1914, pp. 237-249. 254 MINERAL DEPOSITS by hot springs; it is difficult to conceive its origin from sea water. For barite and manganese carbonates as marine shore deposits see p. 269. Evidence is scant as to the sedimentary deposition on a large scale of sulphides other than pyrite or marcasite. "The Kup- ferschiefer" of Mansfeld, in which the sulphides may be of syngenetic origin, will be described elsewhere (p. 413). SEDIMENTARY IRON ORES It is conceded that iron ores, such as magnetite, can be de- posited by mechanical concentration as placers along rivers or the seashore (p. 245), or again we may easily conceive hematite or limonite derived from deep decay of rocks along the littoral, or from the. oxidation of pyrite deposits, as at Rio Tinto, Spain, swept out into the sea and deposited close to the shore. Iron ores are also formed by chemical reactions in bodies of water, and these yield a notable proportion of the iron production of the world. In the latter cases the iron has been supplied from the land areas in form of solutions. In many instances both dis- solved iron salts and detrital minerals of iron contribute to the genesis of the deposits. The surface waters extract iron from ferromagnesian silicates as well as from oxides or other minerals; this extraction proceeds most energetically in regions covered by a deep mantle of decayed rock. ! Both iron and manganese are contained in springs and streams. An example of such spring water, rising underneath a deposit of bog iron ore in Holland, is quoted by Clarke: ANALYSIS OF SPRING WATER AT EDERVEEN, NETHERLANDS 1 (Analyst, G. Moll van Charante) (Parts per million) Ca 107.6 A1 2 O 3 3.3 Mr 5 6 Cl 15 2 Fe . 19.6 H,PO, 10 9 Mn K Na 11.4 0.9 10 SO 4 CO 3 SiO 0.9 207.6 18 Organic 56.0 467.0 1 Cited by F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 530. CHEMICAL PROCESSES IN SURFACE WATERS 255 A larger part of the iron dissolved by the surface water is precipitated after a short journey, 1 but some of it is carried down by the streams to lakes and seas, in which it then may be deposited on an extensive scale. The sea water contains about 0.6 milligram per liter of iron and probably more at some places near the shore. LIMONITES IN SWAMPS AND LAKES (BOG IRON ORES) Occurrence. The bog iron ores are found in swamps, lakes, or even in sluggish water courses, and they are especially abun- dant in the recently glaciated regions of northern Europe, Asia, and North America. They consist of dark-brown, rough and cellular masses or loose particles, sometimes oolitic in structure and then designated as "shot ore," and form a layer of varying thickness at the bottom of the swamp or lake. Plants and roots may be replaced by limonite. Such ores are usually mined by means of primitive dredges or scoops. The ore occurs mainly in shallow waters along the shore, to a depth of about 12 feet. After removal a new layer is formed within a few years; according to A. Geikie 2 several inches of limonite accumulated in 26 years in a Swedish deposit. The rate would naturally be subject to great variations according to local conditions. The bog iron ores are now of slight importance to the mining industry, but the easily traceable processes of their formation give us a most welcome key to the origin of other and more obscure deposits. Composition. These ores are always mixed with sand and clay and rarely contain as much as 50 per cent, of iron. The principal mineral contained is limonite, but carbonate of iron is commonly present, also phosphate as vivianite; soluble silica is sometimes recorded. In some low-grade ores from the Netherlands, the analyses of which are quoted by Clarke, 3 there is much more ferrous carbonate than limonite. Varying quantities of manga- nese are present in ores from Sweden, Finland, and Holland. The Swedish ore contains traces of vanadium, molybdenum, copper, lead, zinc, arsenic, nickel, and cobalt. All bog iron ores contain phosphorus, but there is rarely much sulphur. 1 The clogging of water supply pipes by hydroxides of these metals is a common occurrence. 2 A. Geikie, Text-book of geology, 4th ed., 1903, p. 187. 3 F. W. Clarke, Geochemistry, Bull. 616, TJ. S. Geol. Survey, 1916, p. 530. 256 MINERAL DEPOSITS According to Svanberg, cited by Zirkel, 1 the average of 30 analyses of Swedish bog iron ores gave: Fe 2 3 62.57 MgO 0.19 Mn 2 O 3 5.58 P 2 O 6 0.48 SiO 2 12.64 SO 3 0.07 A1 2 O 3 3.58 Ignition 13.53 CaO 1.37 Total 100.01 Origin. 2 The agents by which iron is carried into solution are (1) carbon dioxide from the air and decomposing organisms; (2) sulphuric acid from the weathering of pyrite, and (3) organic acids derived from decomposing vegetable matter. In the ab- sence of air ferric oxide is reduced to the ferrous state and forms soluble double salts with ammonia and humic acid. Precipitation is effected in bicarbonate solutions by the escape of carbon dioxide in the air or through its absorption by plant cells. The ferrous carbonate is easily oxidized to ferric hy- droxides. In the presence of much organic matter ferrous carbonate remains in the precipitate. From ferrous sulphate solution iron is precipitated as limon- ite by oxidation and hydrolysis, or by reaction with calcium carbonate solution, in which case siderite and gypsum will result, the former oxidizing to limonite, or the iron may be precipitated by ammonium humate, always present in swamp waters, or finally by soluble calcium phosphate, in which case vivianite or other iron phosphates result. Less commonly the iron is pre- cipitated as pyrite by alkaline sulphides or hydrogen sulphide. From soluble humates iron is also precipitated by organisms, called iron bacteria, which take up these humates, as well as ferrous carbonate, and coat their cell walls with the segregated limonite, but regarding the real importance of this process we have few data. In these, as in so many other surface reactions, the ferric 1 F. Zirkel, Lehrbuch der Petrographie, vol. 3, 1894, p. 574. 2 R. Beck, Die Lehre von den Erzlagerstatten, 3d ed., vol. 2, 1909, p. 397. F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 529. Ossian Aschan, Zeitschr. prakt. Geol., 1907, pp. 56-62. C. R. Van Hise, Metamorphism, Mon. 47, U. S. Geol. Survey, 1904, p. 826. J. M. van Bemmelen, Zeitschr. anorg. Chemie, vol. 22, 1906, p. 313. J. H. L. Vogt, Zeitschr. prakt. Geol., 1894, p. 30, and 1895, p. 38. F. M. Stapff, Zeitschr. Deutsch. geol. Gesell., vol. 18, 1866, p. 86. CHEMICAL PROCESSES IN SURFACE WA TERS 257 hydroxides are probably precipitated as colloidal complexes of indefinite composition, or "gels," which in time tend to change to crystalline bodies. Much of the ferric hydroxide is doubtless transported for considerable distance in colloid form. Five species of ferric hydroxide are recognized. Arranged by increasing water they are: Turgite 2 Fe 2 O 3 .H 2 O 94.6 per cent. Fe 2 O 3 Goethite 2 Fe 2 O 3 .2H 2 O 89.9 " " Limonite 2 Fe 2 O 3 .3H 2 O 85 . 5 " " Xanthosiderite.... 2 Fe 2 O 5 .4H 2 O 81.6 " " Limnite 2 Fe 2 O 3 ,6H 2 O 74 . 7 " " Both gothite and turgite are red and may be mistaken for hema- tite. Limonite and gothite, and probably also the other com- pounds, soon acquire crystalline properties and then show strong double refraction and fibrous texture. There are several complex ferric silicates and sulphates, which look somewhat like limonite. Examples. Many occurrences of bog iron ores are known in the United States. At Radnor and Drummondville, Three Rivers district, Quebec, the occurrences are extensive. The ores con- tain 0.3 per cent, phosphorus and less than 0.1 per cent, sul- phur. This iron ore was utilized until 1911, being dug in the swamps or dredged in the lakes; 23,000 short tons were mined in 1907, but in 1911 the operations ceased. The production in this district began in 1733. One of the most famous deposits formerly mined is at Katahdin, Maine. Small deposits are found at very many places in New England and have been worked on a small scale. Near Portland, Oregon, at the Prosser mine, limonite ore was found in the sur- ficial hollows of a basalt flow, covered by a later flow of the same rock; it was 6 to 15 feet thick and contained roots and trunks of trees. 1 The earlier basalt was rich in iron and its decom- position furnished the iron to the swamps which covered its surface. THE SIDERITES OF MARINE AND BRACKISH-WATER STRATA Occurrence. Siderite (FeCO 3 ) is an iron ore of some impor- tance, both in epigenetic and syngenetic deposits. It occurs in fissure veins and as a replacement of limestone, but is also found in the sedimentary rocks as a product of the sedimentary proc- 1 J. F. Kemp, Ore deposits of the United States and Canada, 1900, p. 92. 258 MINERAL DEPOSITS esses. The sedimentary siderite ores are called clay ironstone, spherosiderite, or black band. A dense or fine-grained con- cretionary structure is characteristic of the "clay ironstone" occurring in clays or shales and these concretions, more or less ad- mixed with clay and sand and often inclosing vegetable remains, are found abundantly at certain horizons. The variety called "black band" forms continuous beds of dark-colored, compact appearance, in the shales of the coal measures, and often directly underneath or above the coal beds. These ores contain less than 48 per cent, of iron and must be calcined before smelting. Both sulphur and phosphorus are present, sometimes in considerable quantities. Marcasite, pyrite, arsenopyrite, millerite, galena, blende, and chalcopyrite are sometimes found along cracks in the concre- tions of siderite, indicating that the iron solutions carried small amounts of the less common metals, probably as sulphates. After the deposition of the siderite these metals were leached and redeposited as sulphides along available openings. An analysis of siderite ore from Maryland 1 showed 36.05 per cent. Fe, 13.53 per cent. SiO 2 , 6.47 per cent. A1 2 3 , 0.94 per cent. Mn, 0.08 per cent. P, and 0.42 per cent. S. The economic importance of these ores, formerly great, is now small. Near the surface they are sometimes changed to limonite. * The origin of sedimentary siderites is explained along the same lines as that of the bog iron ores. Solutions of .ferrous bicarbon- ate were supplied to the marshes along the sea coast or to the shallow sea where organic matter was abundant. Precipitation of the normal insoluble carbonate took place through absorption of the solvent CQ% by vegetation. Free oxygen was absent, for otherwise the carbonate would have been transformed into limon- ite. Even if the iron had originally been deposited as limonite a reduction and carbonation to siderite may have been effected by the limonitic precipitate being covered by mud containing organic matter. The concretionary ores are not the products of primary pre- cipitation, but are, probably in all cases, segregated into nodu- lar form by the action of percolating solutions around a suit- able nucleus while the sediments were still soft. 1 J. T. Singewald, Econ. Geol, vol. 4, 1909, pp. 530-543. J. T. Singewald, Report on the iron ores of Maryland, Maryland Geol. and Econ. Survey, vol. 9, pt. 3, 1911. CHEMICAL PROCESSES IN SURFACE WATERS 259 Examples. In the United States sedimentary siderites are known in Pennsylvania, Ohio, West Virginia, Maryland, and Kentucky. Their present industrial importance is slight, but they were formerly mined on a more extensive scale. The production in 1916 was only 1,800 long tons, chiefly from Ohio. In Pennsylvania and adjacent States the upper barren Coal Measures contain abundant nodules of siderite in the shales and sandstones, but no valuable deposits In the upper productive Coal Measures, or Monongahela River series, black band ore occurs, for instance, just below or above the Pittsburg coal bed. In the lower Coal Measures the siderites are especially abundant; in Ohio 12 horizons of black bands and concretionary ores are distinguished by Orton. 1 Siderite ores also occur in the Tertiary Claiborne formation of Mississippi. The black bands are common in Germany, but are not mined extensively. They were formerly actively worked in Westphalia and near Saarbriicken, where the ore formed flat lenticular masses as much as 1% meters thick and sometimes several hundred meters in extent. In England the black bands were formerly of the highest im- portance and 40 years ago furnished four-fifths of the total iron output. They are now mined only in North Staffordshire and in Scotland. In Wales the black bands occur in the lower Coal Measures. Kendall 2 enumerates 75 horizons of siderite ore. In Scotland (Ayrshire) the black bands occur both in Coal Measures and in Carboniferous limestone. The ores contain 25 to 40 per cent. Fe, and occur as thin strata, 1^ feet or less thick; several of them are usually close together. THE JURASSIC SIDERITES OF ENGLAND The carbonate ores of the Jurassic "oolite" in England have a much greater importance than formerly, 3 the mine production Geol. Survey, vol. 5, 1884, p. 378. 2 J. D. Kendall, The iron ores of Great Britain and Ireland, London, 1893, pp. 145-199. 3 Henry Louis, The iron ore resources of the world, Int. Geol. Congress, Stockholm, 1910. W. G. Fearnsides, British iron ore resources, Min. Mag., London, Nov., 1917, pp. 241-243. W. Barnes, Mining iron ore in the Midlands, idem, March, 1918, pp. 120-126. 260 MINERAL DEPOSITS being about 11,000,000 metric tons, out of a total output of 15,000,000 tons. While the ores are of low grade they are cheaply mined, largly in open cuts. The largest yield comes from the Cleveland Hills, in the Yorkshire district. The ores form three or four beds, in the shales and sandstones of the Lower Oolite, Upper Lias, and Middle Lias; the thickest bed attains 13 feet with several minor clay streaks. The ore is changed to limonite near the surface, but the bluish green unaltered rock is composed largely of oolitic siderite; a little glauconite is present. Its percentage composition is approximately as follows: SiC>2, 10 to 20; FeO, 40; Fe 2 3 , 1.4; CaO, 1.5; CO 2 , 25; P 2 5 , 0.5 to 2. There is little sulphur and the metallic iron varies between 29 and 35 per cent. A little magnetite is reported in the ore. THE OOLITIC MARINE LIMONITES AND HEMATITES The oolites (name derived from the semblance to fish-roe) con- sist of small rounded grains of concretionary origin, each grain often being formed around a small sand-grain or around a small fossil fragment. They are found in shallow water near the shore, where the action of waves and currents is strong. Oolitic limestones are common occurrences in some sedimentary series. The oolitic iron ores consist of limonite, of hematite, of siderite or of iron silicates. Frequently all these occur together. The concretions are cemented by calcite or siderite or more commonly by an argillaceous substance. THE OOLITIC LIMONITES Occurrence. The oolitic limonites form well-defined and extensive beds in purely sedimentary series of sandstones, shales, and marls. Though several ore beds are usually present in each district they are not always persistent, but may thin out, other beds appearing at different levels. The ores have no relation to volcanism, though in many cases the decay of volcanic rocks may have supplied the iron. Though not particularly characteristic of any one formation the ores are most abundant in Jurassic strata. The percentage of iron is low and that of phosphorus high; favorable features are the presence of calcium carbonate, which makes the ores self-fluxing, and the great extent of the beds. The great iron industries of Germany and France are largely- dependent upon the oolitic limonites. CHEMICAL PROCESSES IN SURFACE WATERS 261 Examples. The so-called "minettes," 1 or oolitic limonites of the German and French Lorraine and of Luxembourg, are of the highest importance as present and future resources of European iron. In France there are at least 50 mines with an annual ore production of nearly 20,000,000 metric tons (1913). The proved reserves are estimated at 3,000 million tons. In German Lorraine, now ceded to France, the production attains similar figures and the estimated reserves are over 2,000 million tons. Dipping gently westward the strata attain a depth of 3,000 feet or more in France. The present mining is done at a depth of 700 feet or less, and in part by tunneling or open cuts (Fig. 91), The ores lie in the Middle (Dogger) part of the Jurassic systems and occur with shales, sandstones, and marls as distinct bed. FIG. 91. Section through the minette measures at Esch; 8, Calcar- eous layer with Harpoceras humphriesianum; 7, calcareous layer with Harpoceras sowerbyi; 6, marl with Harpoceras murchisonae; 5, the minette measure group (see legend); 4, sandstone with Trigonia natris; 2 and 3, upper and lower clays with Harpoceras striatulum; 1, Lias (micaceous marl). After W. Branco. within a vertical distance of 75 to 150 feet. The strata are not absolutely persistent at the same level, but are local accumula- tions, thinning out in lenticular manner. The several beds known are of different thickness, the maximum being 15 feet. A low percentage of iron, varying from 31 to 40, is characteristic, likewise a high percentage of phosphorus, varying from 1.6 to 1.8, the latter making the ores available for the basic process. From 5 to 12 per cent. CaO and from 7 to 33 per cent. Si0 2 are present. The ores are earthy and soft and are of brown, gray, or yellow 1 P. M. Nicou, in "iron ore resources of the world," Int. Geol. Congress, Stockholm, 1910. L. van Werveke, Zeitschr. prakt. Geol, 1895, p. 497, 1901, pp. 396-403. 262 . MINERAL DEPOSITS tints. Limonite forms the bulk of the ore, but there is always calcite and some siderite, often also a little secondary magnetite, and more rarely grains of pyrite, zinc blende, galena, and chalco- pyrite. The small concretions of the size of a pin-head, or a little larger, consist of limonite but, like the Clinton oolites of hematite, have a skeleton of silica, which, according to van Werveke, points to a probable derivation by alteration from glauconite. The cement consists of silica, lime, or clay shale, and grains of glauconite, a ferri-potassic silicate, occur in it. Less important oolitic ores occur in other parts of Europe, likewise in the Jurassic system. Cretaceous oolites in which the limonite is probably derived from glauconite or siderite have been described from Texas. 1 Origin. The origin of these limonites is a disputed question. As already indicated some observers doubt the direct deposition of limonite in the sea water, but hold that the mineral resulted from the oxidation of oolitic siderite or glauconite. In a recent detailed monograph L. Cayeux 2 emphasizes the peculiar fact that the limonitic oolites of France are confined to the older, pre-Cretaceous formations, while in the Cretaceous or later beds the glauconites predominate. He believes that the older oolites are in all cases derived from siderite by replace- ment and oxidation. THE MARINE OOLITIC SILICATE ORES A number of silicates of iron are distinctly sedimentary prod- ucts and common in many waterlaid series of rocks; the most important are glauconite (greensand), chamosite, thuringite, and greenalite. Glauconite seems especially abundant in the Cretaceous, chamosite and thuringite in the Silurian, and green- alite in the Algonkian, but none of them are confined to any definite horizon. Their composition is uncertain and variable, the glauconite KFeSi 2 6 .H 2 alone being distinguished by a large percentage of potash. Glauconite 3 forms in marine deposits on the present sea 1 R. A. F. Penrose, Bull. Geol. Soc. Am., vol. 3, 1891, p. 47. 2 L. Cayeux, Les minerals de fer oolithique de France, Paris, 1909. 3 C. W. von Giimbel, Sitzber. Akad. Mlinchen, vol. 16, 1886, p. 417; vol. 26, 1896, p. 545. L. Cayeux, Contribution 6tude microgr. des terr. se'd., Lille, 1897. W. A. Caspari, Contributions to the Chemistry of the Marine Glauconite, Proc. Edinb. Roy. Soc., vol. 30, 1909, p. 364. CHEMICAL PROCESSES IN SURF A CE WA TERS 263 bottom and also occurs scattered in marine sands of older forma- tions from the Cambrian to the present time, sometimes so abundantly that the rocks are termed greensands. The Creta- ceous greensands of New Jersey form a good example of rocks con- taining abundant glauconite; they are rich in both phosphorus and potassium. The mineral occurs as dark-green granules, often in the interior of shells. According to Murray and Renard, glauconite is formed just beyond the limits of wave and current action, where the muddy APPROXIMATE COMPOSITION OF SEDIMENTARY IRON SILICATES Glauconite Chamosite Thuringite Greenalite SiO 2 53 29 24 30 A1 2 O 3 10 13 17 Fe 2 O 3 21 6 15 35 FeO 2 42 33 26 CaO 1 MgO 3 K 2 4 H 2 6 10 11 9 deposits begin. Organic matter is believed to reduce the iron in the mud to sulphide, which later oxidizes to limonite; at the sa'me time colloidal silica is set free and the colloidal limonite absorbs this as well as potash, forming ferric silicate. 1 In part, however, as shown by Cayeux, the glauconite has been formed somewhat later than the original deposition of the beds and without the intervention of organic matter. Greenalite occurs abundantly, according to Leith 2 in the Algonkian ferruginous cherts of the Mesabi and other iron dis- tricts; it was formerly mistaken for glauconite. It is believed to be a marine deposit. Neither glauconite nor greenalite rocks form iron ores, but the former may be transformed by alteration into limonite, and, according to Van Hise and Leith, the greenalite is the source from which the hematites of the Lake Superior region were in part derived. Chamosite and thuringite 3 form the principal ore minerals in 1 F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 516. The extensive literature is here summarized. 2 C. K. Leith, Mon. 43, U. S. Geol. Survey, 1903, pp. 237-279. 3 A. W. Stelzner and A. Bergeat, Die Erzlagerstatten, 1, 1904, p. 201. E. R. Zalinski, Neues Jahrb. 1904; Beil. B., 19, pp. 40-84. 264 MINERAL DEPOSITS a number of interesting deposits in Thuringia (Germany) and Bohemia, which formerly were mined extensively and which are still being mined on a small scale in the latter region. These sili- cates form oolitic grains in slightly metamorphosed clay slates of the Lower Silurian. The beds still retain fossils. In Germany the ores occur as lenticular beds as much as 7 feet thick. In Bohemia the Silurian series consists of slates, graywacke, and diabase tuffs. These contain beds of oolitic hematite, one bed being 16 feet thick, while other beds consist of oolitic chamosite of considerable thickness. The latter show a groundmass of siderite or chamosite, in which are embedded oolites of dark-gray chamosite. The ores are rich in phosphorus and also carry a little magnetite. Many believe that the iron is derived from the decomposition of the associated diabase tuffs. Be that as it may, these iron ores are certainly of sedimentary origin. THE MARINE OOLITIC HEMATITE ORES Occurrence. Oolitic hematite ores of undoubted sedimentary origin are common in many parts of the world, as, for instance, Germany, France, and Bohemia. They are usually associated with Paleozoic rocks, but appear to be lacking in Mesozoic and Tertiary sediments. Siderite and calcite usually accompany them. Rarely, if ever, do they contain magnetite or metallic sulphides. Differing opinions are expressed as to their origin; they have been explained as replacements of limestone or of sid- erite, or again as primary sediments, the tendency in the United States being in favor of the latter theory of origin. The Clinton Ores. 1 The most important oolitic ores in the United States are those of the Clinton formation in the Appala- chian States; they persist with remarkable regularity wherever 1 C. H. Smyth, Jr., On the Clinton iron ore Am. Jour. Sci., 3d ser., vol. 43, 1892, p. 487. E. F. Burchard, The Clinton iron ore deposits of Alabama, Trans., Am. Inst. Min. Eng., vol. 39, 1908, pp. 997-1055. E. F. Burchard, The red iron ores of East Tenn., Bull. 16, Geol. Survey Tennessee, 1913. Burchard, Butts, and Eckel, The Birmingham district, Alabama, Bull. 400, U. S. Geol. Survey, 1910. D. H. Newland and C. A. Hartnagel, Iron ores of the Clinton formation, Bull. 123, New York State Mus., 1908. CHEMICAL PROCESSES IN SURFACE WATERS 265 this formation appears. The Clinton (Silurian) lies between the Trenton limestone and the Devonian shale, and it invariably contains one or several beds of hematite ore alternating with limestone and shale. The succession of sedimentary rocks in the Birmingham district is as follows. In a general way the section applies to the entire southern Appalachian region. Carboniferous : Feet Pennsylvanian: Pottsville formation ("Coal Measures"). 2,600 to 7,000 Unconformity. Mississippian: Parkwood formation to 2,000 Pennington shale (30-300 feet) \ . > Floyd shale 1,000 Bangor limestone (670 feet) J Fort Payne chert 200 to 250 Unconformity. Devonian : Chattanooga shale x to 25 rrog Mountain sandstone J Unconformity. Silurian: Clinton (Rockwood) formation 250 to 500 Unconformity. Ordovician: Chickamauga (Pelham) limestone 200 to 1,000 Unconformity. Cambro-Ordovician : Knox dolomite 3,300 Cambrian : Conasauga (Coosa) limestone 1,000 + Rome (Montevallo) shale (great thickness). The Clinton ores extend from western New York, through Pennsylvania, Virginia, West Virginia, Kentucky, Tennessee, and northwestern Georgia, into Alabama, where, near Birmingham, they attain their greatest development. The ores constitute beds or lenses at various horizons in the Clinton formation, which forms a striking unit of red shallow water deposits under- lain and covered disconformably by great thicknesses of limestone of the Cambrian and Mississippian ages. Thin beds of ferrugi- nous sandstone, shale and oolitic hematite make up the formation, with frequent cross bedding and some conglomerates. The ores contain calcite and in some places show gradual transition to limestones. The average thickness of the ore beds is only two or three feet but in Alabama they reach 20 feet of merchantable ore with occasional thin shale or sandstone partings. Single ore beds may extend for many miles. In the Birmingham district the Clinton 266 MINERAL DEPOSITS beds outcrop on the east flank of an anticline and they can be traced continuously northward into Tennessee, but the ac- tively working mines extend only for 15 miles along the outcrop. Good ore beds have been found by drilling for several miles eastward but toward the west the formation becomes more calcareous. Four beds are known within 80 feet in the upper part of the formation two of which are worked, with a thickness of from 9 to 20 feet. The iron ores are generally sharply bounded by shale or sandstone; in places they form transitions into ferruginous sandstone. An important iron industry is based upon the deposits in Alabama and the annual production of ore has now attained about FIG. 92. Section showing Clinton iron ores, Birmingham, Alabama; Sc, Clinton (Rockwood) formation, Silurian Oc, Chickmauga (Pelham) lime- stone, Ordovician; Cfp, Fort Payne chert, Mississippian. After E. F. Burchard. 5,000,000 long tons, or about 8 per cent, of the total output of iron ore in the United States. Mining has been carried 4,000 feet on the dip in some of the properties, and entirely similar ore has been shown to exist by borings at a vertical depth of 2,000 feet, 2 miles eastward from the outcrop. Large reserves of ore are available in this district. Clinton ores are also mined north of Alabama in Tennessee though the operations are generally confined to the enriched surface ore. North of Tennessee the mining is profitable in few places. There are several types of Clinton ores; most of them are fairly rich in calcium carbonate. One common type is a fine-grained pebbly conglomerate or sandstone, each pebble or grain coated with hematite and the rock cemented with that mineral and with calcite. Another type consists largely of fragments of bryozoa, shells, trilobites, CHEMICAL PROCESSES IN S URFA CE WA TERS 267 etc., partly coated or replaced by ferric oxide, besides an abun- dance of oolitic grains, usually with a grain of sand as the center. Still another type consists entirely of oolites of hematite in calcite matrix, averaging 1 or 2 millimeters in diameters. ^J fourth type shows small flattened hematite concretions with fragments of fossils changed to hematite; this is the "flax seed ore" which is very common at Birmingham. There is very little siderite or chlorite. Fragments of quartz and other minerals are common. The beds vary along the strike in their calcareous or siliceous admixtures. Phosphorus is present above the Bessemer limit that is above 0.05 per cent. At the surface and down to a depth of about 200 feet the calcium carbonate is in part dissolved and the ore correspondingly enriched. Such ore is called "soft," in contrast to the unaltered or "hard" variety. The poorest ores used carry 25-30 per cent. iron. ANALYSES OF CLINTON ORES (E. C. Harder, Mineral Resources, U. S. Geol. Survey, 1908) Hard ore Soft ore Fe 37.00 50.44 SiO 2 7.14 12.10 A1 2 O 3 3.81 6.06 CaO 19.20 4.65 Mn 0.23 0.21 S 0.08 0.07 P 0.30 0.46 The origin of the Clinton ores is a much discussed subject. The principal views demand either a direct sedimentary origin or a derivation by replacement of limestone. The latter ex- planation is supported by Rutledge, 1 who states that progressive steps in the transformation of limestone to ore may be followed in the field, in thin sections, and in analyses. In view of the constant character of the ore at great depth it is clear that if replacement has occurred at a comparatively late date it has at least not proceeded from the surface. The theory of direct sedimentation is held by C. H. Smyth, who contributed a notable paper to the question of origin. Similar views are advocated by Newlands, Eckel, and Burchard. Smyth thinks that the iron was carried out into shallow marine basins and was there slowly oxidized and precipitated ^mechanically 1 J. J. Rutledge, The Clinton iron ore deposits of Stone Valley, Pennsyl- vania, Trans., Am. Inst. Min. Eng., vol. 39, 1908, p. 1057. 268 MINERAL DEPOSITS around the shells or replaced them. S. W. McCallie believes that the original ore was glaticonite or greenalite, citing as evidence the delicate skeleton of silica remaining when the oolite is dis- solved in acid. The Brazilian Hematites. In the pre-Cambrian metamor- phosed sediments of Minas Geraes in Brazil 1 there are thick beds of rich hematite in a formation of ferruginous sandstone (itabirite) underlain by heavy quartzite. The origin of this undoubtedly sedimentary hematite, which as yet has not been mined, is in doubt. There is no oolitic structure, nor are there fossils. Harder and Chamberlin state that "not having much confidence in the hypothesis that the iron oxide was precipitated directly from sea water by ordinary chemical means we prefer to turn to the iron bacteria as perhaps forming a better hypo- thesis." The Oolitic Hematite-chamosite-siderite Ores. Ores contain- ing hematite, chamosite and siderite have been described from several places 2 and are in all cases of marine shallow water origin. The description by A. O. Hayes of the Wabana ores in Newfoundland is of particular interest. The ores occur in the upper 1,000 feet of flat dipping Ordovician sandstone and shale and contain several workable beds from 10 to 30 feet in thickness, one of which has been mined for a distance of one and one-half miles under the sea. The ores look like hematitic oolites and contain some fragments of marine shells but there is little calcite and no limestone. In average there is, in per- cent., 50-70 hematite, 15-25 chamosite, 0-50 siderite, 0-1 calcite and 1-10 quartz. The oolites consist often of concentric shells of hematite and chamosite such as shown in Fig. 93, and are frequently embedded in a matrix of siderite. The hematite con- cretions, upon treatment with HC1, yield a residual skeleton of silica. It is shown that borings of alga3 penetrate both oolites and 1 E. C. Harder and R. T. Chamberlin, The geology of Central Minas Geraes, Brazil, Jour. Geol, vol. 23, 1915, Nos. 4 and 5. 2 L. Cayeux, Les minerals de fer oolithique de France, Ministere des Travaux publiques, Paris, 1907. W. T. Dorpinghaus, Erzlagerstatten vom Chamosittypus ... in der nordspanischen Provinz Leon, Archiv fiir Lagerstatten-forschung, Berlin, 1914. A. O. Hayes, The Wabana iron ore of Newfoundland, Mem. 78, Geol. Survey Canada, 1915, CHEMICAL PROCESSES IN SURFACE WATERS 269 matrix and that thus the ore was practically in its present con- dition when covered by later sediments. Oxygen given off by these algae may have caused oxidation of chamosite to hematite. Direct precipitation of all three iron minerals is, therefore, advocated, though siderite is believed to be the latest and may replace chamosite. Of exceptional interest are thin beds of pyritic oolite above the "Dominion" bed. This contains graptolites, and the small pyrite concretions lie in an argillaceous matrix with some crystalline quartz. FIG. 93. Ore from Silurian beds at La Ferriere-aux-Etaugs, France. Magnified 22 diameters. The oolites are chlorite With a kernel of siderite; the fine-grained cement is chlorite and siderite. a, Oolite of chlorite, in center of lighter color; partly converted into hematite on the outside, b, Nucleus of corroded pure siderite. c, Same of yellow, altered siderite. d, Grains of siderite in the cement, e, Chloritic oolite, partly crushed and invaded by cement. /, Blackish cement of chlorite and siderite. After L. Cayeux. REVIEW OF THE SEDIMENTARY IRON ORES The descriptions given above show that in marshes, lakes and rivers the hydroxides of iron, mainly limonite, are deposited, and that smaller quantities of ferrous carbonate (siderite), iron disulphide and iron phosphates may be precipitated, 270 MINERAL DEPOSITS Regarding the marine ores, it is certain that glauconite and allied iron silicates are deposited in the sea and that under special reducing conditions siderite and iron disulphide may also form. The probability is also very strong that hematite is developed, in part from oxidation of siderite and glauconite, in part by detrital processes. Whether limonite is ever formed in sea water is much more doubtful for the salt solutions have a strong dehydrating effect. 1 It is more likely that the so-called marine limonites are products of oxidation of siderite and iron silicates. The marine iron ores are all shallow water deposits and the frequent oolitic structure 2 is in part at least due to accompany- ing action of waves and currents. Many of the replacements observed have certainly occurred immediately after deposition. Some geologists like Cayeux hold that the ore was a limestone of organic origin which has been later transformed into hematite and siderite by successive replacements but there seems to be little to support this view. The part played by micro-organisms is as yet difficult to evaluate. It seems certain that many of the blue-green algse develop oxygen in their life processes which would of course promote oxidation; it is also known that sea water at all depths contains air enriched in oxygen. It is also certain that some bacteria of the Crenothrix type 3 segregate hydroxide of iron by oxidation of dilute solutions of FeCO 3 , but this process can probably not go on in sea water. Certain other bacteria accord- ing to Drew of the de-nitrifying type seem to promote the forma- tion of calcareous oolites in the sea, and similar processes may possibly under favorable conditions result in the precipitation of siderite. That either siderite or pyrite can be deposited in large bodies in the open sea must be considered very unlikely. Wherever iron disulphide is formed reducing conditions pre- 1 W. Spring, Neues Jahrbuch, pt. 1, 1899, pp. 47-62. 2 G. Linck, Die Bildung der Oolite und Rogensteine, Neues Jahrb. Beil. B. 16, 1903, p. 495. O. Reis, Geognostische Jahreshefte, vol. 22, 1909, p. 58. 3 S. Wienogradski, Ueber Eisenbakterien, Botan. Zeit., vol. 46, 1888, p. 261. E. C. Harder, Iron depositing bacteria and their geologic relations, Prof. Pap. 113, U. S. Geol. Survey, 1919. A recent paper containing many new and valuable data. F. Lafar, Technical mycology, vol. 1, 1910, p. 272. CHEMICAL PROCESSES IN S URFA CE WA TERS 27 1 vailed; the sulphide has probably precipitated as a colloid and is, therefore, neither pyrite nor marcasite. 1 Any of these oolitic deposits may, of course, have been en- riched, after uplift and erosion, by solution of calcite but the iron was certainly not introduced by atmospheric waters. It is very significant that the oolitic concretions in all these ores yield a delicate concentric skeleton of soft silica, upon treatment with dilute hydrochloric acid. This may indicate that a silicate FIG. 94. Clinton ore, Wolcott, Wayne County, New York. Magnified 20 diameters. Ore essentially formed of remains of bryozoans and crinoids. a, Fragments of bryozoans, calcareous walls preserved, interstices filed with ferric oxide; b, fragment of bryozoan encrusted with ferric oxide, the walls partially replaced by ferric oxide; c, bryozoan structure almost obliterated by ferric oxide; d, crinoid stalk replaced by ferric oxide, cells filled with calcite of uniform optical orientation; e, same, almost entirely replaced; /, calcite cement. After L. Cayeux. had been present in all cases, but a possible alternative is that gelatinous silica and iron ore were precipitated together. So we arrive at the conception of shallow bays in which coral reefs flourished or the detritus of older fossiliferous limestone 1 Bruno Doss, Melnikovit, ein neues Eisenbisulfid, Zeitschr. prakt. Geol. vol. 20, 1912, pp. 453-467. 272 MINERAL DEPOSITS was spread. Into these bays were swept, at intervals, masses of finely divided detritus from the deep mantle of decayed rock of adjacent tropical land areas, undoubtedly rich in hematite as such products always are. The water discharged from the land certainly contained ferrous bicarbonate. In this mud agitated by the waves progressed numerous and complicated reactions. Oolites and shells of calcite were replaced by siderite, which almost simultaneously oxidized to hematite. In the deeper water glauconite was probably deposited, and it also toay soon have been altered to hematite. Somewhat similar condi- tions are found to-day, for instance, on the south side of Molokai, Hawaiian Islands, where such hematite mud is spread out over a large area of shallow coral reef. SEDIMENTARY MANGANESE ORES There is much less manganese than iron in the earth's crust, the average of analyses of igneous rock calculated by Clarke showing but 0.078 per cent, of manganese. Deposits of manganese ore are also much less common than those of iron ore. Nevertheless, many spring waters carry manganese and a minute amount of it is contained in sea water. Sedimentary deposits of manganese are known, marine and lacustrine as well as fluviatile. According to experiments by E. C. Sullivan 1 the manganese in rocks is taken into solution more easily than iron, both by car- bonated water and by dilute sulphuric acid. He also finds that from mixed ferrous and manganese sulphates almost all of the iron is precipitated first by carbonate of calcium before any manganese is thrown down. Fresenius, many years ago, also found that from spring water iron is precipitated first as limonite, while the manganese remains in solution much longer. This accounts for the very general separation of the two metals in the oxidized zone. Manganese is dissolved mainly as bicarbonate and sulphate, possibly also as phosphate. It is easily precipitated by oxida- tion, generally as MnOg in the form of pyrolusite (63.2 per cent. Mn), or as slightly hydrous psilomelane or wad (an impure mixture of manganese oxides), or more rarely as manganite (Mn 2 O 3 .H 2 0). The precipitate is generally a "gel," which 1 E. C. Sullivan, quoted by W. H. Emmons in Bull. 46, Am. Inst. Min. Eng., 1910, p. 803. CHEMICAL PROCESSES IN SURFACE WATERS 273 crystallizes in time, but which appears to have a tendency to adsorb certain oxides, especially those of barium and potassium. According to F. P. Dunnington 1 an acid solution of ferrous sulphate dissolves manganese from the carbonate, as sulphate, with the separation of ferric sulphate and limonite; from the compound solution calcium carbonate precipitates the iron, but the manganese is precipitated only upon access of air. MnS0 4 + CaCO 3 + = CaS0 4 + Mn0 2 + C0 2 . Bog Manganese Ore. It has been stated above that many bog iron ores contain manganese; pure bog manganese ores are also known, though the deposits are not abundant. The material is generally earthy and soft, approaching wad in composition. In part the bog manganese consists of a skeleton of hard and glossy black ore containing cavities filled with a black powder. The deposits are rarely more than a few feet in thickness; a small occurrence near Wiekes, Montana, described by Harder, 2 lies in the flat bottom of a gulch covered by soil and underlain by ochery bog limonite. A much larger and thicker deposit occurs at Hillsborough, New Brunswick; it is said to extend over 17 acres with a thick- ness of 6K feet. An analysis shows Mn, 45.81 ; Fe, 9.95; S, 0.03; P, 0.05, and SiO 2 , 5.36 per cent. 3 J. H. L. Vogt describes a deposit in Norway, about 1 meter thick, in a little valley above a layer of sand and below a cover of peat. The manganese ore alternates with iron ocher; it con- tains Mn0 2 , 71.20; MnO, 8.08; Fe, 1.90; P 2 5 , 0.10, and S, 0.07 per cent. In many of these occurrences the rock from which the metal was leached is a granite or a quartz porphyry. Manganese in Lacustrine and Marine Beds. Many sedimen- tary beds in all parts of the world contain manganese derived from the degradation of old land areas; it occurs as carbonate and stains or concretions of dioxide in tuffs, quartzites, sandstones, clays, shales, and limestones. It is frequently contained in beds of jasper or radiolarian chert. Strongly manganiferous sedi- 1 F. P. Dunnington, Am. Jour. Sci., 3d ser., vol. 36, 1888, p. 177. 2 E. C. Harder, Manganese deposits of the United States, Bull. 427, U. S. Geol. Survey, 1910, p. 137. 3 Ann. Rept., Geol. Survey Canada, vol. 2, 1894, p. 146. E. C. Harder, op. cit., p. 171. 274 MINERAL DEPOSITS ments recrystallize to crystalline schists, the manganese assum- ing the form of rhodonite, rhodochrosite, or manganese garnet (spessartite) . The presence of manganese nodules in deep sea deposits is well known; they are considered to be rather a sub- marine product of segregation from the red pelagic mud than of chemical precipitation from the ocean. Very rarely, however, do these sedimentary rocks contain manganese of economic impor- tance, and it is only by subsequent concentration, especially effective in regions of deep secular decay, that valuable deposits are developed (pp. 338-345). An excellent example of an undoubtedly sedimentary and practically unaltered deposit is described from Newfoundland by N. C. Dale. 1 It is of little economic importance. The metal occurs as carbonate, with some MnO 2 , in nodular form, in shaly and calcareous beds of Cambrian age and is associated with calcium phosphate in nodular form, hematite spherules, and barite in crystals and blades; the psilomelane in the deposit also contains baryum. Such deposits could probably only form in shallow water mud near land areas subjected to secular rock decay. The great manganese deposits of the province of Kutais, in Trans-Caucasia, 2 are apparently sedimentary, if judged from descriptions, but it is not impossible that here, too, enrichment by decomposition has taken place. These deposits, said to be the largest in the world, are beds in Eocene clays, marls, and sand- stones, the last resting on Cretaceous limestone, on the top of an extensive plateau. The ore beds, at the base of the Eocene, are 7 to 16 feet thick, and consist of several strata of oolitic pyrolusite with cementing earthy manganese ore. They are said to extend over an area of 22 square miles. The ores average 40 to 50 per cent. Mn and 0.16 per cent. P. Drake gives a complete analysis of an ore containing, Mn0 2 , 86.25; Mn 3 4 , 0.47; Fe 2 3 , 0.61 ; NiO, 0.3 per cent., and a trace of copper. Barium is present as usual in these ores. The annual production before the war was about 1,000,000 metric tons. X N. C. Dale, The Cambrian manganese deposits of Conception and Trinity Bays, Newfoundland, Proc., Am. Philos. Soc., vol. 54, 1915, pp. 371-456. 2 C. F. Drake, The manganese ore industry of Caucasus, Trans., Am. Inst. Min. Eng., vol. 28, 1898, p. 191. E. C. Harder, op. ait., p. 208. CHEMICAL PROCESSES IN S URFA CE WA TERS 275 SEDIMENTARY PHOSPHATE BEDS 1 Composition of the Calcium Phosphates. Phosphorus enters in the average composition of igneous rocks, according to F. W. Clarke, to the extent of only 0.11 per cent., and the analyses of sediments show smaller percentages. Nevertheless, it plays a most important part in the life processes of plants and animals, in the sea and on the land, and in places its compounds accumulate in large 'masses. Its most common salt is a calcium phosphate; the phosphates of iron, aluminum, lead, and other metals are entirely subordinate. Apatite, the most common calcium phosphate, also con- tains CaF 2 or CaCl2. The formulas may be written Ca 5 (PO 4 )3F and Ca 5 (P0 4 ) 3 Cl, or 3Ca 3 (PO 4 ) 2 .Ca(F,Cl) 2 , the first part of the latter formula being the tri-basic calcium phosphate. Fluorine apatite contains 42.3 per cent. P 2 5 ; chlorine apatite, 41.0 per cent. The pure tri-basic phosphate, which is used as a standard to express the tenor of phosphate rocks, contains 45.8 per cent. P 2 O 5 . The phosphate in sedimentary rocks approaches more or less closely the tri-basic phosphate, but sometimes is almost identical with a fluorine apatite. In deposits of guano a considerable number of acid hydrous phosphates such as, monetite (CaH.PO 4 ) and brushite (CaH.PO 4 .- 2H 2 0) have been found, 2 but they have little practical impor- tance. In the same deposits various complex phosphates of 1 R. A. F. Penrose, Jr., Nature and origin of deposits of phosphate of lime, Bull. 46, U. S. Gepl. Survey, 1888. (Gives bibliography.) David Levat, Etude sur 1'industrie des phosphates, Ann. des Mines, 7, 1895, 135. X. Stainer, Bibliographic ge'ne'rale des gisements de phosphates, Ann. des Mines de Belgique, vol. 7, 1902, et seq. F. W. Clarke, The data of geochemistry, Bull. 616, U. S. Geol. Survey, 1916, pp. 519-528. O. Stutzer, Die wichtigsten Lagerstatten der Nicht-Erze, Berlin, 1911, pp. 265-461. Eliot Blackwelder, The geologic role of phosphorus, Am. Jour. Sci., 4th ser., vol. 42, 1916, pp. 285-298. W. C. Phalen, The conservation of phosphate rock in the United States, Trans., Am. Inst. Min. Eng., vol. 57, 1918, pp. 99-132. Mineral Resources, U. S. Geol. Survey, Annual publication; various authors. Mineral Industry, New York, Annual publication, various authors. 2 F. W. Clarke, Geochemistry, Bull. 616, U.S. Geol. Survey, 1916, p. 520. 276 MINERAL DEPOSITS iron, magnesium, sodium, and ammonium occur, but these also are unimportant. The mineralogical composition of the marine and residual phosphates is complex. 1 Apatite is essentially a high tem- perature mineral and has not been recognized in the marine phos- phates; in the latter hydrous carbono-phosphates play the principal part. The latter are amorphous and doubtless hard- ened colloid precipitates; they are referred to two species: collophanite (9CaO.3P 2 5 CaO.CO 2 .H 2 O + nH 2 0) and fluocol- lophanite. The crystalline minerals which in part are altered colloids, in part crusts and mammillary structures comprise dahllite and francolite both of which are similar carbono-phos- phates with or without fluorine. The marine phosphate rocks, aside from detrital impurities, contain thus calcium carbonate and calcium phosphate; shell fragments and glauconitic granules are frequently present. The poorer kinds may be classified as phosphatic sands, marls, or limestones. The richer varieties are usually oolitic, dark- colored rocks, occasionally with a peculiar whitish efflorescence, and may carry large amounts of organic matter. They are in- conspicuous and in places difficult to recognize. The specific gravity, averaging 2.9 in 70 per cent, phosphate rock, is con- siderably higher than that of limestone and may be used to aid in the identification. A rapid field assay with ammonium molyb- date is the best test. Other Phosphates. Among the iron phosphates, vivianite, Fe 3 (P0 4 )2.8H 2 0, is the best known, and it appears frequently in bog iron ores. Of the aluminum phosphates, wavellite, 4A1P0 4 .2A1(OH) 3 + 9H 2 O, and turquoise, A1PO 4 .A1(OH) 3 + H 2 0, are the best known, the former locally used as a source of phosphorus, the latter a blue semi-precious stone; both are usually products of the uppermost zone of the crust, sometimes even forming in the zone of oxidation. In a similar geological position occur the lead phosphate, pyromorphite, corresponding in formula 1 A. Lacroix, Sur la constitution mineralogique des phosphorites francaises. Compte Rendu, vol. 150, 1910, p. 1213. H. S. Gale and R. W. Richards, Bull. 430, U. S. Geol. Survey, 1910, p. 464. W. T. Schaller, Bull. 509, U. S. Geol. Survey, 1912, pp. 89-100. A. F. Rogers, Am. Jour. Sci., 4th ser., vol. 33, 1912, p. 475. A. F. Rogers, A review of the amorphous minerals, Jour. Geol., vol. 25, 1917, pp. 515-541. CHEMICAL PROCESSES IN SURFACE WATERS 277 to chlorine apatite. Other phosphates, like amblygonite, a fluo-phosphate of lithium and aluminum, monazite, and other phosphates of the rare earths, find their home in the pegma- tite dikes. This ^illustrates the variety of occurrence of the phosphates. Phosphate Deposits . The many kinds of deposits in which calcium phosphate is of economic importance are shown by the following list: 1'. Disseminated in igneous rocks or in their differentiation products of metallic ores. 2. Apatite veins, closely allied to pegmatitic dikes. 3. Marine concretionary beds. 4. Sub-aerial accumulations of animal excrements bat caves, guano islands. 5. Metasomatic deposits by replacement of limestone by means of phosphate solutions, from Nos. 3 and 4. 6. Residual concretions, by action of atmospheric waters on No. 3. Use. The principal use of the calcium phosphate is for soil fertilization, and all the classes enumerated above are so utilized. Under No. 1 comes, for instance, the apatite concentrate from the Adirondack magnetite ores; under No. 2 the apatite veins of Canada and Norway; the occurrences of the remaining classes are described below. For utilization it is necessary to transform the insoluble tri- basic phosphate into soluble form and this is generally effected by a 60 per cent, solution of sulphuric acid; 1 hence the dependence of the phosphate industry on an abundant and cheap supply of sulphuric acid, illustrated, for instance, in the establishment of large sulphuric acid plants at the pyritic copper deposits of Ducktown, Tennessee, for the treatment of the sedimentary phosphates of the Southern States. The treatment with H 2 S0 4 results in a partial decomposition, with the formation of soluble calcium phosphate, also called super-phosphate or mono-calcium phosphate (Ca.H^PO^.HaO); and also some di-calcium phosphate, which is much less soluble. The standard is 77 per cent, of the tri-basic calcium phosphate with less than 3 per cent, of alumina plus iron, but not all of the production reaches this grade. lf rhe reaction is expressed by the following formula : Ca 3 (PO4)+2H 2 SO4 = CaH 4 ( PO 4 ) 2 +2CaSO 4 . 278 MINERAL DEPOSITS Experiments show that even the tri-calcium phosphate or apatite is soluble, particularly in water containing carbon dioxide; its solubility in solutions of CaCOs or in pure water is slight, but the presence of sodium chloride increases the solubility. 1 The marked absorption of phosphoric acid by clays and soils is held to be due to the presence of colloid bodies. Production. 2 Though some phosphates are obtained from apatite deposits and from basic slags, the greater part comes from sedimentary and residual beds. In the United States, the bulk of the production comes from Florida, Tennessee, and South Carolina, in the order named; by far the most is mined in Florida. The yield of the United States in 1913 was about 3,000,000 long tons. Large quantities were exported. In 1917 owing to war conditions the output was one-third less. Scarcity of sulphuric acid and reduced exports contributed to this. The average price of Florida phosphate was $3 per ton in 1916. Of other countries Algeria and Tunis produced about 2,800,000 metric tons, and France about 400,000 tons in 1913, but these figures have been greatly reduced since the war began. The production of the guano islands of the Pacific is now compara- tively unimportant. Origin of the Phosphate Rocks. 3 As all land animals absorb phosphoric acid and segregate it as calcium phosphate in their bones and excrements, it is not difficult to understand the accu- mulation of phosphates wherever animal life is particularly abundant and undisturbed. Besides phosphates, such deposits contain much ammonia and nitrogen, except where subjected to leaching by heavy precipitation. Of this kind are the bone beds which are found occasionally in various formations and in caves. The guano of commerce is deposited by sea birds congregating in enormous numbers on desert coasts and oceanic islands, for instance, along the Peruvian and Chilean coast, on Christmas Island in the Indian Ocean, and in the West Indies. Some 1 H. E. Patten and W. H. Waggaman, Absorption by soils, Bull. 52, Bureau of Soils, Dept. Agriculture, 1908. O. Schreiner and G. H. Failyer, Absorption of phosphates and potas- sium by soils, Bull. 32, Bureau of Soils, Dept. Agriculture, 1906. 2 Mineral Resources, U. S. Geol. Survey, annual issues. 3 Sometimes described as "phosphorites" (Stelzner and Bergeat, 1, p. 442). The name phosphate rock seems more appropriate, especially as some authors (Merrill, Non-metallic minerals, 1910, p. 267) use phosphorite in a somewhat different sense. CHEMICAL PROCESSES IN SURFACE WATERS 279 of these deposits cover whole islands and in places may accu- mulate to a depth of 100 feet, and it is stated that under favorable circumstances the rate of deposition is rapid. The guano of dry climates varies greatly in texture and color, but generally is granular, light colored, and porous. It contains on an average 10.90 per cent, nitrogen, 27.60 per cent, phosphates, and 2 to 3 per cent, potash. 1 The West Indian deposits for instance, those on Navassa 2 and Sombrero islands have been leached and are in part hard and compact, in part porous and friable. The phosphate has been concentrated to 70 or 75 per cent. The material contains from 21 to 40 per cent, of phosphoric acid, 1 to 2 per cent, sulphuric acid, 20 to 45 per cent, lime, usually also much ferric oxide and alumina. The underlying limestone or igneous rock may be locally replaced by the phosphatic solutions. The marine phosphate beds also derive their material from animal life. Sea water contains phosphoric acid, though the quantity is extremely small, and likewise some fluorine, each amounting to about a little less than one part per million. 3 According to Carnot, many shells, particularly those of the older formations, are rich in phosphorus and fluorine. A Cam- brian Obohis contained 36.54 per cent. PzQ*, and 2.78 per cent. F; a recent Lingula yielded 23.20 per cent. PzOs and 1.52 per cent. F. 4 The shells of crustaceans 5 contain up to 26 per cent. Ca 3 P 2 8 . Pteropods, lamellibranchs, gastropods and protozoans also carry phosphorus. Corals likewise contain a small amount of phosphorus and fluorine, and the same substances are found in the bones and teeth of fishes. The marine sediments, then, all hold more or less of phosphates, and it is a matter of some surprise that fluorite does not more commonly occur in sedimen- tary rocks. 6 1 R. A. F. Penrose, Jr., Bull. 46, U. S. Geol. Survey, 1888. 2 E. V. D'Invilliers, Phosphate deposits of the Island of Navassa, Bull, Geol. Soc. Am., vol. 2, 1891, p. 71. 3 A. Carnot, Ann. des Mines, 9th ser., vol. 10, 1896, p. 175. 4 Andersson and Sahlbom, Ueber den Fluorgehalt schwedischer Phos- phorite, Bull. 4, Geol. Inst. Upsala, 1900, p. 79. Neues Jahrb., ref., 1903, 1, pp. 195, 197. 6 F. W. Clarke and G. Steiger, Proc. Nat. Acad. Sci., vol. 5, 1919. pp. 6-8. 6 K. AndrSe, Ueber einige Vorkommen von Flusspath in Sedimenten, Tsch. M. und p. Mitt., vol. 28, 1909, pp. 535-562. H. S. Gale and,R. W. Richards, Bull. 430, U. S. Geol. Survey, 1910, p. 463. 280 MINERAL DEPOSITS In some beds the phosphates occur disseminated in small quan- tities, in part as small concretions, in part remaining in the shell fragments. In the more valuable deposits the phosphates appear in more concentrated form and characteristically assume the forms of nodules, or concretions (sometimes of large size), or oolitic rocks built up of small oolites in part of concentric and fibrous structure. The nodules have often a shell nucleus and, as a result of enrichment, may contain more phosphate in the peripheral than in the central parts. While phosphate nodules have been brought up by the dredge from great oceanic depths, the conditions for their formation are probably best at moderate depths, near shores, where the marine life is most abundantly developed or, as pointed out by some authors, where sudden changes of temperature, owing to con- flicting currents, kill large numbers of marine organisms. The origin of the oolitic and nodular phosphate rocks, in some of which recognizable organic remains are scarce, has been discussed extensively, but is as yet not fully explained. It is believed that ammonium phosphate may form in the organic matter and that this reacts on shell remains, replacing them with calcium phosphate, which eventually accumulates in larger con- cretions. 1 These processes are likely to continue for some time at least after the sedimentation, in the yet soft sediments. 2 After the beds have been uplifted and exposed to weathering enrichment takes place easily, by the removal of calcium car- bonate. This is especially effective in regions of deep rock decay, as in the Southern States. The rock phosphates of Utah and Idaho have remained almost unaltered. The cycle of migration of the phosphates is a fascinating study. From their original home in the igneous rocks they are dissolved by surface waters and absorbed by all living things, vegetable and animal, on land and on sea. After the death of the organisms the phosphates return to the soil or to the sedimentary beds to be dissolved and used anew by other generations. !Renard and Cornet, Bull. 21, ser. 3, Acad. Belgique, 1891, p. 126. L. Kruft, Neues Jahrbuch, Beil. Bd. 15, 1902, pp. 1-65, Ref. in Zeitschr. prakt. Geol, vol. 10, 1902, p. 301. R. Delkeskamp, Zeitschr. prakt. Geol, vol. 12, 1904, p. 299. 2 L. Cayeux, for instance, presents a figure, showing a small concretion of phosphate molded against a grain of glauconite; the latter itself being formed after the sedimentation; Contrib. a l'6tude micrographique des terrains sedimentaires, Mem. Soc. g6ol. du Nord, 4, pt. 2, Lille, 1897. CHEMICAL PROCESSES IN S URFA CE WA TERS 28 1 Occurrences of Phosphate Rocks. Deposits of phosphate rock are found in the marine beds of all ages and in almost all countries, at least from the Cambrian, when the segregation of phosphoric acid by the inhabitants of the sea appears to have begun, to the Tertiary, and in the present oceans such deposits certainly continue to form. In description it is impracticable to separate the primary marine deposits from those altered by weathering. Large deposits, enriched by weathering, are worked in the Cretaceous beds of northern France. In the southwestern part of that country, in the departments of Lot and Lot-et-Garonne, phosphates occur in irregular fissures with clay in Jurassic limestone. 1 These deposits are probably formed by replacement effected by descending solutions from sedimentary phosphate beds. Phosphate beds are now mined on a large scale along the frontier of Algeria and Tunis. 2 The beds occur in the lower Eocene, which covers Cretaceous strata, and consist in part of larger concretions in marl, sometimes carrying the rich phos- phate only as a crust; other beds are formed by a soft material, consisting of small and smooth brown or yellowish grains of phosphate cemented by calcite and also containing many fossils and much bituminous matter. The thickness of the richest phosphatic stratum is said to be 10 to 15 feet. The deposits found in the United States are mainly in three regions (1) the Atlantic coast belt of Tertiary rocks in the Carolinas and Florida; (2) the Tennessee area of Silurian and Devonian strata; (3) the Utah-Idaho region of Carboniferous beds. The phosphates of the Utah-Idaho region 3 were discovered only recently, but are of great extent and prospective value; at *L. de Launay, Gites mineraux, vol. 1, 1913, p. 679. 2 M, Blayac, Description geologique de la re'gion des phosphates du Dyr et du Kouif, Ann. des Mines (9), 6, 1894, pp. 319-330. L. de Launay, Les richesses minerales de 1'Afrique, Paris, 1903, p. 206. O. Tietze, Die Phosphatlagerstatten von Algier und Tunis, Zeitsc.hr. prakt. Geol, 1907, p. 229. 8 F. B. Weeks and W. F. Ferrier, Bull. 315, U. S. Geol. Survey, 1907, pp. 449-462. H. S. Gale and R. W. Richards, Bull. 430, idem, 1910, pp. 457-535. Eliot Blackwelder, Butt. 430, idem, 1910, pp. 536-551. R. W. Richards and G. R. Mansfield, Bull. 470, idem, 1911, pp. 371-439. G. R. Mansfield, Am. Jour. Sci., 4th ser., vol. 46. 1918, pp. 591-598. 282 MINERAL DEPOSITS present, owing to difficulties and cost of transportation, they are mined only on a small scale near Montpelier, Idaho. They extend north of Ogden, Utah, into Idaho, Wyoming and Montana, and the best deposits are in the ranges which constitute the northern continuation of the Wasatch. Their position is in the Park City formation of the Upper Carboniferous (Pennsyl- vanian), which has an average thickness of 600 feet and consists of limestones, cherty in part, phosphate beds, and shales. The phosphate horizon is in the middle of the formation and the beds have an average thickness of 200 feet. (See Figs. 8, 9, and 95.) The rocks are massive brown to gray phosphatic shales and beds of rock phosphate with some limestone. The richest bed mined at Montpelier, carrying 70 per cent, or more of CagPaOs, lies at the base of the phosphate section and is 5 or 6 feet thick. It is a black to dull-gray oolitic rock, with grains of all sizes up to pebble-like bodies one-half inch in diameter. Large sections of the phosphatic beds, in places a thickness of 75 feet, carry from 30 to 50 per cent, of Ca 3 P 2 8 . The beds are folded and locally have steep dips. The rock is hard and the mining is carried on by underground operations. Very little enrichment is noted. The phosphates of western Tennessee 1 have been worked since 1894 and at present yield about 400,000 tons per annum. They are of three classes. 1 . Brown residual phosphates, resulting from leaching of Ordovician phosphatic limestones; the beds are from 3 to 8 feet thick and carry as much as 80 per cent, of tri-calcium phosphate. 2. The blue or black bedded phosphates, which oc- cur in beds of Devonian age, show variations from oolitic through compact and conglomeratic to shaly forms. The high-grade rock is seldom more than 20 inches thick. The nodular variety, which is embedded in a green sand, carries about 60 per cent. Ca 3 P 2 08. 3. The white phosphate, which is a post-Tertiary product of replacement or filling of cavities of limestone of Carboniferous age. None of it is now mined. The phosphate beds of North and South Carolina, 2 discovered 1 C. W. Hayes, The Tennessee phosphates, Sixteenth Ann. Rept., U. S. Geol. Survey, pt. 4, 1895, pp. 610-630; Seventeenth Ann. Rept.; U. S. Geol. Survey, pt. 2, 1896, pp. 513-550; Twenty-first Ann. Rept., U. S. Geol. Survey, pt. 3, 1901, p. 473. 2 G. S. Rogers, Phosphate deposits of South Carolina, Bull. 580, U. S. Geol. Survey, 1914, pp. 183-220. CHEMICAL PROCESSES IN S URFA CE WA TERS 283 in 1867, extend along the coast for a distance of 60 miles. They are contained in loose beds of Miocene age, rich in fossils. The land deposits lie at a shallow depth and consist of so-called pebble rock, a solid mass from which the calcium carbonate has been leached and partly replaced by phosphate; the solution cavities give this material the appearance of a mass of separate pebbles. 8 feet FIG. 95. Section showing beds of phosphate, Montpelier, Idaho. After Weeks and Ferrier, U. S. Geol. Survey. The rock varies from 1 to 3 feet in thickness and is covered by a green sandy marl. Similar deposits have been dredged in the rivers; they consist essentially of water-rounded fragments of the land rock. The mining is carried on by steam-shovel or dredge operations. 284 MINERAL DEPOSITS The phosphate deposits of Florida 1 are at present the most productive in the world and large quantities are, under normal conditions, exported to almost all European countries. The deposits follow in the main the northwestern coast of the State but lie some distance from the shore. There are several types, all contained in the Alum Bluff formation or above it, or in the underlying Vicksburg limestone, both formations of Oligocene age. The clays, marls and sandstones of the Alum Bluff contain in several horizons abundant smooth yellowish or brown nodules or ovules of phosphate which are considered by Matson and Sellards as of primary deposition and the source of FIG. 96. Dredging Florida phosphates. Upper bench is sandy over- burden. Photograph by F. B. Van Horn. all the other deposits. These beds are worked in some places but are not of great importance. The material is very similar to the phosphates of Gafsa in Tunis. The so-called "land pebble" deposits which are the most important are rudely stratified detrital and residual masses; they rest on the Alum Bluff formation and are believed to be derived from this source. They are believed to be of Miocene or E. H. Sellards, Fifth Ann. Rept., Florida Geol. Survey, 1913, pp. 23-80. E. H. Sellards, Trans., Am. Inst. Min. Eng., vol. 50, 1914, pp. 901-916. G. C. Matson, The phosphates of Florida, Bull. 604, U. S. Geol. Survey, 1915. See also successive issues of Mineral Resources, U. S. Geol. Survey, and Mineral Industry. CHEMICAL PROCESSES IN SURFACE WATERS 285 Pliocene age. They contain concretions of white phosphate, averaging 65 to 70 per cent, tribasic calcium phosphate, while the finer matrix often contains 20 to 30 per cent, of the same. They lie at elevations of about 100 feet and form parts of gravel beds with casts of shells, shark's teeth, and bones of mastodon, horse and rhinoceros. The average depth of deposit is about 12 feet, that of the sandy overburden up to 40 feet. The workable deposits average several acres, though some cover as much as 40 acres. The type of deposit called "rock phosphate" occurs in the Vicksburg limestone of lower Oligocene age "and rest in depres- sions on its surface. They are of secondary origin and are believed to be leached from the overlying Alum Bluff formation. The concentration was affected by chemical and mechanical means; the result is a mass of rounded or subangular phosphate concretions ("rocks" or "pebbles") in places rudely sorted in layers of coarse and fine, and contained in a matrix of sand, clay or soft phosphate. Pleistocene sand, up to 50 feet thick, covers the deposits. There are many deposits covering from 5 to 35 acres, while some are mere pockets in the limestone. The average thickness is 30 feet; transitions to the underlying limestone are sometimes ob- served. The concretions or nodules vary from a few inches to 10 feet in diameter, are close grained and light gray and some- times show cavities lined with secondary and mammillary phos- phate. Shells are rare but fragments of bones and shark's teeth occur here and there. The processes which have operated in the concentrations from the overlying Alum Bluff beds are re- placement of limestone by phosphate and solution of residual limestone with resulting slumping; in places there has been mechanical transportation. The soft phosphates are thought to be formed by replacement of porous limestone. The per- centage of "recoverable" phosphate in the deposits does not average much above 15 per cent. The so-called "river pebbles" are Pleistocene deposits in the present rivers but are not worked now. Mineralogically, the Florida phosphates are held by Matson to consist of collophanite with francolite (p. 284). The material is mined by steam shovel, by hydraulic method or by dredge (Fig. 96); it is then washed to remove the clay and afterward crushed, screened and hand-picked. 286 MINERAL DEPOSITS The chemical composition of the marine and residual phos- phates is shown in the following analyses of which I and III represent unaltered marine deposits and II a residual occurrence. ANALYSES OF PHOSPHATES I II III Insoluble 1.82 6.69 3.05 SiO 2 0.30 A1 2 O 3 0.50 2.14 1.09 Fe 2 O 3 0.26 0.61 0.64 MgO , 0.22 0.33 0.57 CaO 50.97 46.03 48.58 0.09 K 2 0.47 H 2 O- 0.48 0.79 H 2 O+ 0.57 3.47 CO 2 1.72 3.93 4.60 P 2 O S 36.35 31.50 29.74 SO 3 2.98 2.75 Cl 0.11 Fl 0.40 1.86 2.12 Organic 7 . 45 99.04 97.35 100.79 I. Crawford Mountains, Utah. Geo. Steiger, analyst, Bull. 430, U. S. Geol. Survey, 1909, p. 465. No titanium, organic matter not determined, trace chlorine. II. Florida. Land pebble, G. H. Eldridge. III. Gafsa, Tunis. O. Tietze, Zeitschr. prakt. Geol, 1907, p. 248. Analysis calculated on dry material, 3.81 per cent. H^O. PzOs equivalent to 64.93 per cent. Ca 3 P 2 O 8 ; 2.35 per cent. CaFl 2 ; 4.67 per cent. CaSO 4 ; 10.45 per cent. CaCO 3 . CHAPTER XVII DEPOSITS FORMED BY EVAPORATION OF BODIES OF SURFACE WATERS THE SALINE RESIDUES 1 INTRODUCTION The deposits thus far described have been in the nature of insoluble residues, or chemical precipitates of relatively insoluble substances in lakes, rivers, and seas. There are, however, other deposits which also may be considered as chemical precipitates in surface waters but which consist of soluble salts formed by the evaporation of waters in closed or partially closed basins. They contain the easily soluble substances leached from the crust, brought down by the rivers to oceans and lakes, and finally concentrated under certain characteristic conditions. Closed basins are typical of dry climate and of deserts. The slow crustal movements tend to create them everywhere, by folding, subsidence, and uplift, but in the deserts the streams have not the power to cut outlets and to keep the drainage lines established. On the contrary, the movement of the debris from the mountain ranges in broad alluvial fans or aprons increases the tendency toward closed basins. The dry climate accelerates evaporation and the precipitation of the salts; dust storms transport vast masses of fine detritus; blinding salt flats extend between the barren mountain chains. Thus, at present, salt beds are found in the Cordilleran deserts along the western side of the whole American continent, in the Sahara, and in the arid, central part of Asia. Similar conditions existed in the past in different parts of the world: The Permian in central Europe, the Triassic in the Rocky Mountain region, and the Silurian in eastern North America all these ages were at times characterized by arid wastes and deposition of salt and gypsum. Saline deposits may then form : (1) in bays of the sea ; (2) in lakes ; 1 George P. Merrill, The non-metallic minerals, 1910. F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, pp. 217-259. 287 288 MINERAL DEPOSITS (3) in playas or intermittent lakes ; (4) on arid slopes by rapid evapo- ration of storm waters. In the latter two classes capillary ascent of the solutions often help to bring the salts to the surface to form " efflorescences " or saline crusts as illustrated by the crusts of "alkali" (carbonate, chloride and sulphate of sodium, sulphate of calcium) which so often interfere with agriculture in dry countries. . In regions of calcareous rocks, as in the undrained basins of Mexico, soft or compact beds of calcium carbonate locally called "caliche" or "tepetate" often cover the gentle slopes below the mountains. These belong in class 4. Minor saline deposits may result from evaporation at the surface of waters from ascending springs. Bodies of soluble salts are rarely formed below the surface; but reactions may take place in a buried deposit by which new salts are formed or concentrations of disseminated substances are effected. In places it may be difficult to distinguish these strictly speaking epigenetic bodies from the syngenetic salts (p. 313). No traces of metallic ores are found in the saline residues. Minute amounts of gold and silver have been found in salt from sea water (p. 13). Regarding traces of gold in the potassium deposits in Germany the evidence is conflicting. 1 TYPES OF WATER From a geological standpoint there are two types of water in the seas and closed basins. The first, which may be called the oceanic type, contains dominant sodium chloride and is char- acteristic of the sea as well as of partly evaporated lakes in regions where sedimentary rocks prevail; the Great Salt Lake of Utah is an example. When such water is subjected to extreme evaporation, as in the Dead Sea, a "residual" type rich in magnesium chloride results. The second main type is that of generally smaller closed basins in regions of great volcanic activity; this type contains an abundance of sulphate and car- bonate of sodium, besides more or less chloride; it indicates the result of the first leaching of loose volcanic ejecta and also shows the influence of the discharge of hot springs containing 1 A, Liversidge, Jour. Chem. Soc., vol. 71, 1897, p. 298. E. E. Lungwitz, Eng. and Min. Jour., April 6, 1905. DEPOSITS FORMED BY EVAPORATION 289 sodium carbonate and borate. The water of Mono Lake, Cali- fornia, is a good example. Certain sedimentary series, such as the Cretaceous of the Western States, contain abundant alkaline sulphates. Leach- ing of these beds by atmospheric waters takes place and these products may be carried down into salt flats and small lakes. By chemical reactions (p. 59) carbonate of sodium forms from other sodium and calcium salts and the lakes often contain much of this salt besides the sulphates. Such alkali lakes occur in Wyoming, for example. Borates characteristic of volcanic regions are generally lacking in these lakes. The first, "oceanic" type of waters yields deposits of gypsum, common salt, and finally potassium and magnesium salts. The second, "volcanic" type yields soda, glauber salt, borates, probably also nitrates, as well as more or less sodium chloride. COMPOSITION OF SALTS IN WATER OF SEA AND CLOSED BASINS I II III IV V 10.45 23.34 Cl. 55 292 55 69 70 25 Br . ... 188 tr 1 55 SO 4 B 4 O 7 7.692 6.52 0.21 CO 3 207 tr Li 01 Na 30 593 32 92 6 33 K ' 1 106 1 70 1 70 Ca 1 197 1 05 5 54 Me 3 725 2 10 14 42 Fe 2 O 3 .. . tr tr SiO, 01 tr 100 100 100 54.07 12.86 0.32 4.24 23.42 25.88 37.93 ...... 1.85 tr. 0.04 5.36 0.10 tr. tr. 0.14 100 100 I. Oceanic type. Average of 77 analyses, W. Dittmar, Challenger Kept., vol. 1, 1884. Salinity 3.5 per cent. II. Oceanic type. Great Salt Lake, E. Waller, School of Mines, Quart., vol. 14, 1892. Salinity 23 per cent., subject to variations. III. Residual type. The Dead Sea, Compt. Rend., vol. 62, 1866, p. 1329. A. Tereil, analyst. Salinity 20.7 per cent. IV. Sulphate type. Devil's Lake, North Dakota, F. W. Clarke, Geo- chemistry, 1916, p. 163; H. W. Daudt, Analyst. Salinity 1.1 per cent. V. Volcanic type. Mono Lake, Bull. 60, U. S. Geol. Survey, 1890, p. 53. T. M. Chatard, analyst. Salinity 5.1 per cent. 290 MINERAL DEPOSITS The general relation of the salts dissolved in oceanic waters to those in rivers is as follows : In ocean waters: C1>SO 4 >CO 3 ; Na>Mg>Ca In river water: CO 3 >S0 4 >C1; Ca>Mg>Na. Change from river to lake water involves a concentration of chlorides and relative loss of magnesia, silica, and lime. NORMAL SUCCESSION OF SALTS When water evaporates until precipitation of the dissolved salts begins, the least soluble salts will generally fall down first, while the most soluble salts will remain in the solution until the last. Experiments by J. Usiglio 1 on sea water showed that the carbonates of calcium and magnesium, with a little ferric oxide, were precipitated when one liter of the water was reduced from one-half to one-fifth. Gypsum was precipitated when the volume was one-fifth to one-seventh, but continued in lessening amounts until only 30 cubic centimeters of the original liter remained. Sodium chloride was precipitated abundantly upon reduction of volume to 100 cubic centimeters, but continued until the volume of the water was only 16 cubic centimeters; even then some of the salt remained in solution. Chloride and sul- phate of magnesium fell down within the same limits but in increasing quantities, and the residual "bittern" contained mainly the chlorides of magnesium and potassium, bromide of sodium, sulphate of magnesium, and chloride of sodium. Naturally the quantity of NaCl greatly exceeded that of the other salts. The whole series of these salts is rarely represented in the saline deposits; the best known and almost only example of such complete evaporation is found in the great Prussian potash and salt deposits. Often, as in the "Red Beds " of the Western States, the process ceased after the gypsum was laid down, and changes of climate or invasion of the sea may have prevented the formation of sodium chloride. Actually the conditions and the results of precipitation are far more complex than the experiments mentioned would seem to show. The influence of temperature and time may vary the details of the precipitation greatly, and double or complex salts are often formed. l Annales chim. phys., 3d ser., vol. 27, 1849, pp. 92-172. DEPOSITS FORMED BY EVAPORATION 291 A saline solution containing the same salts as sea water but in different proportions would yield materially unlike results upon evaporation. In brief, temperature, concentration, and time are always factors of great importance in the origin of saline residues. The study of the stability fields of these salts has received much impetus by the labors of J. H. van't Hoff and his numerous associates, 1 undertaken mainly to elucidate the problems of the potassium deposits of Prussia. The occurrence of thick beds of anhydrite is explained by the work of van't Hoff and Weigert, 2 who established that the mineral forms from gypsum in sodium chloride solutions at 30 C. In sea water the transformation takes place at 25 C. Crys- tals of gypsum, sinking through a salt solution at that tempera- ture, are converted into anhydrite. This is an example of the more general rule of dehydration of minerals in contact with salt solutions, at temperatures considerably below their normal inversion temperature. Many minerals are deposited in nature from solutions in a lower state of hydration than is produced at ordinary tempera- tures in the laboratory. Thus natron, the ordinary sodium carbonate (Na 2 CO 3 .10H 2 O) rarely occurs as a natural product, although it is stable at temperatures below 37 C. The more common product is trona (Na 2 COs.NaHCO3.2H 2 0), which is ordinarily stable only above 37 C. From pure sodium sulphate solution mirabilite (Na 2 S0 4 .10H 2 0) is ordinarily deposited, but in the presence of sodium chloride thenardite (Na 2 S0 4 ) is formed. From a solution of magnesium sulphate in the presence of magnesium chloride kieserite (MgS0 4 . H 2 O) is precipitated instead of the heptahydrate. The presence of a co-solute, by lowering the osmotic pressure, acts in the same direction as a rise of temperature. 3 This principle is undoubtedly also applicable to minerals in rocks and veins and explains many anomalies of mineral occurrence. 1 Mainly published in the Sitzungsberichte K. preuss. Akad. d. Wiss., from 1897 to the present time. The results are summarized by van't Hoff in a book entitled "Zur Bildung der Oceanischen Salz-Ablagerungen," Braunschweig, 1905, and 1909, and in "Physical chemistry in the service of the sciences," Univ. Chicago Press, 1903. * Sitzungsber. Akad., Berlin, 1901, p. 1140. 3 J. V. Elsden, Principles of chemical geology, 1910, pp. 85-86 292 MINERAL DEPOSITS STRUCTURAL FEATURES In desert valleys filled by temporary or permanent lakes we often find a succession of salt beds of no great thickness alter- nating with detrital matter of eolian or fluviatile origin. If the basin is large and deep a considerable mass of salt may accumu- late. The laws governing the deposition of saline residues in the depressions of the deserts have been ably set forth by J. Walther. 1 The salt deposits of marine origin are frequently of great thick- ness. In some cases they aggregate 1,000 to 2,000 feet, and it will be readily recognized that any theory based on a single cycle of evaporation of sea water, containing only 3.5 per cent, of salts would meet with great difficulties. The "bar theory," presented by C. Ochsenius 2 in 1877, but already suggested by previously expressed views of Miller, Lyell and Bischof, attempts to explain these thick salt beds. Och- senius believed that salt deposits of the purity and thickness of those in central Germany could not have been formed by the flooding of a series of shallow sounds and lakes. A periodical repetition of evaporation and flooding would necessitate an im- probable number of uniform epochs of subsidence and elevation. Walther's theory involving a leaching of the small saline con- tent of old sediments and its gradual concentration in desert valleys was rejected by Ochsenius as not properly representing the conditions prevailing in the saline deposits of central Germany. The bar theory premises a bay of the ocean separated from the open sea by a practically level bar which permits only about the same quantity of water to enter as is evaporated from the surface. A dry climate and absence of fresh-water tributaries to the bay are also premised. Under these conditions the sea water entering over the bar continuously carries a new supply to the bay; the surface layers, becoming denser, always sink and the concentration continually increases in the enclosed body of water. The salt deposits on the bottom are increasing in thickness and the heavy " bittern" solution, with the remaining magnesium salts, correspondingly rises toward the surface. When these dense solutions reach the surface of the bar, the 1 J. Walther, Lithogenesis der Gegenwart, Jena, 1893, 1894, pp. 776-800. J. Walther, Das Gesetz der Wiistenbildung, Leipzig, 1912. 2 C. Ochsenius, Die Bildung der Steinsalzlager, Halle, 1877, p. 172. C. Ochsenius, Bedeutung des orographischen elementes "Barre," Zeitschr. prakt. Geol.., 1893, pp. 189-201; 217-233. DEPOSITS FORMED BY EVAPORATION 293 movement is reversed and the residual "bittern" will flow out- ward into the ocean. No accumulation of potassium-magnesium salts will occur. Should, however, the bar have increased in height just at this time, the bittern would be retained in the now closed basin and the deposition of the potassium salts would follow. 1 The Gulf of Karaboghaz, on the eastern side of the Caspian Sea, is frequently referred to as an excellent illustration of the bar theory. 2 GYPSUM AND ANHYDRITE 3 Occurrence. Gypsum (CaSO 4 .2H 2 O) and anhydrite (CaSO 4 ) usually occur in sedimentary beds as saline residues. Both are also found occasionally as gangue minerals in ore deposits and gypsum is in places a product of ascending springs or of reactions of acid waters on calcareous beds. Efflorescences of gypsum may be produced by capillary action over gypsiferous beds or along saline lakes. Anhydrite, on account of its slow transfor- mation into the hydrous compound, has no economic value, while gypsum is one of the most important non-metallic minerals. Gypsum in sedimentary deposits frequenty forms almost pure beds of considerable thickness. It appears as snow white fine grained aggregates; characterized by softness (H:2), low specific gravity (2.3), perfect cleavage and great solubility in dilute hydrochloric acid. Anhydrite likewise forms white granular aggregates, but is easily distinguished from gypsum by its greater hardness, its greater specific gravity (2.9), its pseudo- cubical cleavage and resistance to weak HC1. Anhydrite 1 H. Everding, Deutschlands Kalibergbau, Berlin, 1907, pp. 37-40. 2 F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 165. 3 F. A. Wilder, Eng. & Min. Jour,, vol. 74, 1902, p. 276; Mines and Minerals, Dec., 1909; Mineral Industry, Annual issues. G. P. Grimsley, Michigan Geol. Survey, vol. 9, pt. 2, 1904. G. P. Grimsley and E. H. S. Bailey, Kansas Geol. Survey, vol. 5, 1899. E. C. Eckel, Cements, limes, and plasters, 2d ed., New York, 1907. G. I. Adams, Gypsum deposits of the United States, Bull. 223, U. S. Geol. Survey, 1904. H. Ries, Economic geology, New York, 1916, pp. 244-259 (with references). R. C. Wallace, Gypsum and anhydrite in genetic relationship, Geol. Mag., vol. 1, 1914, pp. 271-276. D. H. Newland and H. Leighton, Bull. 143, N. Y. State Mus., 1910. A. F. Rogers, Notes on the occurrence of anhydrite in the United States, School of Mines Quarterly, vol. 36, 1915, pp. 123-142. 294 MINERAL DEPOSITS slowly alters to gypsum, and many occurrences of apparently solid gypsum contain remnants of anhydrite. Beds of gypsum and anhydrite occur in many water laid forma- tions all over the world. Usually gypsum predominates but alternating beds of the two are common. Beds of anhydrite up to 300 feet in thickness are found in the Permian of central Germany in connection with the potash salts (p. 312). Anhydrite is not abundant in the United States but occurs with gypsum overlying thick salt beds in Louisiana and southern Texas; beds of anhydrite are also known from southern Cali- fornia, Nevada, Kansas, Nova Scotia, and New Brunswick. The gypsum beds of the United States are rarely more than 30 or 40 feet thick though there may be several in any one section. They are interstratified with limestone or shale; in places they are of great purity and snow, white; but frequently gypsum is also intergrown or interbedded with thin streaks of shale or limestone. The compact, translucent variety is called alabaster and is used for ornamental objects; gypsum in larger plates or crystals is called selenite. Recent surface deposits, mixed with clay are known as "gypsite." A remarkable series of gypsum beds, in part alternating with anhydrite have been described from the pre-Cambrian(?) of the Palen Mountains 1 in southern California. Economically important gypsum deposits are found in the Salina (Silurian) formation in northern New York and extend parallel to the south shore of Lake Ontario. Gypsum beds are also extensively worked in Michigan where they are of Mississippian age (Lower Carboniferous). Equally important beds of the same age are exploited in Nova Scotia and New Brunswick. Iowa, Kansas, Texas, Oklahoma, New Mexico and other states are rich in gypsum of Permian age; in the western part of this region gypsum occurs at several horizons in the "Red Beds" whose age ranges from upper Carboniferous to Jurassic. Exceptionally thick, but not easily utilized deposits of un- certain age overlie the "salt domes" (p. 310) of Louisiana and Texas. Tertiary deposits are known from California and Quaternary "gypsite" is abundant in Kansas, Oklahoma and Texas. The Tertiary beds in the basin of Paris, France, are rich in gypsum, hence the name "Plaster of Paris." 1 E. C. Harder, Bull. 430, U. S. Geol. Survey, 1910, pp. 407-416. DEPOSITS FORMED BY EVAPORATION 295 Uses. Gypsum finds extensive use in various industries. Ground in its natural state, it is employed as a fertilizer (land plaster), to counteract alkali in soils, to retard the setting of cement and for numerous chemical purposes. It is often used as a "filler" or adulterant. Most important is, however, its use as structural material. For this purpose it is calcined at 350 F. when a large part of the water is expelled. After grinding and mixing with water gypsum forms again and the whole sets to a hard mass called stucco or plaster of Paris. The use of gypsum is increasing rapidly. In 1917 the production in the United States was 2,700,000 tons. Stability and Solubility. As noted above gypsum is trans- formed to anhydrite in sea water at 27 C. In pure water it begins to change slowly to anhydrite at 66 C . At or above 27 C. , a temperature often reached in salt lakes in tropical countries alternating beds of gypsum and anhydrite may form, as indeed is often observed. In nature both hydration and dehydration takes place but the changes are very slow. The solubility of gypsum is a complicated problem owing to the existence of metastable forms the hemi-hydrate and the soluble anhydrite and it has only lately been worked out by van't Hoff and Meyerhoffer. 1 The solubility of gypsum in water reaches a maximum of 0.21 per cent, at 40 C., and decreases slightly above this temperature. At 66 C. the solubility of anhydrite is, of course, equal to that of anhydrite, but beyond this point it decreases rapidly so that at 100 C. it is 0.06 per cent, and at 200 C. only about 0.005 per cent. 2 Other calcium salts, having a common ion, depress the solubility of gypsum, but sodium chloride increases it about three times owing to formation of CaCl 2 , so that a saturated solution of NaCl can hold 0.54 per cent, at CaSO 4 at 23 C. and 0.75 per cent, at 82 C. SODIUM SULPHATE AND SODIUM CARBONATE Occurrence.- Most of the soda of commerce is an artificial product from common salt, but both the carbonate and the sulphate of sodium are often contained in saline desert lakes or in residues from such lakes. The ordinary white efflorescence on the playas of the deserts consists of these salts together with more or less sodium chloride and a little of the chlorides and 1 Summarized by Cameron and Bell, Bull. 33, U. S. Bureau of Soils, 1906. *A. C. Melcher, Jour. Am. Chem. Soc., vol. 32, 1910, pp. 50-66. 296 MINERAL DEPOSITS sulphates of potassium and magnesium; the soda lakes contain all these salts. In the United States the commercial utilization has been attempted at Owens Lake, in California, at the Ragtown lakes, in Nevada, and at the Wyoming soda lakes. T. M. Chatard's 1 work on Owens Lake, where sodium carbonate forms a little over one-third and sodium sulphate about one- seventh of the dissolved salts, showed that the order of deposition upon evaporation is: (1) trona (Na 2 CO3.NaHCO3.2H 2 0); (2) sodium sulphate; (3) sodium chloride, and (4) the easily soluble normal sodium carbonate. The deposits at Ragtown are even richer in carbonate of soda, 2 but the evaporation by solar heat did not prove successful as a commercial process. One or two of the Wyoming lake deposits are rich in soda. At Green River borings in the Wasatch sandstone (Eocene?) at depths of 125 and 700 feet disclosed well water forming an almost concentrated solution of sodium carbonate, which for a time was utilized for the manufacture of caustic soda; the process was based on reaction with caustic lime. The alkali lakes in the arid regions of Wyoming 3 and New Mexico form deposits leached from surrounding Mesozoic and Cenozoic sediments. The thickness of the salt beds amounts to 15 feet at most and they extend over as much as 100 acres. The salts consist mainly of mirabilite, epsomite, natron, and halite. Sodium sulphate is much more soluble in warm than in cold water, but as the similar variation for sodium chloride is small, "a mere change of temperature between summer and winter in salt lakes may cause mirabilite (Na 2 S04.10H 2 0) to separate out or to redissolve." The Great Salt Lake, according to Gilbert, deposits sodium sulphate during winter. SODIUM NITRATE The alkaline nitrates are very soluble salts which are found in larger masses only under exceptional conditions. Sodium nitrate is present in the soil and is produced by the so-called J T. M. Chatard, Natural soda, Bull. 60, U. S. Geol. Survey, 1890. 2 F. W, Clarke, .Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 238. S A. R. Schultz, Deposits of sodium salts in Wyoming, Bull. 430, U. S. Geol. Survey, 1910, pp. 570-589. W. C. Knight and E. E. Slosson, Alkali lakes and deposits, Bull, Wyo. Exper. Sta., 1901, p. 49. DEPOSITS FORMED BY EVAPORATION 297 nitrifying bacteria 1 or by reactions between organic nitrogenous matter and alkaline salts. Sodium and potassium nitrates of great purity are sometimes found as efflorescences and veinlets on sheltered cliffs of various rocks and in caves and are in many cases produced by organic agencies. Calcium nitrate is known from limestone caves. Naturally, nitrate deposits are most common in arid countries. In minor quantities nitrates are widely scattered in the Western States and very frequently they are associated with volcanic rocks, 2 particularly rhyolite but also tuffs, basalts and lake beds in regions of volcanic activity. The volcanic origin of these nitrates is not accepted by all writers but nevertheless it is the most probable theory advanced. There are two sources of nitrogen which may be utilized by nature for the development of ammonia salts and nitrates. 1. The nitrogen in the air, which may be fixed by organisms or by electric atmospheric discharges and entrainment in rain water. 2. The nitrogen from the interior of the earth, which possibly is contained in the magma as a nitride of boron or of some metal. At any rate the volcanic gases and exhalations frequently con- tain nitrogen and ammonia; it is held by many that a fixation of nitrogen from this source as nitrates is well possible. 3 The only place where nitrates are present in abundance is in the Atacama desert in northern Chile. 4 These wonderful deposits 1 H. S. Gale, Nitrate deposits, Bull. 523, U. S. Geol. Survey, 1912. 2 G. R. Mansfield, Nitrate deposits in southern Idaho and eastern Oregon, Bull. 620, U. S. Geol. Survey, 1916, pp. 19-44. Whitman Cross, Am. Jour. Sci., 4th ser., vol. 4, 1897, p. 118. W. Lindgren, Prof. Paper 43, U. S. Geol. Survey, 1905, p. 121. C. DeKalb, Min. and Sci. Press, May 6, 1916. 3 F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 256. 4 The literature is very extensive and only part can be quoted. L. Darapsky, Das Departement Tal-tal, Berlin, 1900. Ref. Zeitschr. prakt. Geol., 1902, p. 153. R. A. F. Penrose, Jour. Geol., vol. 18, 1910, pp. 1-32. S. H. Loram, Min. and Sd. Press, Jan. 15 and 29, 1910. W. F. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, pp. 253-259. L. W. Strauss, Min. and Sci. Press, June 13 and 20, 1914. J. T. Singewald, Jr., and B. L. Miller, Econ. Geol, vol. 11, 1916, pp. 103-114. Lorenzo Sundt, Econ. Geol, vol. 12, 1917, p. 89. A. H. Rogers and H. R. Van Wagenen., The Chilean Nitrate Industry, Bull. 134, Am. Inst. Min. Eng., Feb., 1918, pp. 505-522. Discussion, Bull. 136, pp. 845-848. S. H. Salisbury, Jr., Mineral Industry, Annual issues. 298 MINERAL DEPOSITS practically supply the world with nitrates; the annual production now (1916) amounts to nearly 3,000,000 metric tons. The deposits are situated in the provinces of Tarapaca and Anto- fagasta in the interior dry valleys between the Coast Range and the Andes, at elevations ranging from 1,000 feet to 3,000 feet, and they extend for 300 miles parallel to the coast. The lowest depressions are often occupied by salt flats with a little nitrate and, in the higher region, by borax flats. The nitrate deposits lie on the gentle slopes of the valleys. The nitrate bed is a superficial formation of considerable though irregular extent; it lies below an overburden of a few feet of loose crumbly material with subangular gravel, becoming harder toward the bottom. This overburden contains some nitrate and often much sodium chloride, sodium sulphate and gypsum, as well as a little sodium iodate. The "Caliche" or nitrate bed is a reddish brown sandy gravel cemented with salts; it averages a few feet in thickness. Below the "caliche" lies rudely stratified sand, gravel or clay, often of considerable thickness. The "caliche" averages about 25 per cent, sodium nitrate and the lower limit of workable material is placed at 15 per cent. Associated with the nitrate are a large amount of sodium chloride, more or less of the sul- phate and borates or calcium and sodium, and a small but con- stant quantity of sodium iodate. Small quantities of the nitrates of potassium, calcium and barium as well as a little calcium iodate and iodo-chromate (lautarite and dietzeite) are found. Very curious is the occurrence of a small amount of sodium perchlorate. The material mined is usually of the following composition: Per cent. Sodium nitrate 14r-25 Potassium nitrate 2-3 Sodium chloride 8-50 Sodium sulphate 2-12 Calcium sulphate 2-6 Magnesium sulphate 0-3 Sodium biborate 1-3 Sodium iodate 0.05- 1 Sodium perchlorate 0.1-0.5 Insoluble 0-50 The origin of the nitrate deposits of Chile is a much debated question and few authors are in agreement DEPOSITS FORMED BY EVAPORATION 299 The theory advanced many years ago by Pissis, the Chilean geologist, and followed lately, for instance, by Rogers and Van Wagenen accounts for the deposits by fixation of atmospheric nitrogen by thunderstorms and its descent from the Andes in the underground circulation and ascent to the surface by capillarity. Penrose and others hold that the nitrate came from beds of bird guano accumulated at the time when the Coast Range did not exist and that the nitrates were gradually leached and mingled with the salt waters of a closed basin. Others are inclined to consider the deposits caused by ordinary bacterial fixation or by oxidation of nitrogenous vegetable matter. Singewald and Miller think that the nitrates have been carried down by the ground water and emphasize that only the usual processes in operation everywhere, have been active. The accumulation is simply caused by the abnormally dry climate. All these explanations appear inadequate or forced. The nitrate deposits are not marine or lacustrine. Their extent corresponds in a most remarkable way to the Jurassic and Cre- taceous tuffs and lava flows which occupy so much space in this region and the conclusion is inevitable that there must be some causal connection . 1 It is probable that the nitrates are of volcanic origin and that the nitrogen was contained in the rocks men- tioned from which they have been leached under unusual climatic conditions. This view of the origin is also supported by F. W. Clarke. 2 The constant presence of borates is an additional suggestive fact. The world's need of iodine is now supplied by the nitrate region of Chile. The production was 709,000 kilograms in 1915. BORAXES 3 General Occurrence. Borates and other boron compounds appear in nature under conditions indicating widely differing modes of origin. As complex and insoluble borosilicates like 1 Unpublished observations by W. L. Whitehead and W. Lindgren. 2 Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 258. 3 G. E. Bailey, The saline deposits of California, Bull 24, California State Min. Bur., 1902. M. R. Campbell, Reconnaissance of the borax deposits of Death Valley and the Mohave Desert, Bull. 200, U. S. Geol. Survey, 1902. C. R. Keyes, Borax deposits of the United States, Trans.) Am. Inst. Min. Eng., vol. 40, 1909, pp. 674-710. 300 MINERAL DEPOSITS tourmaline and datolite they are disseminated in igneous and metamorphic rocks or in pegmatite dikes and fissure veins, but are here of no economic importance except that tourmaline oc- curring in this manner is sometimes utilized as a gem stone. As boric acid and borates of calcium and magnesium they ap- pear in volcanic exhalations, of which the most famous are the "soffioni" of Tuscany, Italy, from which large amounts of boric acid have been recovered. Borates, principally in the form of borax (Na2B4O7.10H 2 O) occur in hot springs and in lakes of vol- canic regions. Borax was first obtained from such lakes situated in Tibet. Von Schlagintweit reports it as a hot-spring deposit in the province of Ladak, on the headwaters of the Indus. Ac- cording to A. Forbes a calcium borate is being deposited at the hot springs of Banos del Toro, Chile. The thermal waters of the California Coast Ranges and Nevada (page 61) often contain boron, sometimes in large quantities. The borates from these springs are sometimes accumulated in little lake basins and there deposited by evaporation as borax crystals. About 40 years ago much borax was won from the Borax Lake, Lake County, Cali- fornia. The evaporated salts contained 62 per cent, sodium carbonate, 20 per cent, sodium chloride, and 18 per cent, borax. The borates occur abundantly in the playas, or shallow basins intermittently covered by water, in southern California, Nevada, Oregon, Argentina, and Chile; the salts are borax, ulexite (CaNaB 5 O 9 .8H 2 O), and colemanite (Ca 2 N 6 On.5H 2 O). The Quaternary borax beds are probably derived from leach- ing of deposits of colemanite in Tertiary lake beds, formed during volcanic epochs. The deposits in southern California now furnish most of the world's boron salts. Finally, boron is contained in sea water and appears in small quantities in the form of magnesium borates, principally boracite (MgyC^NieOao), in saline residues. Marine Borate Deposits.- The marine deposits are mainly confined to the beds of potassium salts in central Germany, but J. H. van't Hoff, Zur Bildung der ozeanischen Salzablagerungen, 1905. General: F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, pp. 243-253. C. G. Yale and H. S. Gale, in Mineral Resources, U. S. Geol. Survey, annual publication. 1 H. S. Gale, , The origin of colemanite deposits, Prof. Paper 85, U. S. Geol. Survey, 1913, pp. 3-9. DEPOSITS FORMED BY EVAPORATION 301 SANTA j -WT- RBAR ^VENTURA\ L S . AI 7 E| - E ^ FIG. 97. Sketch map showing distributions of borate deposits in California and Nevada. After H. S. Gale, U. S. Geol. Survey. 302 MINERAL DEPOSITS borates have also been observed in sodium chloride, anhydrite, or gypsum. The. principal occurrence is as boracite and several other rare borates in the carnallite region (page 313) that is, in the deposits of the last mother liquors of evaporating sea water. The boracite usually forms small crystals or concre- tions, but one occurrence is recorded of a mass weighing about 1,400 pounds. A few hundred tons of borates are annually obtained by crystallization. The boron compounds .then remained with the most easily soluble salts and were finally precipitated as a magnesium salt because of the predominance of that metal over calcium in the sea water. Borax Marshes. The deserts of San Bernardino and Inyo counties in California, and also those of Nevada and Oregon, are rich in borate deposits (Fig. 97). The desolate plains between the barren ranges contain many shallow basins, which at times, during the brief seasons of rain, are covered with thin sheets of water. The evaporation of this water leaves dazzling expanses of white salt deposit or efflorescence, some of which may become covered by the fine sand carried by the desert storms. These deposits were discovered about 1870 and for many years yielded a large production of borax at Searles Marsh (60 miles north of Barstow), Death Valley, and other places. Though enormous quantities of these salts remain they are now of little or no im- portance; the richer and more easily worked colemanite deposits have replaced them. The crusts are rarely more than 1 foot thick, the percentage of borax varying considerably. According to Bailey, the crude salt from Searles Marsh yielded 50 per cent, sand, 12 per cent, sodium chloride, 10 per cent, sodium carbonate, 16 per cent, sodium sulphate, and 12 per cent, borax. Borings showed 20 feet of clay and sand with crystals of calcium borate, underlain by a bed of solid trona 28 feet thick, and below this 350 feet of clays impregnated with hydrogen sulphide. On ground that had been worked over a new crust formed by capillary action that was thick enough to remove in 3 or 4 years. The area pro- ductive of borax amounts to about 1,700 acres, slightly depressed below the general level of the playa, on which in wet seasons stands about 1 foot of water. Though no ulexite was found at Searles Marsh, it is common in many other playa deposits, both in California and Nevada and in Argentina; it usually forms concretions of silky fiber, known as "cotton balls." DEPOSITS FORMED BY EVAPORATION 303 Tertiary Lake Beds. The borates in the marshes and playas have undoubtedly been leached from the older deposits in the Tertiary lake beds, which have been recognized at many points in Inyo, San Bernardino, Kern, Los Angeles, and Ventura counties, California. These beds yield colemanite almost exclusively and it is evident that the borax and ulexite of the marshes are mainly products of secondary reactions of the leached colemanite with the sodium salts of the playas. These colemanite deposits begin near the Pacific coast at Piru, Ventura County, and near Saugus, Los Angeles County, where, according to Keyes, they lie in a series of yellow clays and sand- stones probably Miocene in age and several thousand feet in FIG. 98. Lila C. borate mine, Inyo County, California, closed in 1915. thickness. The mineral is present as nodules in clay, and above the borate beds are strata of gypsum. Other important beds are in the foot-hills of the Calico Moun- tains north of the Mojave River; they have a steep dip and have been mined by shafts to a depth of 400 feet. These large masses of low-grade colemanite shales, with 7 to 20 per cent, boric acid, are not mined now, attention being confined to two solid beds of the mineral 7 to 10 feet thick. Rhyolite tuffs lie underneath the borate beds. f ' The richest colemanite beds - a*re, however, in the Furnace Creek region of the Funeral Range, which overlooks Death Valley, in Inyo County (Fig. 98). About 4,000 feet of Tertiary non- fossiliferous sediments are recognized here, which form a broad 304 MINERAL DEPOSITS belt obliquely crossing the range and dipping 20 to 45 N. E. The lower and thicker part, according to Keyes, consists of con- glomerates and sandstones, above which are olive-colored clays interbedded with basalts. The upper part of the clay series carries gypsum, colemanite, and thin layers of limestone. The borate beds are traceable for 25 miles. Within the colemanite- bearing beds, which may be as much as 50 feet thick, the bluish clays are thickly interspersed with milky white layers or nodules of the white coarsely crystalline mineral, mingled with more or less gypsum, according to Keyes. The solid layers may be 15 feet thick. Near by the clays are impregnated with fine particles of colemanite and yield 10 to 25 per cent, boric acid, but these low- grade deposits are not utilized at present. Production and Uses. The borate industry is now concen- trated in southeastern California and has shown great expansion in the last years. In 1916, 103,000 tons of crude colemanite were mined, averaging about 25 per cent. BaOs. The richest mineral is hand-sorted, and the poorer grades are milled, roasted, and screened, the last process effecting the separation of the colemanite from the gangue. The products are shipped direct to the sea board, where the material is manufactured into borax and boric acid. The further treatment involves boiling with soda for the manufacture of borax or with sulphuric acid if boric acid is desired. Under the influence of these new discoveries the price of borax has gradually decreased; in 1916 it was from 6 to 8 cents per pound. Borax is extensively used in industrial chemistry, in metal enameling, in medicine and in the household. Origin. The colemanite deposits which in places occur with gypsum and limestone are surely not of marine origin and can hardly be supposed to be saline precipitates from evaporating lake water. The mode of occurrence in specimens suggests replacement and it has indeed been shown lately by H. S. Gale 1 that some colemanite deposits are of epigenetic nature. The replacing boron solutions may have been leached from the sediments or they may have ascending waters in genetic connec- tion with basalt flows formed in close connection with the beds. The strontianite deposits found in similar lake beds (p. 380) form a somewhat analogous occurrence. Further investigations of the colemanite deposits is highly desirable. l Prof. Paper 85, U. S. Geol. Survey, 1913, pp. 3-9. DEPOSITS FORMED BY EVAPORATION 305 Colemanite and the allied species priceite have been formed in recent borax marshes, for instance, at Searles lake and would undoubtedly be formed from alkaline borate solutions in contact with calcite or calcium carbonate waters. Van't Hoff 1 has produced pandermite (Ca 8 B2oO38.15H 2 O) and colemanite from the heptaborate (Ca2B 6 On.7H 2 0) and states that ulexite, pandermite, and probably colemanite can be formed at temperatures from 25 C. upward. Regarding the separation of borates he states that while the first salts precipitated in oceanic waters are calcium salts, different relations exist with the borates; for these the saturation point is not reached until carnallite is precipitated. SODIUM CHLORIDE Occurrence. 2 Sodium chloride or common salt forms beds in sedimentary rocks and in most cases its derivation by evapora- tion of saline solutions is clear. Only a small part of the four million tons of salt produced in the United States is mined in solid form. Most of it is obtained from brines derived from solution of salt beds by natural waters or by water forced down into bore-holes to the saline strata; much also is produced by evaporation of sea water or water of saline lakes, such as the Great Salt Lake of Utah. . Salt beds' are present in formations of different ages, but are perhaps most common in the Permian and Triassic strata; the oldest saline rocks in the United States are those of the Silurian in New York State. As may be easily understood from the general statements on previous pages, strata of calcium sulphate are ordinarily associated with salt beds and should appear below them; owing to recurrent and shifting epochs of desiccation the order may be reversed and gypsum beds appear above .the salt. It is also very common to find crystals or streaks of anhydrite or gypsum with salt, as well as streaks of clay. In thickness 1 J. H. van't Hoff, Zur Bildung der ozeanischen Salzablagerungen, 2, 1909, pp. 45-75. 2 U. S. Geol. Survey, Mineral Resources (annual reports). See especially W. C. Phalen, Mineral Resources, 1907-1911. Mineral Industry (annual issues). G. D. Harris, Bull. 7, Louisiana Geol. Survey, 1908. J. O. von Buschman, Das Salz, Leipzig, 1906 and 1909, 2 volumes. W. C. Phalen, Technology of salt making in the United States, Bull. 146, U. S. Bureau of Mines, 1917 (with description of deposits). 306 MINERAL DEPOSITS salt beds vary enormously from the thinnest strata to masses 1,500 feet or even more in depth. A bore-hole near Speeren- berg, in the German potash region, penetrated 3,900 feet of FIG. 99. Sections of salt wells, Tully, New York. After F. J. H. Merrill. salt, but here, as in so many other places, the apparent thickness may be deceptive, being due to movements of folding and fault- ing. Besides, the plasticity of salt is remarkably great, much DEPOSITS FORMED BY EVAPORATION 307 greater than that of the accompanying clays and anhydrite, and this, as the German geologists have found, leads to most astonish- ing and confusing stratigraphic relations. From the calcium sulphate secondary sulphur often results and may form thick beds. Hydrocarbons and carbon dioxide are often contained as inclusions in the salt. The difficulties of FIG. 100. Section of lower Michigan basin. After A. C. Lane. accounting for the great thickness of salt beds have already been ref erred u to . It is evident that by evaporation of sea water with 3 . 5 per cent, salt in a basin 100 meters deep a bed less than 2 meters of salt would accumulate. The theories of Ochsenius and Walther attempt to explain this in different ways, as described on FIG. 101. Section of Permian salt formation in Kansas. From Mineral resources of Kansas. previous pages. The special conditions in Louisiana will be referred to later. Examples. Salt beds occur in New York State in the red Salina shales of the Silurian and underlie a considerable area. Much of the salt is recovered from artificial brines. The salt forms pure lenticular masses and layers interbedded with soft shales, limestone, and gypsum, the salt-bearing formation having a variable thickness up to 470 feet (Fig. 99). At Ithaca several 308 MINERAL DEPOSITS beds of salt occur at a depth of 2,244 feet with a total thickness of 248 feet. A magnesian limestone, containing gypsum, lies above. Rock salt is mined at several places, one shaft lately opened at Cuylerville being 1,100 feet deep and reaching a salt bed 21 feet in thickness. Salt has also been mined near the outcrops of the beds at Livonia. Similar beds are worked in Ohio by bore-holes and brines. The greatest salt production in the United States is derived from Michigan. 1 The salt occurs as large beds at different horizons in the Salina formation and also in the sandstones of the Mississippian or Lower Carboniferous (Fig. 100) . The salt is recovered by means of natural and artificial brines; bromine, in which these brines are unusually rich, is recovered as a by-product of the final mother liquor. Deep mining has been undertaken under considerable difficulties near Detroit. Kansas is likewise among the great producers. 2 Some salt is obtained from salt springs in the Carboniferous and on the "salt plains" leached from Permian beds. From the latter the principal product is derived; it occurs interstratified with shales, the total thickness of the salt beds being at most 500 feet. Some of the beds are said to be over 200 feet thick, but generally they are much less (Fig. 101). In the western arid States playa deposits of salt are common in the dry basins between the ranges; they are usually thin, though at Danby, in southern California, there are solid beds 22 feet in thickness, according to Bailey. 3 The most noted deposit is that of Salton, Imperial County, where, the basin lies below the level of the sea. Before the recent flooding by the Colorado River an important production was maintained here. A large area is covered by salt crusts 10 to 20 inches in thickness. Below this lies a thin mud deposit covering another salt crust. Deeper borings encountered 22 feet of black mud containing salt and soda, and this covers 270 feet of hard clay. 4 The desert regions of northern Africa and central Asia offer similar occurrences in abundance. 'A. C. Lane, Mineral Industry, vol. 16, 1907; also vol. 19, 1910. A. C. Lane, Water-Supply Paper 30, U. S. Geol. Survey, 1899. 2 M. Z. Kirk, Mineral resources of Kansas, Univ. Geol. Survey, 1898. 3 G. E. Bailey, Bull. 24, California State Min. Bur., 1902, p. 128. 4 G. E. Bailey, idem, p. 126. DEPOSITS FORMED BY EVAPORATION 309 Large deposits of impure salt mixed with clay have been worked for a long time in the Alpine Triassic of Tyrol; they lie between limestone beds. Another important saline region fringes the outside of the Carpathian chain in Roumania, Transylvania, and Galicia and is contained in Miocene sands and clays. The beds are generally greatly disturbed, brecciated, and pressed. The best-known place where mining is carried on is the celebrated mine of Wieliczka, in Galicia, now about 1,000 feet deep, which is much visited by tourists on account of the FIG. 102. Vertical section of salt dome, based on borings at Calcasieu parish, Louisiana. Black areas represent sulphur. After Kirby Thomas. picturesque and extensive workings with elaborate carvings in solid salt. The salt beds of the Stassfurt region will be described later. The Salt Deposits of the Gulf Coast. 1 The greatest salt deposits in the United States have been discovered by borings in 1 G. D. Harris, Bull. 7, Geol. Survey Louisiana, 1908. A. F. Lucas, The possible existence of deep-seated oil deposits on the Gulf Coast, Bull. 139, Am. Inst. Min. Eng., 1918, pp. 1119-1134. G. S. Rogers, Intrusive origin of the Gulf Coast Salt Domes, Econ. Geol, vol. 13, 1918, pp. 447-485. E. R. DeGolyer, The theory of volcanic origin of the salt domes, Bull. 137, Am. Inst. Min. Eng., 1918, pp. 987-1000. 310 MINERAL DEPOSITS Louisiana and the adjoining coast belt of Texas; they show many unusual features and some difficulty has been experienced in explaining their genesis. Above the low and swampy coast west of New Orleans rise a number of low mounds or knolls and below these most of the salt has been found. It does not occur in regular beds, but as enormous subterranean domes, surrounded on all sides by thick and often steeply dipping beds of Quaternary and Tertiary clays and sands. At some places a thin-bedded Cretaceous limestone appears at the surface. Fig. 102 gives a suggestion of the strange relations encountered. At Petite Anse, according to Harris, the drill shows 2,263 feet of almost pure salt, followed by 70 feet of foreign matter, below which the drill again enters rock salt of unknown thickness. On Cote Carline the drill entered salt at 334 feet and continued in salt without change till the drilling ceased at 2,090 feet. At Belle Isle 1 the Knapp Well No. 1 penetrated 2.000 feet of salt and, below this, 236 feet of anhydrite and gypsum. Another well at Humble is said to have penetrated 5,410 feet of salt. Oil, gas, and sulphur are often met in the drill-holes. Gypsum and anhydrite, in beds 200, 400, or even 600 feet thick cover the salt in some places, or again the salt may be overlain (as at Spindletop, Texas) by several hundred feet of a porous lime- stone carrying oil. The dip of the loose strata forming the outside of the dome is steep and bore-holes only a short distance from those disclosing salt may fail to encounter it. Naturally the published sections, based on a few bore-holes, are more or less problematical as to structure. These enormous salt resources are as yet little utilized. Rock salt was mined in 1915 at Weeks Island, where the shaft is 645 feet deep, and at A very Island, at a depth of about 500 feet. The shafts are sunk in heavy, wet ground until the impermeable salt is reached. In places there is considerable danger of flood- ing the mine by driving into the loose strata. According to R. T. Hill, these wonderful salt domes are de- posited by ascending solutions; the uplift of surrounding strata is caused by the hydrostatic pressure of salt solutions and oil rising through fissured rocks. According to L. Hager and A. C. Veatch, the domes are uplifts caused by laccolithic intru- sions. According to G. D. Harris, the uplifts are produced by the expanding power of growing salt crystals, the concentrated X A. F. Lucas, Trans. Am. Inst. Min. Eng., vol. 57, 1917, p. 1034. DEPOSITS FORMED BY EVAPORATION 311 solutions rising at the intersections of fissures; the salt is derived from underlying Paleozoic or Mesozoic beds. None of these views are fully convincing. A brief considera- tion of the relations of solubility of sodium chloride will show that only a very small quantity could have been precipitated as the temperature of the ascending solutions was lowered and that, therefore, the quantity of primary salt required by Harris' hypothesis would be incredibly large. Recent literature has shown the existence of many salt domes along the coast and some quite a distance inland. 1 The same kind of salt domes are also found on the isthmus of Tehuantepec back of Puerto Mexico, 2 accompanied in places by oil and gas. The foreign literature indicates that such salt domes also exist in northern Germany and in Transylvania. 3 Hopkins shows clearly that the salt dome at Palestine, Texas, is caused by a highly localized vertical uplift of quaquaversal form. Rogers and DeGolyer arrive at similar conclusions. It is probable that these domes are Permian or Carboniferous salt beds forced up through the softer sediments. This is made pos- sible by the extraordinary plasticity of rock salt, which easily yields to deformation. The nature of the force producing these uplifts remains in doubt. Composition, Production and Use. Rock salt is usually very pure aside from the occasionally occurring admixture of clay, the tenor in NaCl ranging from 96 to 99 per cent. Calcium sulphate is the principal impurity and is often present to the amount of over 1 per cent. Salt from some desert lakes contains sodium carbonate and sulphate. In 1917, 1,605,000 short tons of rock salt was mined in the United States. The total produc- tion of salt for the same year was nearly 7,000,000 short tons. The average price was $2.86 per ton. The wide range of uses of salt for culinary, preservative, and industrial purposes need not be specified. Large amounts are used in the manufacture of other sodium salts, particularly the carbonate. In 1917, 450 tons of bromine were produced in Michigan, Ohio, and West Virginia, the normal price is 50 cents per pound. 1 O. B. Hopkins, Bull. 616, U. S. Geol. Survey, 1917, p. 28. 2 Burton Hartley, Econ. Geol., vol. 12, 1917, pp. 581-588. 3 F. F. Hahn, Econ. Geol., vol. 7, 1912, pp. 120-135. 312 MINERAL DEPOSITS THE GERMAN POTASSIUM SALTS 1 If carried to its conclusion the process of evaporation of sea water will result in the deposition of the easily soluble chlorides and sulphates of potassium and magnesium, also chloride of cal- cium. Evidently this seldom takes place, in part for the reason given on page 292. Almost the only locality thus far discovered where the whole sequence of salts is present is in central Ger- many, north and south of the Harz Mountains, in formations of Permian age. These deposits are now mined by about fifty companies and yielded in 1913 about twelve million tons of potassium salts, of which 85 per cent, is used as soil fertilizer and the remainder for general industrial purposes. The value of this production is about $45,000,000. The mining is done exclusively by shafts from 1,000 to 2,500 feet deep. Circular shafts lined with concrete or iron tubing are used and the greatest caution is necessary to prevent influx of water during sinking or working; if the water once breaks in, the mine will probably have to be abandoned. The old workings are filled with waste or rock salt. Some of the products are sold for fertilizers without further chemical treatment. Carnallite is the most important of these, and next to it comes kainite; as mined, both are mixed with 30 or 40 per cent, of common salt. Other chemical products from carnallite and other salts are chloride and sulphate of potassium and potassium-magnesium sulphate. Kieserite is refined to magnesium sulphate. The larger part of the bromine production of the world is also obtained from the mother liquor resulting from the solution of the Stassfurt salts. The remaining part is derived from the brines of Michigan. The potassium salts lie as a relatively thin series of strata over a heavy bed of rock salt in the Permian and are in turn covered by Triassic sandstones and limestones, and finally by the Tertiary and Quaternary beds. They form a series of 1 H. Precht, Die Salz Industrie von Stassfurt, 1889. R. Ehrhardt, Die Kali-Industrie, 1907. E. Pfeiffer, Handbuch der Kali-Industrie, 1887. Beyschlag, Everding, Erdmann, Loewe, and Paxmann, Deutschlands Kalibergbau, 1907. Summaries: F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, pp. 221-228; JG. P. Merrill, Non-metallic minerals, 1905; R. Meeks, Mineral Industry, 1906.^ DEPOSITS FORMED BY EVAPORATION 313 faulted synclines and anticlines, in places approaching closely to the surface, but here generally changed by secondary leach- ing processes (Fig. 103). The general section is as follows, 1 counted from the bottom of the Triassic sandstone. No. 10 11 Thickness in meters 20 to 30. . . 50 1 to5 5 to 15.... 100 to 150. 40 to 90. . . 4 to 10 30 to 40... Character of strata 20 to 40.. 40 to 60 . . 300 to 500. Red clay with a little anhydrite and rock salt. Rock salt. Anhydrite with salt. Red clay with anhydrite and rock salt. Younger rock salt. Main anhydrite. Salt clay. Carnallite zone. Carnallite (KCl.MgCl,+ 6H 2 0). Kieserite zone. Kieserite (MgSO 4 + H 2 O) . Polyhalite zone. Polyhalite (2CaSO 4 .MgSO 4 . K 2 SO 4 + 2H 2 O). Older rock salt, with narrow streaks of anhydrite, interpreted as annual deposits. Older anhydrite. Limestone (Zechstein Kalk) . Marine deposit. ( Black copper-bearing shale. \ Conglomerate. Lower Permian and Carboniferous beds. The older series (Noa. 8 to 11) closed with deposition of potassium salts. The younger series (Nos. 1 to '7) contains no potassium salts. 12 I 70 to 100. 13 4 to 10... 14 0.5to4.. Rock salt is really present throughout the section, for the car- nallite zone, which yields the greatest quantity of crude product, contains only about 55 per cent, of carnallite, including also 25 per cent, rock salt and 16 per cent, kieserite. The kieserite zone yields 65 per cent, rock salt and only 17 per cent, kieserite. Rock salt, kieserite, sylvite (KC1), carnallite (KCl.MgCl 2 + 6H 2 0), and kainite (KCl.MgS0 4 +3H 2 O) are the main products. The German geologists have shown that extensive secondary changes have taken place in the salt beds in part immediately after the deposition, in part much later, following the Triassic sedimentation and progressing even now. These post-Triassic X H. Everding, Deutschlands Kalibergbau, 1907, p. 36. 314 MINERAL DEPOSITS changes have occurred both in the croppings and at greater depth. The minerals just enumerated may occur in all three generations, but in addition a large number of more or less com- plicated secondary compounds were formed. Kainite is in part a secondary product derived from carnallite by the leaching of MgCl 2 . It forms along the crests of the anticlines. Under some circumstances a secondary mixture of Kochstedt 1 2 3 4 Kilometers 1. Tertiary beds 2. Triassic shales (Keuper) 3, Triassic limestone (MuscheikalK) 4. Upper sandstone) 5. Middle " > Buntsandstem 6. Lower " J 7. Upper clay 8. Secondary salt 9. Younger salt beds 10. Main anhydrite PERMIAN TRIASSIC w 12 13. 14. 15. i& 17. Salt clay Potassium salts Older salt beds Permian limestone (Zechstein) Upper Permian sandstone ( Rotliegendes) boniferous ( Rotliegendes) niferous FIG. 103. Section of the Stassfurt-Egeln anticline. After Everding. potassium chloride, kieserite, and salt would be formed instead of kainite, and this constitutes an important product under the name of "hartsalz." Secondary deposits of the older type ap- pear between the carnallite and the salt clay. In all these transformations the products are very complex. Van't Hoff 1 and his associates have studied the various com- 1 Van't Hoff, Die ozeanischen Salzablagerungen, 1905 and 1909. DEPOSITS FORMED BY EVAPORATION 315 binations of salts in order to ascertain their fields of existence at temperatures ranging from 25 to 83 C. In this way they have arrived at the temperatures of stability of the various salts and consequently ascertained the minimum temperature at which they were formed. Sylvite and carnallite are stable in concen- trated NaCl solutions from to 85 C. Many of the rarer salts (langbeinite (2MgS04.K 2 S04), for instance) are stable under these conditions only from 37 C. upward. Some of the temperatures required may seem high; kieserite with sylvite, for instance, forming above 72 C. There is, indeed, some danger in using the laboratory results to explain the proc- esses of nature, for the important element of time probably plays a considerable part in these reactions. It has been shown by measurements, however, that the tem- perature in the middle depths of evaporating salt lakes is surprisingly high. Interesting results were obtained by A. V. Kaleczinsky, 1 who found the temperatures of certain Hungarian salt lakes to be as much as 71 C. during the summer at a depth of 1.3 meters, while the surface and the bottom layers were much cooler, about 20 C. OTHER SOURCES OF POTASSIUM SALTS The exhausted agricultural lands of all countries need potas- sium salts, 2 together with phosphates and nitrogen compounds. Germany is the only country in which potassium salts in easily available form occur on a large scale. The imports of the United States in 1913 amounted to about 1,800,000 metric tons while the highly stimulated domestic production since the war broke out amounts to but 20 per cent, of the former consump- tion. This brief statement indicates the acute situation. 1 A. V. Kaleczinsky, Ref. Ann. Phys. (4), 7, 1902. 2 W. C. Phalen, Potash salts, Mineral Resources, U. S. Geol. Survey, annual publication, 1910-16 (.with literature). W. C. Phalen, Occurrence of potash salts in thejbitterns of the eastern United States; Bull. 530, U. S. Geol. Survey, 1911. " H. S. Gale, The search for potash in the United States; Bull. 530, U. S. Geol. Survey, 1911. Also in Bull. 580, U. S. Geol. Survey, 1914, pp. 265-317. A. R. Schultz and Whitman Cross, Potash-bearing rocks of the Leucite Hills, Wyo.; Bull. 512, U. S. Geol. Survey, 1912. B. S. Butler and H. S. Gale, Alunite, Bull. 511, U. S. Geol. Survey, 1912. 316 MINERAL DEPOSITS Promising beds of potassium salts similar to those of Ger- many have been discovered in the Oligocene of Alsace. Other deposits occur in Tyrol, Spain, Galicia and India, but none of these have as yet contributed to the world's supply. An active search for potassium salts has been carried on in this countoy since 1910 Potassium in Rocks and Minerals. Granites, pegmatites, some phonolites, and some leucite rocks contain a considerable amount of potassium, varying from 5 to 12 per cent. Unfor- tunately there is great difficulty in transforming the insoluble potassium silicates contained in the orthoclase, leucite, or glassy base in these rocks into soluble salts. Some pegmatite dikes, composed largely of orthoclase, yield even more than 12 per cent, of potash. If orthoclase or any potassium-bearing rock is ground to fine powder and slimed with water a certain small per- centage of potash salt is converted into soluble form, probably as a potassium silicate, and it is said that such finely ground powder has some slight fertilizing power. Processes have been patented by A. S. Cushman and others based on electrolytic treatment of the slimed rock or treatment with quicklime and calcium chloride and subsequent calcining, methods by which soluble potassium salts are said to be set free. None of these processes has thus far been applied on a large scale. Greensand marls for instance, the Cretaceous beds in New Jersey contain from 3 to 6 per cent, of potash besides some phosphoric acid, the former in glauconite, the latter in cal- cium phosphate. These marls are used in their crude state as fertilizers and the recovery in soluble form of their potassium content has been proposed, but the practical application has not as yet been attempted. Another source of potassium has been sought in the mineral alunite 1 , which is a hydrous potassium-aluminum sulphate of inconspicuous appearance, white or pink, compact or fine granu- lar, rarely coarse granular. The formula of the mineral is K 2 O.3Al203.4S03.6H 2 0; and it contains from 8 to 11 per cent, of potash. It occurs mainly in volcanic regions, as a product of rock alteration, probably caused by waters containing sulphuric acid, and is much more common than the soluble natural alum which sometimes appears as efflorescences. The alunite is found disseminated in the rocks or in well-defined veins. Nota- ble western occurrences are at Goldfield, Nevada; Marysvale, DEPOSITS FORMED BY EVAPORATION 317 Utah; and the Rosita Hills, Colorado. Among the foreign de- posits which have already been utilized are those at La Tolfa, in Italy; Almeria, in Spain; and Bullah Delah, in New South Wales. The transformation of alunite into soluble potassium sulphate is easily effected by calcination; part of the sulphuric acid and all of the water is volatilized, leaving soluble potas- sium sulphate and an insoluble residue of alumina. The manu- facture of potassium sulphate from alunite began at Marysvale in 1915. Potassium in Brines. Potassium salts are easily soluble and therefore remain with calcium and magnesium chlorides in the last residues or mother liquors, the so-called "bitterns." Many natural brines pumped from bore-holes in salt-bearing beds con- tain some potassium and under favorable circumstances may be used for the recovery of these salts. Some of the Michigan brines from the Marshall formation of the lower Carboniferous (Fig. 92) contain from 3 to 5 grams per liter of potassium chlo- ride; salt, calcium chloride, and bromine are recovered from these brines, but their potassium content appeals to be too small for profitable recovery. In places certain well-defined strata yield natural brines or residual "bitterns." One such bittern from Fairport Harbor, in Ohio, on Lake Erie, contains, accord- ing to W. C. Phalen, in grams per liter, 7.4 KC1, 110.1 NaCl, 134.4 CaCl 2 , 43.2 MgCl 2 . Such a brine could possibly be utilized for the recovery of potassium. This stratum is almost 400 feet above the topmost salt bed from which artificial brines are pumped in Ohio. Lakes in dry regions, especially in areas of former volcanic activity, contain appreciable quantities of potassium. The water of Owens Lake, in eastern California, yields almost 3 grams of potassium chloride per liter. Both soda and potash are now recovered from the new plant at this locality. Evaporation in the Quaternary lakes of the Lahontan basin in Nevada and California has at many places resulted in deposits of salt of moderate thickness. Changes in drainage among these basins sometimes resulted in the residual brines, /richer in potash, being drawn off into a neighboring depression, and thus it happens, as at Searles Marsh, in San Bernardino County, California, that the salt bed, which here is almost 60 feet in thickness and covers an area of at least 11 square miles, is saturated with a strong brine unusually rich in potassium. In 318 MINERAL DEPOSITS the dissolved salts of Searles Marsh there is almost 7 per cent. of K 2 0. A large plant is now under construction for the recovery of potash at this place. Some small lakes in the "sand hill country" of Nebraska and Colorado contain a remarkably high percentage of potassium which it is thought may have become concentrated from the potash resulting from repeated burnings of the grass in the sur- rounding country. The Nebraska lakes now (1917) yield about one-half of the potash production of the United States. The earliest source of potassium was, as is well known, ashes of vegetable matter. Seaweed is especially rich in this metal and also contains iodine. Potash salts are now also recovered from the flue dust of cement plants. Bromine and Calcium Chloride. The very soluble salts of bromine and calcium chloride are recovered from residual salt brines in Michigan and West Virginia. The principal production before the war came from the Stassfurt salts of Germany. CHAPTER XVIII MINERAL DEPOSITS RESULTING FROM PROCESSES OF ROCK DECAY AND WEATHERING 1 GENERAL CONDITIONS The uplifted sedimentary beds, the lavas of the volcanoes, the granular crystalline rocks uncovered by erosion all these, when exposed at the surface of the earth are subject to a series of changes, the sum total of which is called weathering. The agents are water, air, heat, and vegetable and animal life. Water is essential without it very little decomposition could take place. Oxygen is also essential, and indeed we often speak of weathering as synonymous with oxidation. Carbon dioxide dissolved in water decomposes the minerals and hastens the process of solu- tion. Change of temperature acts mainly by promoting disin- tegration, most powerfully by the expansion of water when freezing in cracks and crevices, a force sufficient to break and dislodge heavy rock masses. Vegetable life furnishes carbon dioxide and disintegrates the soil by the vital energy in the roots, and bacterial life changes its composition. Animals burrow in the ground, loosening it and effecting chemical changes. Weathering differs from alteration and metamorphism in that its ultimate result is the destruction of the rock as a unit; its chemical processes are far more radical and intense than those of the depths. Weathering effaces the texture of the rocks and segregates their chemical compounds in ways wholly different from those of the other processes mentioned. Metals closely associated in the primary rocks part company and seek new associates. Segregations of large masses of single minerals are usually a result of the process. The ordinary silicates and 1 N. S. Shaler, The origin and nature of soils, Twelfth Ann. Rept., U. S. Geol. Survey, pt. 1, 1891. G. P. Merrill, Rocks, rock-weathering, and soils, 1897. C. R. Van Hise, Metamorphism, Mon. 47, U. S. Geol. Survey, 1904. F. K. Cameron and J. M. Bell, The mineral constituents of the soil solu- tion, Bull. 30, Bureau of Soils, Washington, D. C., 1905. j T. L. Lyon, E. O. Fippin and H. O. Buckman, Soils, New York, 1916, p. 764. 319 320 MINERAL DEPOSITS the carbonates of iron, magnesium, and calcium are unstable in the belt of weathering. The uppermost thin mantle of the products of weathering we call the soil; in it the disintegration and chemical changes are carried to their limit; it is mixed with the products of life, and its constituents and reactions are, of course, of more interest to the agricultural chemist than to the student of ore deposits. The depth to which weathering extends differs greatly; in some desert regions, recently glaciated areas, or areas covered by fresh lava flows it is practically absent, disintegration being the only visible effect. In regions of heavy vegetation and rainfall the weathering may extend to a depth of 100 or even 200 feet; along fractures in particularly permeable and soluble rocks like limestone oxidation may be carried to still greater depth; in mineral deposits its effects are in places felt for several hundred or in extreme cases as much as 2,000 feet. As a rule, however, weathering does not extend deeper than 50 feet, and its more intense effects are usually limited to the zone above the surface of underground water. Disintegration and decomposition work together, but the former is likely to extend deeper than the latter. The upper layers, ordinarily colored red or brown by ferric iron, gradually change into paler-colored, more or less softened and disintegrated rock. In some areas, notably over limestone strata, there is a sharp change to the unaltered rock so sharp, indeed, that the red clayey soil has often been taken for a different formation resting on the calcareous rock. Erosional transportation attends disintegration, and removal of material by solution accompanies decomposition, both tending strongly to reduce the volume of the rock. On the other hand, hydration and the peculiar quality of adsorption which the soils possess tend to increase the volume. On the whole weathering lessens the volume. According to G. P. Merrill the granites of the District of Columbia may lose by weathering 13.5 per cent, of their volume; T. L. Watson calculates the loss of granites of Georgia at 7 to 72 per cent. The most marked loss is the shrink- ing in residual clays derived from limestone; often it is more than 95 per cent. Whitney long ago arrived at the conclusion that 1 meter of residual clay in Wisconsin was derived from a thick- ness of 35 to 40 meters of limestone or shale. Except in the easily soluble rocks the decomposition is never ROCK DECAY AND WEATHERING 321 complete, for, as brought out by Cameron and Bell, even in the fine soils abundant grams of the original minerals remain unaltered. Other conditions being equal, weathering is most complete in tropical and moist countries. In the United States the most intense action of this kind has taken place in the Ap- palachian region south of the glaciated area, and this region contains the majority of ore deposits caused by weathering. Air contains approximately by volume 21 per cent, oxygen, slightly less than 79 per cent, nitrogen (with argon), and 0.03 per cent, carbon dioxide. In the air contained in rain water both oxygen and carbon dioxide are greatly concentrated. In the soils carbon dioxide and air are absorbed; soils and clays of various kinds contain from 14 to 40 cubic centimeters of gas per 100 grams, with 14 to 34 per cent, of carbon dioxide and consid- erably less oxygen than the air indeed, in some soils oxygen appears to be absent. 1 Decaying vegetation still further in- creases the percentage of carbon dioxide. As the ground-water level is approached the oxygen decreases rapidly, as shown by the measurements made by B. Lepsius 2 in bore-holes, and below this level there is probably little left. Naturally the processes of weathering are hastened by the presence of sulphuric acid derived from the decomposition of pyrite or exhaled from solfataric vents. W. Maxwell 3 has shown interestingly how extensive a part this acid plays in the develop- ment of soils on the slopes of volcanoes. The processes characteristic of weathering are oxidation, hy- dration, and solution. In the surface waters calcium and mag- nesium carbonates ordinarily prevail, with a considerable amount of alkaline carbonates and relatively much soluble silica, both derived from the decomposition of the silicates. Under special conditions, as in volcanic regions or in sediments rich in salts, the surface waters may be materially different in composition, being predominantly sulphate solutions. The ground waters contain in addition small amounts of iron and manganese, carried mainly as bicarbonates, also phosphoric acid and sodium chloride. In the weathered zone will remain the residual, almost insol- 1 CameroA and Bell, op. cit., p. 26. 2 Quoted in F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey, 1916, p. 477. 3 W. Maxwell, Lavas and soils of the Hawaiian Islands, Honolulu, 1898. 322 MINERAL DEPOSITS uble minerals, like quartz, hydrated aluminum silicates more or less closely approaching kaolinite in composition, ferric oxides (as limonite, gothite, or hematite), and manganese dioxide, all mingled to form a red or brown clayey soil. All these reactions involve the development of colloid bodies like aluminum silicates and hydroxides of iron, which before their transformation into crystalline minerals are characteristic absorbents of many salts. The colloids of manganese, for instance, have a tendency to adsorb potassium and barium. The zone of weathering has indeed been called the realm of the colloids. DECOMPOSITION OF MINERALS The silicates of the rocks are decomposed by water rather than dissolved, for the resulting solution does not usually contain the elements in the same proportions as the original mineral. Owing to hydrolysis the solution in most cases gives an alkaline reaction. 1 Cameron and Bell believe that the more rapid decomposition shown to take place in the presence of CC>2 is in large part due to its combination with the hydrolyzed bases by the formation of bicarbonates rather than to direct solvent action. Orthoclase, for instance, would give KAlSi 3 Os+HOH = KOH+HAlSi 3 O8, and the carbon dioxide would form potassium bicarbonate with KOH. Kaolin would form from the unstable silicate HAlSi 3 O 8 . The biotite, amphibole, and pyroxene, perhaps previously altered to chlorite below the water-level, break up into soluble earthy carbonates, with limonite, hydrous aluminum silicate, and silica, the latter three in colloidal state. These ferromagnesian minerals are attacked first, so that the ordinary surface waters contain more of the carbonates of calcium and magnesium than of other salts. The soda-lime feldspars come next while the alkali feldspars are more resistant. Quartz is only partially attacked. The decomposition of orthoclase is usually expressed in the following equation: The ultimate product is kaolin, or allied colloidal bodies. Muscovite or sericite do not result from weathering, although the colloidal aluminum silicate may adsorb potassium 'and form amorphous compounds related in composition to the white micas. i F. W. Clarke, Bull. 167, U. S. Geol. Survey, 1900, p. 156. ROCK DECAY AND WEATHERING 323 Zeolites are undoubtedly unstable in the zone of weathering. Muscovite or sericite is slowly attacked; Cameron and Bell 1 treated 2 grams of powdered muscovite with a liter of pure water for 14 months in paraffine cylinders and obtained in the solution 10.4 parts K per million; when treating it with carbon dioxide in water they obtained 18.3 parts per million. The same quan- tity of orthoclase with pure water yielded a solution with 1.7 parts per million of K; when saturated with COg it yielded 2.5 parts per million. Muscovite thus yields its potassium more easily than orthoclase. Albite treated in the same way gave 1.0 and 1.1 parts of sodium per million respectively. Earlier experiments leading to similar results have been undertaken by A. Daubree, R. Miiller, A. S. Cushman and F. Henrich. 2 Magnetite is soluble with difficulty, but finally yields hematite and limonite. Pyrite easily yields limonite and sulphuric acid. Apatite appears to be rather easily soluble, especially in car- bonated water. Cameron and Bell 3 treated powdered chlo- rine apatite suspended in water at 25 C. for 7 days, passing C0 2 through the liquid. The solution showed 0.0378 gram CaO, 0.0096 gram P 2 6 , and 0.0026 gram hydrochloric acid per liter of solution. In soils and clays the phosphates are decomposed or hydrolyzed, soluble phosphates being formed, but the per- colating water contains these only in minimal quantities. It is stated 4 that humus suspended in water can adsorb calcium and a considerable amount of phosphoric acid from the calcium phosphates. The reactions of the iron phosphates are in the main similar to those of the calcium salts. Lachowicz 6 found that organic matter in soil is a distinct solvent for ferric phosphate. Cameron and Bell ascertained that carbon dioxide greatly aided the solu- tion of ferric phosphate, 5 grams of which, shaken for 5 days with 100 c.c. H 2 and later with 100 c.c. of saturated solutions 1 Op. cit., p. 33. 2 A. Daubre"e, fitudes synthe"tiques de ge"ologie experimentale. i R. Miiller, Jahrb. K. k. geol. Reichsanstalt, vol. 27, 1887. F A. S. Cushman, Bull. 92, Bureau Chemistry, U. S. Dept. Agr., 1905. F. Henrich, Ueber die Einwirkung von Kohlensaurehattigen Wasser auf Gesteine Zeitschr. prakt. Geol, 1910, pp. 84-94. 3 Bull. 41, Bureau of Soils, U. S. Dept. Agr., 1907. 4 Patten and Waggaman, Absorption by soils, Bull. 52, Bureau of Soils, U. S. Dept. Agr., 1908. 6 Gesteins und Bodenkunde, 1877, p. 329. 324 MINERAL DEPOSITS of C0 2 and CaS0 4 , yielded respectively 74, 171, and 118 mil- ligrams of phosphoric acid. Zircon, pyrope garnet, tourmaline, and similar minerals are almost insoluble. Quartz also shows great resistance and appears practically insoluble in the zone of weathering, except when exposed to the action of a stronger solution of alkaline carbonates. C . W. Hayes, 1 M. L. Fuller, 2 and C. H. Smyth 3 have observed a marked corro- sion of quartz pebbles in conglomerates, but the exact nature of the reaction is uncertain. Cherty and fine-grained quartz is a little more soluble. 4 The theory of the origin of the Lake Superior iron ores, supported by Van Hise and Leith, is based on the supposed solubility of such material. It was formerly thought that certain organic acids had the power of dissolving quartz, but this is now considered very questionable. TOTAL CHEMICAL CHANGES BY WEATHERING The studies and analytical work of G. P. Merrill have greatly advanced our knowledge of weathering, and many others have contributed valuable data. A compilation of a number of characteristic gradational analyses is given in F. W. Clarke's "Data of geochemistry" 6 and allows an estimate of the final effects of weathering. The analyses show consistently an appar- ent increase in alumina and water and decreases in SiOg, CaO, MgO, K 2 O, and Na 2 O; in short the composition tends toward that of a ferruginous kaolin, except that in the weathering of acidic rocks residual quartz prohibits the decrease of silica to the amount characteristic of kaolin. Comparing equal volumes we find little actual change in the quantity of alumina, and for purposes of comparison this oxide is assumed to be constant. By recalculating the analyses on this basis, the percentage lost or gained by each constituent may be ascertained. In the analyses quoted, the loss of silica is the largest, varying from 8 to 32 per cent, by weight of the original rock and from 15 to 52 per cent, of the original quantity of silica. The abstraction of silica 1 Bull. Geol. Soc. Am., vol. 8, 1897, p. 213. 2 Jour. Geology, vol. 10, 1902, p. 815. 3 Am. Jour. Sci., 4th ser., vol. 19, 1905, p. 277. 4 G. Lunge and C. Millberg, Zeitschr. angew. Chemie, 1897, p. 393. 6 Butt. 616, U. S. Geol. Survey, 1916, pp. 486-490. See also, C. K. Leith and W. J. Mead, Metamorphic geology, New York, 1915, pp. 3-24. ROCK DECAY AND WEATHERING 325 as soluble silicates is, then, the dominant factor in the weathering of silicate rocks. Compared with the decrease in silica, the losses of bases in silicate rocks are small. Calcium, magnesium, sodium, and potas- sium are removed, but the loss is ordinarily only partial. The leaching of both potassium and sodium is characteristic and is markedly different from certain processes in the alteration of wall rocks of ore deposits, where sodium is completely removed and potassium enriched. The analyses quoted show that from one-seventh to one-fifth of the iron oxides are carried away. The water invariably increases. RESIDUAL CLAY 1 Occurrence. The residual clays, as might be expected, are found mainly where decay of rocks has proceeded unchecked for a long time and where the products have not been swept away by strong erosion or by glacial action. Such clays are found in all parts of the world; in the United States they occur chiefly in the southern Appalachian region. Acidic granular rocks like granite and gneiss particularly those rich in feldspar and poor in ferromagnesian silicates, like peg- ma'tite dikes yield the best and purest clays, but even these mast often be purified by washing in order to remove residual quartz and limonite. At varying depths, 100 feet at most, these clays gradually change into unaltered rocks. Uses and Properties. The ordinary varieties of residual clays are used for brickmaking, the purer for fire-bricks, the finer grades for pottery; for the last use the deposits of the United 1 The literature of clays is exceedingly voluminous. For information more detailed than can be given here, consult: H. Hies, Clays, occurrence,'properties, and uses, 1908. H. Ries, A review of the theories of origin of white residual kaolins, Trans. Am. Ceramic Soc., vol. 13, 1911. I. E. Sproat, Refining and utilization of Georgia kaolins, Bull. 128, U. S. Bureau of Mines, 1916. H. O. Buckman, The chemical and physical processes involved in the formation of residual clay, Trans. Am. Ceramic Soc., vol. 13, 1911. J. H. Watkins, White-burning clays of the southern Appalachian states, Trans. Am. Inst. Min. Eng., vol. 51, 1916, pp. 481-501. A full bibliography of the older literature by H. Rosier is contained in Neues Jahrb., Beil. Bd. 15, 1902, p. 231. 326 MINERAL DEPOSITS States are insufficient, hence large quantities are imported from England. These purer grades of white burning clays are usually called kaolin, ball clay, paper clay or china clay, and are also used in manufacturing paper and as fillers in paints, putty and crayons. About 200,000 tons of these fine clays are produced in the United States, the average price being $8 per ton. Their composition^before and after washing, is indicated by the follow- ing analyses. ANALYSES OF CRUDE AND WASHED KAOLIN, WEBSTER COUNTY, SOUTH CAROLINA Crude Washed Kaolinite Si0 2 A1 2 3 Fe 2 O 3 62.40 26.51 1.14 45.78 36.46 0.28 46.5 39.5 FeO 1 08 CaO 57 50 MgO 01 04 Alkalies 98 25 H 2 O Moisture 8.80 0.25 13.40 2.05 - 14 Total Clav substance . . . 100.66 66.14 99.84 93.24 100 It is sometimes difficult to determine with the microscope the particles of kaolinite (H4Al 2 Si 2 9 ) in an altered rock, on account of , their minute flaky size and low double refraction. It is probable that kaolinite is present in the residual clays, but besides this well-defined mineral there may be other hydrated silicates of alumina separated in colloidal form, as gels, during the weathering. Among these hardened gels there are some highly hydrated forms like halloysite and allophanite which are de- composed by HC1. Others corresponding approximately to kaolinite are only attacked by H 2 SO 4 . Both classes occur in sediments as well as in residual deposits. Paper clays are mined in Cretaceous beds of South Carolina and Georgia, plastic kaolins or ball clays are obtained in Tertiary beds in 1 H. Ries, Economic geology, 1917, pp. 170-186. ROCK DECAY AND WEATHERING 327 Florida and Western Tennessee. Residual clays are mainly mined in North Carolina. The most important property of clay is plasticity, by means of which it can be kneaded or molded into a desired shape, which it retains when dry. Not all the residual clays are plastic, nor is the pure mineral kaolinite. It is now generally be- lieved that the plasticity of clay depends upon the presence of colloids. 1 The tensile strength of air-dried clays varies from 15 to 400 pounds or more per square inch, according to Ries. The fusi- bility varies according to the impurities present. In low grades of clay incipient fusion may occur at about 1,000 C., while in refractory clays, which are low in fluxing impurities, it may not occur until 1,300 or 1,400 C. is reached. The melting-point of kaolin is about 1,800 C. Origin. The best residual clays are derived largely from the decomposition of the feldspars as indicated on p. 322 by carbon dioxide. The process is hastened by sulphuric acid, as is attested by the great development of pure kaolin in the upper levels of pyritic mineral deposits. The decrease in volume by decomposition of orthoclase, if the silica were liberated in soluble form, would be 54.44 per cent. In kaolinization anorthite simply loses its calcium oxide and takes up C02 and H^O. Pure orthoclase loses 43.24 per cent. Si0 2 and all of its potassium; albite loses 45.87 per cent. SiO 2 and all of its sodium. The origin of kaolin has been the subject of much discussion. About 10 years ago H. Rosier 2 published a long and important paper which gave rise to an animated discussion. 3 Rosier con- cludes that kaolin is not formed by weathering, but only by pneumatolytic or allied processes by the action of thermal waters. It is impossible to accept these results and they have been vigorously contested by Stremme and Barnitzke; 4 the latter showed that the celebrated deposit at Meissen, in Saxony, where a high grade of chinaware is made, is decidedly a product of 1 H. E. Ashley, The colloid matter of clay and its measurement, Bull. 388, U. S. Geol. Survey, 1909, p. 65. *Neues Jahrb., Beil. Bd. 15, 1902, p. 231-393. 3 SeeH. Rosier. Zeitschr. prakt.Geol., 1908, p. 251; Stremme, idem, p. 122. 4 Ueber das Vorkommen der Porcellanerde bei Meissen und Halle. Zeitschr. prakt. Geol., 1909, pp. 457-472. 328 MINERAL DEPOSITS weathering, gradually changing in depth into unaltered porphyry and syenite. The chemistry of kaolin is given in full by H. Stremme, 1 and has lately been summarized by Doelter. 2 It has been shown by W. Lindgren 3 and others that kaolin does not form in deposits that are due to ascending thermal waters, except possibly very close to the surface, where they may mingle with atmospheric waters. The idea that the mineral may form by pneumatolysis, or the action of water or gases liberated at high temperature from igneous magmas, is assuredly untenable; a strongly hydrous mineral, parting with its water at the comparatively low temperatures of 300 to 400 C., could not possibly originate together with such minerals as topaz and tourmaline. The frequent association of kaolin with cassiterite veins for instance, in Cornwall has been held by L. v. Buch, A. Daubre"e, J. H. Collins, 4 and H. Rosier to indicate a derivation by the action of hydrofluoric acid on feldspars, but as the kaolin deposits, during the metallogenetic epoch, were under the same general conditions of pressure, temperature, and depth as the tin deposits, this view must be abandoned. Extensive observations in the United States have shown that in mineral deposits kaolin is scarcely ever a primary mineral, but has been derived largely by the action of sulphuric acid on the feldspar minerals of the rocks and on sericite, which is often abundantly developed in ore deposits formed under widely differ- ing physical conditions. In view of this, it seems odd that Rosier expresses astonishment at finding a large amount of muscovite with the kaolin from Cornwall and suggests that the former may be a secondary product. G. Hickling 5 has investigated the china clays of Cornwall and shows that they form essentially a sheet covering the corroded surface of the granite and that they have resulted from the weathering of sericitic granite, the sericite being due to previous alteration by thermal waters. Kaolin, then, is formed abundantly in the zone of weathering and in smaller .amounts for a considerable distance below this zone. 1 Die Chemie des Kaolins, Fortschritte der Min., Krist. u. Petr. Jena, 1912. 2 C. Doelter, Handbuch der Mineralchemie, vol. 2, 1914, pp. 31-91; 125-137. 3 The origin of kaolin, Econ. Geol, vol. 10, 1915, pp. 89-93. 4 J. H. Collins, Min. Mag., vol. 7, 1886-1887, p. 217. 6 Trans. Inst. Min. Eng. (England), voL 36, 1908-1909, p. 10. ROCK DECAY AND WEATHERING RESIDUAL IRON ORES (LIMONITE AND HEMATITE) 329 Origin. During the processes of weathering only a small part of the iron is carried away in solution; the greater part remains in the rock altered to limonite (2Fe 2 3 .3H 2 O), to gothite (Fe 2 3 . H 2 0), or to indefinite colloidal mixtures of various hydroxides of iron; hematite may also be present. In places basic sulphates or phosphates may remain, as well as somewhat indefinite and unstable ferric silicates. Nontronite, H4Fe2Si2Og, the equivalent of kaolin, is said to be present in weathered rocks. In the zone of weathering the iron shows a strong tendency to move out- ward and segregate in irregular or mammillary masses, separated by clayey material, though much of it, of course, remains inti- 10 FEE* FIG. 104. Section showing oxidation of iron carbonate to limonite in Tertiary beds, Cass County, Texas. After E. F. Burchard, U. S. Geol. Survey. mately mixed with clay. The same is true of manganese, some of which may be associated with the limonite, though when much manganese is present, it also tends to separate by itself. The "centrifugal" tendency of iron hydroxide is well seen in many oxidized mineral deposits, often also in the weathering of pebbles. A fine instance was observed in the cobbles of andesite in the Tertiary river-bed at Iowa Hill, California. The outside of these cobbles is hard and consists of an impure limonite; the center contains soft yellowish kaolin. During the concentration the ferric hydroxides (see p. 257) were probably transported as colloids, which hardened and be- came crystalline, as shown by the radial structure of many con- cretions. The chemical character of these ores has rarely been 330 MINERAL DEPOSITS studied in detail; probably it will be found that barite, oxidized zinc minerals, and compounds containing manganese, nickel, and cobalt are present. Many of the limonites are rather pure and they are of considerable economic importance. Classification. One class of residual brown iron ores is derived from the decomposition of deposits of siderite or pyritic ores, both usually formed by ascending waters, or from the weathering of black bands or glauconite beds (Fig. 104). Such limonites in places reach considerable depths, dependent on the penetrating power of oxygenated waters. The decom- posed croppings of pyritic ores are not often used as iron ores. Another class consists of local segregations of limonite or allied hydroxides in the decayed rock and residual clay near the surface. These masses are particularly common in limestone areas. Little or no siderite is found near the surface, but it may appear in the limestone at greater depth. When oxygen is exhausted the iron is more easily transported as a bicarbonate and the metasomatic replacement of calcite by siderite may then occur. There are, however, few deposits of limonite which change in depth to large irregular replacements of siderite, so that it may be assumed that the rate of solution and downward transportation of the precipitated limonite is slow. Finally, a third class of residual iron ores, consisting of limonites mixed with hematite, occurs as widespread sheets formed by the gradual decay of strongly ferriferous rocks. Brown Hematites of the Appalachian Region. 1 In the United States the residual iron ores are most abundant in the Appala- chian region, mainly in Alabama, Georgia, Virginia, and Ten- nessee. The annual production of such ores is about 2,000,000 long tons, a small part, of course, of the yearly output of iron ores in the United States. These so-called "brown hematites" 1 C. W. Hayes and E. C. Eckel, Iron ores of the Cartersville district, Georgia, Bull. 213, U. S. Geol. Survey, 1902, pp. 233-242. E. C. Eckel, Limonite deposits of eastern New York, etc., Bull. 260, U. S. Geol. Survey, 1904, pp. 335-342. R. J. Holden in Mineral resources of Virginia, r 1908. E. C. Harder, The iron ores of the Appalachian region in Virginia, Bull. 30, U. S. Geol. Survey, 1908, pp. 215-254. E. C. Harder and E. F. Burchard, Mineral Resources, U. S. Geol. Survey, particularly pt. 2, chapter on Iron, 1908. E. F. Burchard and E. C. Eckel, Birmingham district, Bull. 400, U. S. G ol. Survey, 1910, pp. 145-167. ROCK DECAY AND WEATHERING 331 are mined in many small deposits; their content in iron ranges from 38 to 52 per cent, (limonite 59.89 per cent. Fe); most of them are comparatively rich in phosphorus. Most of the southern limonites lie in Cambro-Silurian strata and extend along the "Great Valley," between the pre-Cam- brian on the east and the Paleozoic rocks on the west. They are classed as vaUey ores, mountain ores, and Oriskany ores. The valley ores appear as irregular deposits of shallow pockets in clay derived from the decomposition and solution of Cambro- Silurian limestone or dolomite. The ores lie as scattered lumps in the clay, not so much on the eroded surface of the limestone, but rather higher up (Fig. 105). Each deposit is soon exhausted, FIG. 105. Vertical section showing structure of the valley brown ore deposits of the Rich Hill mine, Virginia. After E. C. Harder, U. S. Geol. Survey. and few extend below a depth of 50 feet. The ores are mixtures of limonite, gothite, and clay; the composition ranges from 40 to 56 per cent. Fe, 5 to 20 per cent. SiO- 2 , 0.05 to 0.5 per cent. P, and 0.3 to 2.0 per cent. Mn. Many of these ores were evidently concentrated under con- ditions different from those of to-day; most of them are probably of Tertiary age as shown particularly in the deposits south of Birmingham, Alabama. It is not unlikely that the same applies to many "mountain ores." The mountain ores, according to Harder, show greater varia- tion in occurrence and appearance. They are found as small discontinuous pockets in residual material above the Lower 332 MINERAL DEPOSITS Cambrian quartzite at or near the contact with the overlying formation, which is generally a limestone. While these ores are mainly superficial, they are sometimes worked to a depth of several hundred feet. The composition ranges from 35 to 50 per cent. Fe, 10 to 30 per cent. SiO 2 , 0.1 to 2 per cent. P, and 0.5 to 10 per cent. Mn. These limonites are often glassy and concretionary. The occurrences are classed by Harder as follows: 1. Pocket deposits in clay, in part replacements of limestone, in part manganiferous (Fig. 106). 2. Small replacement deposits in shale, along fractures. 3. Deposits in quartzite or sandstone, not abundant, including a. Breccia deposits accompanied by replacement. 6. Vein deposits along faults. Brown-clay -with -ore-fragments: T- FIG. 106. -Vertical section showing the structure of mountain brown ore occurring as mammillary masses in clay. Mary Creek mine," Virginia. After E. C. Harder, U. S. Geol. Survey. The sandstones of the Cambro-Silurian are often ferruginous in this region. Some of the varieties rich in hematitic cement change along the strike to beds of siliceous hematite, several feet thick and of possible economic importance. The Oriskany ores 1 are mined in Virginia and form irregular replacements along local folds or fracture zones on the flanks of 1 C. M. Weld, The Oriskany iron ores of Virginia, Earn. Geol, vol. 10, 1915, pp. 399-421. ROCK DECAY AND WEATHERING 333 greater anticlines. They occur in the calcareous Oriskany sandstone which is overlain by the Romney shale (Devonian) and underlain by the Helderberg limestone (Silurian). The ore largely replaces the sandstone, sometimes also the limestone and may be from 10 to 100 feet wide. The greatest depth reached is 600 feet; at this or lesser depth the ore grades into unaltered rock. The iron is considered by some authors to be derived from the Romney shale but is more likely derived from the sandstone itself. The ore is made up of earthy masses and rounded concretions of fibrous limonite filled with clay or sand. The ore from the Callie mine contains about 0.2 per cent, of zinc. 1 At the Callie mine the ore production is about 2,700 tons FIG. 107. Vertical section showing the Oriskany brown ore deposit at the Callie mine, Virginia. After E. C. Harder, U. S. Geol. Survey. per month. Probably hematite or turgite are also present for the ore does not contain enough water for limonite; it averages 43 per cent. Fe, 10 to 25 per cent. Si0 2 , 0.06 to 0.5 per cent. P, and 0.5 to4 per cent, manganese. Cobalt and nickel are reported to be present in traces. The Oriskany ores, like the other " brown hematites," are subjected to a rough concentration in log washers in order to remove the clay. ir The foot- wall limestone is said to contain the same amount of Zn. Letter from S. E. Doak. 334 MINERAL DEPOSITS Iron Ores of Bilbao, Spain. 1 -The great deposits of Bilbao, in northern Spain, have for many years yielded several million tons annually, the ores being exported to England. According to Adams, both replacement and residual ores are present. The ores are superficial and limited to areas of -Cre- taceous limestone, which is 250 feet thick and dips northeast. The white siderite ore, which is found at some depth, is altered near the surface to red hematite with 80 to 90 per cent. Fe 2 O 3 . The ores are of Bessemer grade. Adams believes that, during the progress of denudation, the calcareous beds became replaced by siderite by the aid of downward-percolating solutions, de- rived partly from the overlying calcareous shale. Through long- continued rock decay the siderite was altered to hematite and limonite, which now, with much clay, cover the limestone areas like a sheet. One of the largest iron-bearing areas is 2 miles long and 3,300 feet wide; the iron ore in this area had a thick- ness of about 100 feet. 2 Residual Ores of Cuba. 3 Iron ores have been mined for a number of years in the vicinity of Santiago, Cuba, but these ores, of contact-metamorphic origin, consist of hematite with some magnetite and contain a high percentage of sulphur. The three new districts described by Spencer and others are likewise in the eastern part of the island, but are of an entirely different type. They are the Mayari and Moa districts in Oriente province, and the San Felipe in Camaguey (Fig. 108). The ores occur as 1 F. D. Adams, Notes on the iron deposits of Bilbao, Jour., Canadian Min. Inst., 1901. John, Zeitschr. prakt. Geol., 1911, pp. 208-212. P. Grosch, Geol. Rundschau, vol. 5, 1914-15, pp. 392-400. 2 Stelzner and Bergeat, Die Erzlagerstatten, vol. 2, 1906, p. 1049, with list of literature. 3 A. C. Spencer, Three deposits of iron ore in Cuba, Bull. 340, U. S. Geol. Survey, 1907, pp. 318-329. C. M. Weld, The residual iron ores of Cuba, Trans., Am. Inst. Min. Eng., vol. 40, 1909, pp. 299-312. J. F. Kemp, The iron resources of the world, Int. Geol. Congress, Stockholm, 1910, pp. 793-795; Trans., Am. Inst. Min. Eng., vol. 51, 1916, pp. 3-30. See also seven papers on the same subject by J. S. Cox, Jr., C. K. Leith, W. J. Mead, A. C. Spencer, C. W. Hayes, W. L. Cumings, B. L. Miller, D. E. Woodbridge, and J. E. Little, in Trans., Am. Inst. Min. Eng., vol. 42, 1911, pp. 73-152. Also C. K. Leith and W. J. Mead, idem, vol. 53, 1916, pp. 75-78. ROCK DECAY AND WEATHERING 335 residual mantles resulting from the weathering of serpentine and for the most part lie on plateaus at rather high elevations. They were probably formed during the Tertiary before the uplift of the present plateaus. Near the surface the material is earthy and dark red, sometimes cemented with shot-like lumps of hema- tite scattered over the surface; underneath lie yellowish ores changing rather abruptly into decomposed and soft serpentine. In places a layer of cherty material is found immediately above the ! serpentine. In the Mayari district the average depth of the ore is about 15 feet and it extends over an area of lOjby 4 miles. Hundreds of millions of tons are said to be available, Map of the EASTERX PART OF CUBA Showing Iron Ore Districts FIG. 108. Sketch map of eastern part of Cuba. After W. L. Cumings and B. L. Miller. allowing for parts of the area which are below the workable grade. The ore is removed by drag-line steam shovels. According to analyses the ore is fairly uniform, the metallic iron varying in percentage from 40 to 50. It is remarkably free from phosphorus and evidently contains hematite, limonite, a little magnetite, and also some free aluminum hydroxide. It is, in brief, a typical iron-rich laterite (see p. 351). There is much water; according to Kemp the Moa ores yield 25 to 30 per cent, hygroscopic and 10 to 12 per cent, combined water; silica is low and alumina high. The concentration of nickel and chro- mium is also remarkable; the latter metal is removed during the smelting; the former is favorable to the quality of the iron. 336 MINERAL DEPOSITS ANALYSES OF SERPENTINE AND ORE FROM THE MAYARI DISTRICT, CUBA (After C. K. Leith) 2.90 10.24 72.35 50.56 SiO 2 A1 2 O 3 39.80 1.39 Fe 2 O 10 14 Fe MgO Cr Ni+Co 7.10 33.69 0.20 0.97 p 001 s 06 H.O + .. . .13.31 1.66 0.84 0.016 0.20 10.96 99.561 99.166 1. Serpentine, at depth of 29 feet. 2. Iron ore, at depth of 6 feet. Analyses by Spanish- American Iron Co. The porosity of the ore is exceedingly great amounting to 75 per cent, of its volume but lessens near the surface. In considering the alteration of serpentine to ore in terms of weight it is found that the alumina has remained nearly constant. The changes in the composition of the serpentine during its alteration to ore is shown by Leith and Mead in Fig. 109, which is based on many analyses at uniform intervals. The diagram illustrates the rapid destruction of the serpentine by leaching of Si0 2 and MgO, the marked relative increase of iron and alumina and a gradual loss of nickel. Toward the surface hematite (with a little magnetite) develops from limonite and bauxite from kaolin. In the middle part of the ore body iron has increased in proportion to the alumina, owing probably to re-deposition and oxidation of ferrous iron dissolved by the reducing action of the vegetation. Silica is lost throughout and magnesia is wholly removed. In 100 pounds of typical serpentine there are 1.5 pounds of alumina and 10 pounds of ferrous oxide. When the magnesia and silica are removed in solution and the iron oxidized there remain approximately 11.75 pounds of limonite, 3.8 pounds of bauxite and kaolin, and, at the most, 2 pounds of minor constitu- ents. This residual of 17.55 pounds contains 7.8 pounds, or 44.4 per cent., of metallic iron and is an iron ore. Distribution and Stability of Residual Iron Ore. The residual iron ores are widely distributed in countries of warm climate, ROCK DECAY AND WEATHERING 337 where secular decay has progressed without interruption for a long time. It seems, however, that great concentration has been effected only from relatively soluble rocks like limestone and serpentine. Many of the laterites of India, Africa, and other tropical countries are rich in ferric oxide and have the same characteristic concretionary pellets and shots on the surface. Extensive limonite deposits similar to those of Cuba have lately been discovered in Borneo and on Mindanao, in the Philippines. Vegetation plays an important part in the origin of many of these deposits. Underneath the mat of roots and decayed vegetation the soil in tropical countries is often white.or yellowish, 338 MINERAL DEPOSITS indicating that the iron is in the ferrous state, probably as car- bonate. When, as happened on the high volcanic plateau of Molokai, Hawaiian Islands, 1 the vegetation is destroyed the soil immediately turns red and hard and shows characteristic pellets of ferric oxide. In part at least the rock is thus changed directly to hematite without passing through the intermediate stage of limonite. According to H. Wolbling, 2 the natural ferric hydroxides have great stability and cannot readily be changed to ferric oxide, probably not by exposure to air and salt solutions. The freshly precipitated hydroxides are, however, easily converted to ferric oxide and these colloids may easily be crystallized. 3 His experi- ments show that by the precipitation of ferric solutions with calcite or siderite at 100 C., Fe 2 O 3 is easily formed, containing only 1 or 2 per cent. H 2 0, while during slow and wet oxidation of ferrosalts, ferric hydrates of iron are obtained. Wolbling also asserts that there are yellow forms of FegOs, as well as red forms of the hydroxides. It is certain, at any rate, that the ferric oxide, as well as the hydrates, is very stable when once formed and is not easily altered. No one can fail to be impressed by certain similarities of the Cuban residual ores to those of the Mesabi range (p. 366). Similar large expanses of rock, weathered under a tropical sun and covered by residual ferric oxide, undoubtedly yielded the material for the sedimentary hematite deposits. RESIDUAL MANGANESE ORES* The minerals of the residual manganese ores consist of pyrolu- site (MnC>2, 63.2 per cent. Mn), psilomelane (MnO2, with H 2 O, K 2 0, and BaO; 49 to 62 per cent. Mn), wad (perhaps MnO 2 . *W. Lindgren, The water resources of Molokai, Water-Supply Paper, 77, U. S. Geol. Survey, 1903, p. 19. 2 H. Wolbling, Bildung der oxydischen Eisenerzlager, Stahl und Eisen, 1909, p. 1248; also Zeiischr. prakt. Geol, vol. 17, 1909, p. 495. 3 C. Doelter, Ueber Umwandlung amorpher Mineralkorper in Krys- talline, Tsch. M. und p. Mitt., vol. 28, 1909, pp. 556-559. *R. A. F. Penrose, Jr., Manganese, its uses, ores, and deposits, Ann. Rept. Arkansas Geol. Survey, vol. 1, 1890. T. L. Watson, Trans., Am. Inst. Min. Eng., vol. 34, 1904, p. 207. T. L. Watson, Preliminary report on the manganese deposits of^Geor- gia, Bull. 14, Georgia Geol. Survey, 1908. ROCK DECAY AND WEATHERING 339 nMnO+H 2 0, varying percentage of metal), more rarely braunite (3Mn 2 O 3 .MnSi03 (?) ; 69.7 per cent. Mn), and manganite (Mn 2 O 3 . H 2 O, 62.4 per cent. Mn). The most common ores are pyrolusite and psilomelane, both occurring frequently in botryoidal, renif orm, or mammillary con- cretions. Harder has shown that these two minerals may form alternating layers in the concretions. Earthy or rough, slaggy forms are also common. Like limonite they are largely colloid deposits, later converted into crystalline minerals. Primary Sources. Nearly all workable manganese deposits are of secondary formation that is, they are concentrated from manganese minerals more sparsely distributed in rocks. Pyrolu- site, psilomelane, and wad are always secondary, formed under the influence of weathering, even where they descend to con- siderable depths below the water level. In igneous rocks manganese is always present but only in small amounts. The largest percentages (about 0.36 percent.) are found in syenite and its porphyries and in basalts. Sedimentary rocks may contain manganese in the form of oxide and carbonate. Manganese nodules occur in some deep- sea deposits. Analyses of limestones often show a small amount of manga- nese. In many cherts and jaspers of the sedimentary series manganese is characteristically present as rhodonite or rhodo- chrosite. On previous pages it has been shown that important deposits of manganese may be produced by sedimentation. In crystalline schists, especially in those of more basic composi- tion, small quantities of manganese are found. In some crystalline schists spessartite (manganese garnet), rhodonite, and piedmonite (manganese epidote) appear in considerable quantities. Finally, rhodochrosite and rhodonite are rather common in ore deposits of hydrothermal or contact-metamorphic origin, and much manganese is present in some metamorphic specularite and magnetite deposits. E. C. Harder, Manganese deposits of the United States, Bull 427, U. S. Geol. Survey, 1910 (with bibliography and notes on foreign occurrences). E. C. Harder and D. F. Hewett, Mineral Resources, U. S. Geol. Survey, annual publication. D. F. Hewett, Some manganese ore in Virginia and Tennessee, Bull. 6iO, U. S. Geol. Survey, 1916, pp. 37-71. 340 MINERAL DEPOSITS Manganese Deposits in the United States. From the rocks above mentioned manganese may be concentrated by processes of weathering, and its ores are found in concretions embedded in residual clay or ocher and accompanied more or less closely by limonites. During this process some other metals, notably nickel, cobalt, zinc, and barium, have a tendency to accompany the pyrolusite and psilomelane. In general such deposits are superficial or of slight depth and closely parallel the residual limonites already described. In California small deposits of secondary manganese ores occur FIG. 110. Generalized section, showing the occurrence of manganese ore at'Batesville, Arkansas, a, Boone chert (Mississippian) ; b, Cason shale with manganese_deposits (Ordovician) ; c, Polk Bayou limestone (Ordovician) ; d, surface clay with manganese deposits. After E. C. Harder, U. S. Geol. Survey. in areas of the radiolarian cherts or jaspers of the Franciscan formation (Jurassic ?) In Arkansas residual ores have been mined at Bates ville, 1 where they occur both in the Cason manganiferous shale, of upper Ordovician age, and in clay derived from this formation (Fig. 110). Penrose believed that the manganese was derived from the pre-Cambrian area in southeast Missouri and de- 1 R. A. F. Penrose, Jr., op. tit. ROCK DECAY AND WEATHERING 341 posited in the sedimentary formation, but the later work of Ulrich and others has shown that erosional epochs have inter- vened within the formation period assumed by Penrose and that the ores are original marine deposits, reconcentrated dur- ing two subsequent land stages, first during the late Silurian and Devonian partial emergence, and second during the post- Paleozoic erosion of the Boone chert. 1 In the Appalachian region small deposits occur in granites and schists of the Piedmont region, but chiefly in the Paleozoic sedi- /- Manganese ore kidneys Ir i White sandy mass FIG. 111. Sketch showing distribution of manganese ore lumps in clay at the Crimora mine, Virginia. After E. C. Harder, U. S. Geol. Survey. ments of the Cambro-Silurian belt that is, in the general area of the residual iron ores. At the Crimora deposit, in "Virginia (Fig. Ill), the ore is found as "masses of various sizes scattered through variegated clays in an elliptical basin in a canoe-shaped syncline of the Cambrian quartzite," into which the manganese penetrates as dendritic forms and crystalline coatings. 2 1 E. C. Harder, Bull. 427, U. S. Geol. Survey, 1910, p. 117. * E. C. Harder, idem, p. 60. 342 MINERAL DEPOSITS ANALYSIS OF BEST QUALITY CRIMORA ORE [T. L. Watson, Mineral resources of Virginia, p. 248] MnO 2 81.703 BaO 0.829 MnO 7.281 CaO 0.880 Fe 2 O 3 0.533 MgO 0.630 CoO 0.354 P 2 O 5 0.171 NiO 0.096 (NaK) 2 0.467 ZnO 0.623 H.,0 1 3.405 A1 2 O 3 0.896 SiO 2 2.132 Total 100.000 Mn 57.297 The manganese deposits of the Appalachian region occur in a decomposed surface zone of many different rocks (Figs. 112 and 113), but most of the deposits are, according to Harder, asso- ciated with the top stratum of an impervious. Cambrian quartz- ite overlain by limestone. Penrose 2 holds that they were laid down in local basins during the deposition of the rocks in whose residual clays they are now found. Harder 3 believes that the metal was in the first place obtained from the crystal- line rocks of the Piedmont region and that since the emergence of the sediments repeated concentration by rock decay has been going on. In central Texas, 4 in Mason, Llano, and San Saba counties, oxidized manganese ores occur as products of weathering of crystalline schists containing spessartite, piedmontite, and tephroite. As stated above, many ore deposits contain manganese as carbonate and silicate, and in the oxidized zone the metal is often highly concentrated in- the form of psilomelane, etc., mixed with limonite; these ores often contain gold and silver, but rarely much copper, lead, or zinc. Considerable quantities of such ores, used in part as flux for lead smelting and in part, if of high grade, for the manufacture of spiegeleisen, are mined at Lead- ville, Colorado. 5 Here the oxidized ore is apparently derived from a manganiferous siderite. 1 Probably by difference. 2 R. A. F. Penrose, Jr., op. tit, 3 E. C. Harder, op. tit., pp. 99-101. 4 R. A. F. Penrose, Jr., op. tit., p. 432; Sidney Paige, Bull. 450, U. S. Geol. Survey, 1911. 6 S. F. Emmons and J. D. Irving, Bull. 320, U. S. Geol. Survey, 1907, p. 26. ROCK DECAY AND WEATHERING 343 The largest part of the manganese obtained in the United States is derived from ores of the Lake Superior region, where manganese occurs as oxides associated with specularite, and from the zinc residues of the great zinc deposit of Franklin Furnace, FIG. 112. Sketch showing occurrence of manganese breccia ore at Reynolds Mountain, Virginia. After E. C. Harder, U. S. Geol. Survey. FIG. 113. Sketch showing development of breccia ore by replacement. White areas, chert or sandstone; black, manganese ore. One-fifth natural size. After T. L. Watson. New Jersey, where the manganese is contained in the franklinite [(Fe,Zn,Mn)O.(Fe,Mn) 2 O 3 ] associated with zincite [(Zn,Mn)O] in a deposit of deep-seated, probably contact-metamorphic, origin. 344 MINERAL DEPOSITS Brazil. The high-grade manganese deposits of Minas Gerses, Brazil, have been described by J. C. Branner and 0. A. Derby. 1 In the main they appear to be residual ores derived from the weathering of lenses in the crystalline schists containing rhodo- chrosite, tephroite, and spessartite. The ores are concretions, masses, and vein-like deposits of psilomelane in the soft decom- posed rock. India. Manganese ores are extensively distributed in India and their occurrence and origin have recently been described in a detailed manner by L. L. Fermor. 2 To a large extent these rich ores are formed by the combined replacement and decom- position of Archean rocks containing manganese silicates. In part the rocks are crystalline schists with spessartite and rhodo- nite, in part probably non-metamorphosed peculiar igneous rocks, one of which, for instance, consists of spessartite (spandite) and orthoclase with 3.70 per cent, apatite. To a smaller extent the ores are contained in jaspery quartzites and also in laterite, which is purely residual. Many deposits of the first class contain enormous masses of psilomelane, pyrolusite, and braunite; during the process of weathering almost all the silica and alumina have been removed. Fermor finds no evidence that the alteration has been caused by sulphuric acid, but holds that in some manner, not yet fully understood, it has been effected by surface waters. Many of the deposits extend to depths far below the water level and Fermor believes that the oxidation may be of very ancient date, perhaps Archean. In some ways these con- centrations by surface waters recall the Lake Superior iron deposits. Origin. The manganese ores here described as products of weathering and rock decay are in the main similar in origin to the corresponding deposits of iron ore. It is explained on page 272 that iron and manganese, although acting in a similar manner, are usually laid down separately in residual and sedimen- tary deposits because of the greater solubility of the manganese carbonate. Where sulphates are present the ferrous salt is de- composed easily by oxygen, while manganese sulphate requires 1 Literature summarized by E. C. Harder, Bull. 427, U. S. Geol. Survey, 1910, p. 183. 2 L. L. Fermor, The manganese ore deposits of India, Mem., Geol. Survey India, vol. 37, 1909. ROCK DECAY AND WEATHERING 345 the presence of calcium carbonate or some such mineral. 1 On the whole manganese is not transported far from its original source and is characterized by a strong tendency to segregation into local concretions and masses. It is believed that in the main the ordinary surface waters effected the concentration and that the metal has been transformed through the intermediate stage of carbonate. Production and Uses. The normal domestic output of man- ganese ores containing above 35 per cent. Mn was small. Forced production under war conditions has increased the output to about 300,000 tons (in 1918) which is one-third of the amount normally needed. Heavy imports come from Brazil and India. For the manufacture of spiegeleisen, an alloy with iron con- taining less than 20 per cent. Mn, low grades of manganiferous iron ore may be used, but for other purposes the ores should contain at least 40 per cent. Mn, less than 12 per cent. SiOz, and less than 0.3 per cent, phosphorus. The higher grades of manganese ores are used extensively for the manufacture of ferromanganese alloys, which are employed for many purposes in the iron-smelting industry, particularly for hardening steel. 2 The pure manganese dioxide ores also find an extensive chemical use, for the generation of chlorine and for the manufacture of cells for dry electric batteries. RESIDUAL BARITE Barite as residual material and nodular concretions is not uncommon in the residual soils of Virginia and Georgia and in Washington County, Missouri. In Virginia the Cambro-Sil- urian limestone, according to T. L. Watson, generally contains a notable percentage of barium, and in many places in Georgia the Weisner sandstone, of the .same age, also carries barium suphate. In Missouri the barite is concentrated in the soil from veins in the Ordovician Gasconade limestone. The barium may have been transported as the carbonate, which is slightly more soluble than the sulphate, and precipitated by water carrying sulphate. Much of the barite produced in the United States is obtained from residual clays (p. 376). 1 F. P. Dunnington, Am. Jour. Sci., 3d ser., vol. 36, 1888, p. 177. 2 Mineral Resources, U. S. Geol. Survey, pt. 1, 1908, p. 138, and in later 346 .MINERAL DEPOSITS RESIDUAL ZINC ORE In the Appalachian region, in western Virginia and eastern Tennessee, the Cambro-Silurian limestones contain in places sulphides of lead and zinc distributed in brecciated and crushed zones. At such localities the deep residual soil often contains calamine and smithsonite, the hydrated silicate and the carbon- ate of zinc, with some cerussite and galena. These ores occur next to the limestone at the bottom of the clay (Fig. 114), not scattered through it like limonite and pyrolusite. 1 FIG. 114. Section in open cut at the Bertha zinc mines, Virginia, show- ing'relations of the residual ore to the limestone chimneys and the residual clay. After T. L. Watson. RESIDUAL OCHERS 2 The residual ochers are impure deep-red, yellow, or brown pulverulent materials consisting usually of predominant limonite and hematite with more or less clay and are generally used for pigments. They are no doubt colloid precipitations. The terms Indian red, sienna, and umber, the latter two for the darker yellowish-brown and brown shades, are in use. Not all mineral pigments are natural products, for roasted pyrite, siderite, slates, 1 W. H. Case, The Bertha zinc mines at Bertha, Virginia, Trans., Am. Inst. Min. Eng., vol. 22, 1894, pp. 511-536. T. L. Watson, Lead and zinc deposits of Virginia, Bull. 1, Virginia Geol. Survey, 1905. T. L. Watson, Mineral resources of Virginia, 1907. T. L. Watson, Lead and zinc deposits of the Virginia-Tennessee region, Trans., Am. Inst. Min. Eng., vol. 36, 1906, pp. 681-727. 2 E. F. Burchard and J. M. Hill, Mineral Resources, U. S. Geol. Survey, annual publication, "Mineral Paints." G. P. Merrill, Non-metallic minerals, 1910, pp. 104-111. ROCK DECAY AND WEATHERING 347 and shales are also used. 1 The southern Clinton iron ores are also employed for these purposes. The residual iron ore deposits of the Southern States contain material which may be classed and is used as ocher. Especially interesting are the Cartersville deposits, 2 in Georgia. These ochers occur only in the Weisner (Cambro-Silurian) quartzite, in the lower part of the residual zone immediately above the yet solid rock, and also in shattered zones in the quartzite itself. The quartzite contains about 90 per cent. SiOa, 1 .5 per cent. FeS 2 , 0.5 per cent. Fe 2 3 , and also an unusual percentage of barium sulphate (4.46 per cent, in the analysis given by Watson). The calculated constituents of the ocher are 66 per cent, limonite, 25 per cent, clay, and 9 per cent, quartz; a little hematite is probably also present. Hayes and Watson are in agreement regarding the origin of the ocher, considering it as resulting from a metasomatic replace- ment of the cement and the quartz grains of the quartzite by limonite. The process begins by the permeation of the grains by dendritic limonite. This direct formation of the ocher is scarcely probable, but more likely it has progressed by means of an intermediate stage of siderite. The replacement of quartz by iron carbonate is a well-known phenomenon, illustrated, for instance, in the Coeur d'Alene lead deposits of Idaho. The annual domestic production of natural pigments amounts to about 57,000 tons. The mining is done mainly in open pits, and the material is crushed, washed in a log-washer, and allowed to settle in tanks. RESIDUAL PHOSPHATES As described more fully on page 275, many sedimentary beds contain much phosphate of calcium, often in oolitic or concretionary form. When these beds are exposed to surface waters an enrichment usually takes place by solution of calcium carbonate, provided the beds are permeable to the circulating 1 B. L. Miller, The mineral pigments of Pennsylvania, Rept. No. 4, Topo- graphic and Geologic Survey Commission of Pennsylvania, Harrisburg, 1911. F. T. Agthe and J. L. Dynan, Paint-ore deposits near Lehigh Gap, Penn- sylvania, Bull. 430, U. S. Geol. Survey, 1909, pp. 440-454. 2 C. W. Hayes, Iron ores in the Cartersville district, Georgia, Trans., Am. Inst. Min. Eng., vol. 30, 1901, pp. 403-419. T. L. Watson, The ocher deposits of Georgia, Bull. 13, Georgia Geol. Survey, 1906. 348 MINERAL DEPOSITS waters. Many important phosphate deposits for instance, those of Florida, South Carolina, and Tennessee have been thus enriched. DEPOSITS OF HYDRATED SILICATES OF NICKEL The original home of nickel, cobalt, and chromium is in the peridotitic and pyroxenic rocks and in the serpentines derived from them, although traces of these metals are also frequently noted in analyses of other basic rocks. The primary condition of the nickel in the rocks is not always known; probably it occurs both as silicate and as sulphide, the latter in microscopic grains, the former as an admixture in iron-magnesium silicates. From the serpentines and peridotites the nickel is sometimes concentrated in commercially important quantities by processes of weathering and the ores thus formed are always the green hydrated silicates of nickel. Chromite, which always occurs in these basic rocks, does not readily yield oxidized minerals in the zone of weathering. Sulphates of chromium have been observed in a quicksilver mine in California, but no silicate analogous to garnierite exists. Nickel silicates are diverse and uncertain in composition. The most important are genthite, H 4 Ni2Mg2(Si04)3.4H 2 O; connarite, H 4 Ni 2 Si 3 Oio; and garnierite, (Mg,Ni)Si0 3 +nH 2 0. According to an analysis by A. Liversidge garnierite contains 38.35 per cent. SiO 2 ; 32.52 per cent. NiO; 10.61 per cent. MgO; 0.55 per cent. A1 2 3 and Fe 2 O 3 ; 11.53 per cent. H 2 O (at red heat) and 6.44 per cent. H 2 (at 100 C.). Such deposits are superficial and the oxidizing surface waters have been the carrying and concentrating agency. The ores rarely extend far below the water-level and in some cases are contained in the residual clays of the completely weathered rock. These nickel ores are often accompanied by cobalt in the form of separate masses of asbolite, a rather indefinite mixture of hydrous oxides of manganese and cobalt. These deposits do not contain sulphides, and copper is rarely present. The accompanying minerals are quartz, chalcedony, opal, and various obscure hydrous magnesium silicates, some- times also a little magnesite. Nickel ores of this kind are not uncommon, but have attained commercial importance only in New Caledonia. The nickel mine at Riddles, in southern Oregon, has been ROCK DECAY AND WEATHERING 349 described by several authors. 1 The parent rock is a peridotite containing 0.10 per cent, of NiO. The olivine separated from the rock contained 0.26 per cent, of NiO and all observers agree that the nickel ores are formed from this silicate. In the finest joints of the rock silica and nickel-magnesium silicates are deposited, and between them lies the oxidized rock converted to a limonite with some clay and chromite. One of the two most important nickel-bearing districts of the world is in New Caledonia. 2 The island is about 250 miles long and 30 miles wide; one-third of the area is underlain by post- Cretaceous serpentine and peridotite. The lower slopes are covered by a deep mantle of decayed rock ("variegated clay") which really is an iron ore containing, in per cent., 18 silica, 69 ferric oxide, 0.45 alumina, 1 .64 nickel oxide and 10 water. The garnierite deposits are found at elevations of from 400 to 2,500 feet, sometimes on fairly steep slopes, or in the saddles of ridges and spurs. Underneath the "variegated clay" at depths of from 20 to 75 feet the nickel ores occur often descending into the serpentine along fissures and accompanied by chalcedony and opal. There are many small deposits; the largest contained only 600,000 tons. The ores are worked by open cuts and care- fully graded and sorted. Glasser classifies the deposits in vein- like, brecciated, impregnations and earthy masses. In the latter there is much dark brown "chocolate ore" in which the green silicate is not visible. The clayey ore averages, in per cent., 23 water, 5-7 nickel oxide, 10-12 ferric oxide, 25 magnesia, 40 silica, no lime, 1.1 chromic oxide, 0.12 cobalt, and 1.5 alumina. Most of the ore is exported. Some of it is sun dried and bri- quetted for local smelting to 45 per cent, nickel matte with limestone and gypsum flux. The New Caledonia deposits were discovered by the geologist Gamier in 1864; the mines were opened 10 years later, and the cheaply mined rich ores made all nickel deposits elsewhere un- profitable. In 1906 the maximum output of 144,000 metric tons 1 Diller and Clarke, Bull. 60, U. S. Geol. Survey, 1890, p. 21. G. F. Kay, Bull. 315, U. S. Geol. Survey, 1907, p. 120. 2 E. Glasser, Rapport sur les richesses mine'rales de la Nouvelle Cal6donie, Ann. des Mines (10), vol. 5, 1904, pp. 29-154, 503-701. G. M. Colvocoresses, Eng. and Min. Jour., Sept. 21 and 28, 1907. | W. G. Miller, Nickel Deposits of the World, reprinted from Report of Royal Ontario Nickel Commission, Toronto, 1917, pp. 234-264. 350 MINERAL DEPOSITS was reached. Lately, owing to the active competition of the Sudbury mines (p. 814), the output has been materially reduced. In 1916, 30,100 tons of ore as well as 5,000 tons of nickel matte were exported to England and France. A small quantity of cobalt ore, a black, earthy asbolite, was exported for a number of years but at present can not compete with the ore from Cobalt, Ontario. The island also produces much chromite (p. 794). BAUXITE i Introduction. Clay, as more or less impure kaolin, is the most abundant product of rock decay, but although it carries 39.8 per cent, alumina its use as a source of metallic aluminum has not been found possible. Corundum is not abundant enough to be used for this purpose. Cryolite (Na 3 AlFl 6 ), a mineral obtained from pegmatitic masses occurring in Greenland, was formerly an important aluminum ore and is still used, in smaller quantities, in the electrolytic processes for the extraction of aluminum. In certain places the weathered zone, however, contains the hydroxides of aluminum and of these bauxite is the most im- portant aluminum ore. There are three aluminum hydroxides: Diaspore, A1 2 3 .H 2 O, with 85 per cent. AhOs; gibbsite or hydrargillite, A1 2 3 .3H 2 O, with 65.4 per cent. A1 2 O 3 ; and bauxite, A1 2 O 3 .2H 2 O, with 74 per cent. A1 2 O 3 . The independence of bauxite as a mineral species is, however, questioned and many authors consider it a hardened and in part crystallized hydrogel of indefinite composition. The Georgia bauxite, according to T. L. Watson, corresponds well to gibbsite. F. Laur states that the French bauxites also agrees with the formula given above for this mineral. Bauxite forms compact, earthy, also very commonly pisolitic masses, the individual concretions often having a diameter of several centimeters. It is gray, cream-colored, yellowish or brown and 1 C. W. Hayes, Bauxite, its occurrence, geology, etc., Sixteenth Ann. Rept., U. S. Geol. Survey, pt. 3, 1895, pp. 547-597. T. L. Watson, Bull. 11, Georgia Geol. Survey, 1904 (Bibliography). F. W. Clarke, Geochemistry, Bull. 616, U. S. Geol. Survey. 1916, pp. 493-501. W. C. Phalen, Mineral Resources, U. S. Geol. Survey, Annual pub- lication. J. W. Richards, Mineral Industry, Annual publication. ROCK DECAY AND WEATHERING 351 is usually admixed with silica and ferric oxide. Its occurrence and structure lend probability to the view that it has originated as a colloid precipitate. The bauxites contain in places crys- talline gibbsite as crusts or veinlets; diaspore has been identified more rarely and quite naturally as it usually formed at higher temperature than that prevailing in residual deposits. The bauxites always contain titanium, averaging as much as 4 per cent. Ti0 2 , and some vanadium but in this they merely share the peculiarities of residual and sedimentary clays. Some investigators state that bauxite contains residual rutile while others have failed to find any titanium mineral. Most probably the titanic dioxide is present in colloid state. Little or no hydroxide of aluminum forms in ordinary rock weathering. Cameron and Bell 1 state that during an examina- tion of several thousand soils from all parts of the United States, hydroxide of aluminum was observed in only one sample, which came from southern California. Bauxite, it may be concluded, is thus rarely formed in the temperate region. In tropical countries, on the other hand, the deep residual soil very often contains aluminum hydroxide. This has Jaeen called laterite (later, brick) and is variously defined. We may say that true laterite is essentially a mixture of the hydroxides of iron and aluminum with more or less free silica, but there are all gradations toward an ordinary ferruginous clay. The laterite may be derived from any igneous or sedimentary rock but ser- pentine and limestone are specially favorable. The iron ore from Mayari, Cuba (p. 334) is a laterite exceptionally rich in iron. Many so-called laterites are not true residual but transported deposits. Laterites may or may not contain bauxite of economic value; they have been described from many lands and the literature is very extensive. 2 *Bvll. 30, Bureau of Soils, U. S. Dept. of Agr., 1905, p. 28. 2 A. Streng, Zeitschr. deutsch. Geol. Gesell., vol. 39, 1887, p. 621. (Germany). A. Bauer, Neues Jabrb., Festband, 1907, p. 33 and 1898, pt. 2, p. 192. (Seychelle Islands). R. D. Oldham and T. H. Holland, Records, Geol. Survey India, vol. 32, pt. 2, 1905, pp. 175-184. L. L. Fermor, The manganese deposits of India, Mem. Geol. Survey India, vol. 37, 1909, pp. 370-380. 'G. C. DuBois, Min. pet. Mitt., vol. 22, 1903, p. 1. (Surinam). A. Lacroix, Nouv. Arch. Mus. Hist. Nat. (Peris), 5th ser., vol. 15, 1913 352 MINERAL DEPOSITS In apparent contradiction to this many of the worked bauxite deposits are found in temperate regions such as Georgia, Arkansas, France, Hungary, etc., but this is explained by the fact that these are not being formed at the present time but are of Tertiary age when a climate like that of Cuba prevailed in large parts of the temperate zone. Origin. The desilication of clay in low latitudes has been discussed extensively. The action of nitric acid, supposedly derived from rain during tropical thunderstorms, has been sug- gested as the cause. T. H. Holland 1 has mentioned the pos- sibility of bacterial action. Clay is decomposed by sulphuric acid and by sodium hydroxide or sodium carbonate and at some places aluminum hydroxide may have originated in this way. W. Maxwell 2 has demonstrated this origin for some of the soils of Hawaiian volcanoes and it applies also to a deposit of alum and bauxite on the upper Gila River 3 in New Mexico. Nevertheless it is clear that sul- phuric acid does not always produce this effect, for diaspore and hydrargillite occur rarely (Rosita Hills, Colorado; Gold- field, Nevada) in the oxidized portions of mineral deposits where the sericitic rocks are acted upon by sulphuric acid solutions. Bauxite also has rarely been observed. In the oxidized zone the sulphuric acid transforms sericite into kaolin, which is frequently accompanied by more or less alunite (K 2 0.3A1 2 O3.4S03+6H 2 0). These suggestions do not suffice to explain the formation of the lateritic aluminum hydroxides. It is now generally con- ceded that this is caused simply by the long continued action of ordinary groundwaters under special conditions of moisture and heat. W. J. Mead 4 has shown that there is a complete grada- tion in case of the Arkansas deposits from the original syenite p. 255, reviewed by L. L. Fermor, Geol. Mag., 1915, pp. 28, 77, 123. (French Guinea). J. M. VanBemmelen, Zeitschr. Anorg. Chemie, vol. 66, 1910, p. 322 (General review). J. Morrow Campbell, Laterite, its origin, structure, etc., Mining Maga- zine (London) Aug.-Nov., 1917. (Tropical Africa.) 1 T. H. Holland, Geol Mag., 1903, p. 59. 2 W. Maxwell, Lavas and soils of the Hawaiian Islands, 1898. ' C. W. Hayes, Bull. 315. U. S. Geol. Survey, 1906, pp. 215-223. 4 Econ. Geol., vol. 10, 1915, pp. 28-54. See also Leith and Mead, Metamorphic geology, New York, 1915. pp. 25-38. ROCK DECAY AND WEATHERING 353 (Fig. 115) that the pisolitic structure develops in place and that residual syenite boulders are surrounded by bauxitic material. The texture of the syenite is sometimes visible in the pisolitic bauxite. There is some evidence of downward leaching of the bauxite, for the top layer is usually more siliceous than the lower parts of the deposit. J. Morrow Campbell believes that bauxite only forms in the zone of percolation close to the fluctuating water level and that it never occurs far below the water. .SiOj H,O FIG. 115. Triangular diagram showing the gradation from syenite to bauxite in terms of the principal chemical constituents. Each triangle represents an analyzed sample. After W. J. Mead.] Some bauxites occurring as veinlike masses in limestone of France 1 and Hungary 2 have been explained as the result of the action of ascending waters carrying aluminum sulphate on limestone. Such a mode of origin is admittedly possible. The sedimentary bauxites of which numerous examples may be found in Georgia and Arkansas in the Cretaceous and Tertiary 1 F. Laur, Trans., Am. Inst. Min. Eng., vol. 24, 1894, p. 234. 2 O. Pauls, Zeiischr. prakt. Geol, vol. 21, 1913, pp. 521-572. 354 MINERAL DEPOSITS beds are probably deposits swept out into the sea by the normal processes of erosion from bauxite rich laterites on the shore. Occurrences. The bauxite deposits of commercial importance are of several different types. In the United States they are confined to Arkansas and the southern Appalachian States. In Arkansas 1 the mineral occurs in Pulaski and Saline coun- ties as superficial beds over areas of various sizes up to 20 acres. The deposits are only exceptionally more than 10 feet in depth. They rest on nepheline syenite or on a kaolinized form of that rock; the lower part retains traces of granitic structure, while the upper part is distinctly pisolitic. Tertiary sands and clays in places cover the nepheline syenite and the bauxite. Other deposits of importance, described by Hayes 2 and also by Watson, 3 are found at a number of places in Georgia and Alabama. The principal occurrences are scattered between Jacksonville, Alabama and Cartersville, Georgia, along a belt about 60 miles in length, one of the typical localities being at Rock Run. The bauxite occurs as pockets and irregular masses or curved strata of various colors, with clay and limonite, in the heavy mantle of residual clay overlying the Knox (Cambrian) dolomite, but sharply separated from it. The ore is in part pisolitic and is mined in open cuts, at some places to a depth of 50 feet or more. .The bottom of the clay masses is rarely exposed; before it is reached the pockets of bauxite generally terminate in tapering points. Occasionally associated minerals are gibbsite (A1 2 3 .3H 2 O) and halloysite, which is similar to kaolin in composition but has more water. A suggestive fact is the occurrence of the deposits at or about the 900-foot contour, which coincides with the elevation of a probable Eocene peneplain. The ores were thus accumulated under topographic and climatic conditions different from those which prevail to-day. Deposits differing considerably from those already described have recently been found in Randolph and Wilkinson County, 1 C. W. Hayes, Twenty-first Ann. Rept., U. S. Geol. Survey, pt. 3, 1901, pp. 435-472. J. C. Branner, Jour. Geol., vol. 5, 1907, pp. 263-289. 2 C. W. Hayes, The geological relations of the southern Appalachian bauxite deposits, Trans., Am. Inst. Min. Eng., vol. 24, 1895, pp. 243-254. W. J. Mead, op. cit. 8 T. L. Watson, op. cit. ROCK DECAY AND WEATHERING 355 Georgia. l They occur near the contact of the flat-lying sands and clays of the Tuscaloosa (Lower Cretaceous) and Claiborne (Tertiary) formations. The ore occurs either as beds resting directly upon Cretaceous clay or disseminated as nodules through it. A perfect series of transition to clay exists, as shown by analyses. Bauxite beds 10 feet in thickness have been observed; the mineral is clayey, dense, or pisolitic. Uses and Production. The annual production of bauxite in the United States has been increasing rapidly and in 1917 was 569,000 long tons, most of which was mined in Arkansas. The mines in France, also yield an increasing amount, about 300,000 long tons in 1913. New deposits are being opened in the British and 'Dutch Guianas. The bauxite ores contain 35 to 57 per cent. A1 2 O 3 , a greatly varying percentage of Fe 2 3 , and up to 30 per cent. Si02. Ores with more than 4 per cent. Fe2Oa are not utilized at present. They are mined in open cuts, often necessitating the removal of heavy overburden, washed to re- move the clay, and dried. For purposes of aluminum smelting the ores must be of high grade and low in silica. About 50,000 short tons of aluminum are now produced annually in the United States; exact data are not obtainable. The uses of the metal and its alloys are steadily increasing. Large works for the electric smelting of aluminum are located at Niagara Falls and in Tennessee. Artificial corundum (alun- dum) is made from the ore by the electric furnace. Bricks of bauxite for basic non-corrosive lining of furnaces are widely used. 2 Aluminum salts, especially alum, are also manufactured from bauxite. An addition of bauxite promotes the rapid setting of cements. ANALYSES OF BAUXITE Si0 2 TiO 2 A1 2 3 Fe 2 0, H 2 Analyst Baux, France (pisolitic). 4.8 3.2 55.4 24.8 10.8 Deville. Jacksonville, Ala 21.08 2.52 48.92 2.14 23.41 Hillebrand. Floyd County, Ga... ... 0.80 3.52 52.21 13.50 27.72 Nichols. Pulaski County, Ark... . 2.00 3.50 62.05 1.66 30.31 Wilkinson County, Ga. . 9.38 2.76 57.58 0.96 29.12 E. Everhart. 1 0. Veatch, Butt. 18, Georgia Geol. Survey, 1909, pp. 430-447. 2 Mineral Resources, U. S. Geol. Survey, 1913. 356 MINERAL DEPOSITS Great variations are often shown in one locality. For further analyses see G. P. Merrill, Non-metallic minerals, 1910, p. 91. The average of a long series of analyses of commercial ore from Georgia tabulated by T. L. Watson 1 gives: SiO 2 , 4.274; Ti0 2 , 3.791; A1 2 3 , 58.622; Fe 2 3 , 1.507; and H 2 O, 31.435; total 99.629. This corresponds to A1 2 3 .3H 2 O. 1 Bull. 11, Georgia GeoL Survey, 1904, pp. 45-46. CHAPTER XIX THE HEMATITE DEPOSITS OF THE LAKE SUPERIOR REGION General Character, Distribution. The iron-ores mined in the Lake Superior region in Minnesota, Michigan and Wisconsin amount to from 80 to 90 per cent, of the total domestic output and in 1917 yielded 64,000,000 long tons. The ore is mainly hematite with small admixtures of limonite and magnetite. It occurs as masses, lenses, or flat deposits in pre-Cambrian sedi- mentary rocks of Algonkian and Archean age. The deposits are concentrated by the oxidizing and silica-dissolving effect of waters of meteoric origin, in original sediments called "iron formation" which were rich in carbonate and silicate of iron. They are products of pre-Cambrian weathering which, probably under arid conditions, reached depths not approached elsewhere. Only to a small degree and near the surface does this ore forming activity of the waters persist at the present time. We owe most of our information concerning these deposits to the work of C. R. Van Hise, C. K. Leith, and many others re- corded in a magnificent series of monographs of the United States Geological Survey. These and other papers are cited below. 1 1 R. D. Irving and C. R. Van Hise (Penokee district) Man. 19, U. S. Geol. Survey, 1892. C. R. Van Hise and W. S. Bayley (Marquette district) Mon. 28, 1897. J. M. Clements and H. L. Smyth (Crystal Falls district) Mon. 36, 1899. C. K. Leith (Mesabi district) Mon. 43, 1903. J. M. Clements (Vermilion district) Mon. 45, 1903. W. S. Bayley (Menominee district) Mon. 46, 1904. . C. R. Van Hise, Iron ore deposits of the Lake Superior region, Twenty- first Ann. Rept., U. S. Geol. Survey, pt. 3, 1901, pp. 305-434. C. R. Van Hise and C. K. Leith, The geology of the Lake Superior region, Mon. 52, 1911. S. Weidman, The Baraboo iron-bearing district, Wisconsin, Bull. 13, Wisconsin Geol. and Nat. Hist. Survey, 1904. C. K. Leith, A summary of Lake Superior geology, Trans., Am. Inst. Min. Eng., vol. 36, 1906, pp. 101-153. C. K. Leith, The geology of the Cuyuna iron range, Minnesota, Econ. 357 358 MINERAL DEPOSITS There are seven principal districts in the United States and three or four in Canada, locally called ranges, as follows (Fig. 116): 1. The Mesabi, Vermilion, and Cuyuna ranges of northern Minnesota. 2. The Penokee-Gogebic, Marquette, Iron River and Meno- minee ranges, mainly in northern Michigan. 3. The Baraboo range of southern Wisconsin. 4. The Michipicoten, Gunflint Lake, and other minor districts in Canada. Geology. The following short summary is in part taken from the resume's by Leith and Harder cited above. 1 The principal rocks of the Lake Superior iron ore region are divided as follows: Cambrian: Potsdam sandstone. Algonkian: Keweenawan series (sediments, basic flows, gabbro, Huronian series: Upper Huronian (quartzite, "iron formation," and slate). Middle Huronian (quartzite, ''iron formation," and slate). Lower Huronian (quartzite, conglomerate, dolomite, slate, "iron formation," and intrusives). ! Archean: Laurentian series (granite, gneiss, and porphyry). Keewatin series (greenstone, amphibolite, and "iron formation"). Geol, vol. 2, 1907, p. 145; The iron ores of Canada, Econ. Geol., vol. 3, 1908, pp. 276-291. E. C. Harder and A. W. Johnston, Notes on the geology of the Cuyuna district, Bull. 660, U. S. Geol. Survey, 1917, pp. 1-26. J. F. Wolff (Mesabi range), Eng. and Min. Jour., July 17-Aug. 7, 1914; Trans. Am. Inst. Min. Eng., vol. 56, 1917, pp. 142-169. E. C. Harder, Mineral Resources, U. S. Geol. Survey, 1908, Summary. E. F. Burchard, The production of iron ore, etc., Mineral Resources, Annual publication. Carl Zapffe, The Cuyuna iron ore district, Brainerd Tribune, Suppl., July 1, 1911. 1 Classifications differing somewhat from that here given have been proposed in Canada by Coleman, Miller and Knight, but for present pur- poses the nomenclature current in the United States is adhered to. See W. G. Miller and C. W. Knight, Bull. Geol. Soc. Am., vol. 26, 1915, p. 87; Rept. Ontario Bur. Mines, vol. 22, pt. 2, 1914. THE HEMATITE DEPOSITS 359 360 MINERAL DEPOSITS Of these rocks only the Upper and Middle Huronian and the Keewatin contain deposits of hematite. The Archean or basement complex consists of gneiss and gran- ite with an extensive series of greenstones (basalt, gabbro, am- phibolite), which are largely surface lavas. These lavas are now regarded as the oldest formation exposed; the character of the basement upon which they were outpoured is unknown. Above the Keewatin lavas lie sedimentary rocks of the iron formation. The gneisses and granites are in part certainly intrusive into the Keewatin series. In the Vermilion and Michipicoten districts the productive formation is in the Keewatin series. Unconformably overlying the Archean and similarly covered by the Cambrian is the Algonkian, which in its complete develop- ment consists of four parts separated by uncomformities. The lower three divisions are collectively referred to as the Huronian and the uppermost as the Keweenawan. The principal iron- bearing formations are concentrated in the Huronian, but the development differs materially in the several districts. In the Marquette district all three divisions of the Huronian are present. The lower Marquette series consists of quartzite, dolomite, and slate 3,000 feet in maximum thickness. The middle Marquette series, 3,000 feet in maximum thickness, includes quartzite, slate, and the important Negaunee iron- bearing formation. The upper Marquette series includes quart- zite, schist, slates, and fraginental basic volcanic rocks, each member accompanied by iron-bearing formations. In the Crystal Falls and Menominee districts similar divisions appear. In the Penokee-Gogebic district the Upper and Lower Huronian series are present, but the middle series appears to be lacking. The lower division consists of quartzite, cherty lime- stone, and dolomite; the upper part includes quartz slates and the thick Ironwood "iron-bearing formation" and may aggregate about 13,000 feet in thickness. In the Mesabi district the Lower Huronian consists of con- glomerates, graywackes, and slates standing vertically; it is intruded by the granite of the Giants Range, on the south slopes of which the iron deposits extend from east to west for a distance of 100 miles. The Upper Huronian comprises a basal quartzite, the Biwabik "iron-bearing formation," and the overlying THE HEMATITE DEPOSITS 361 Virginia slate. The total thickness is probably over 2,000 feet. The series dips gently at angles of 5 to 20 and is also gently cross folded. Intrusive into these rocks at the east end of the district are Keweenawan granite and basic igneous rocks. Near these intrusives the sedimentary rocks are highly metamorphosed. Above the Huronian rests the less highly metamorphosed Keweenawan series of sandstones, conglomerates, and igneous basic flows; the thickness is estimated to be over 35,000 feet. It contains no iron deposits. The west end of Lake Superior consists of an eastward- pitching synclinorium of Keweenawan rocks. The next underly- ing series, the Upper Huronian, takes less part in this synclinal structure and borders the other edge of the Keweenawan areas. We have thus in the Lake Superior country six series, consist- ing from top to bottom of the Keweenawan, Upper, Middle, and Lower Huronian, Laurentian, and Keewatin, all but the last two separated by unconformities. Above them and separated by a marked unconformity rests the Cambrian Potsdam sandstone. The "Iron Formations." The iron ores of the Lake Superior region are believed to be derived by concentration by means of meteoric waters from lean "iron formations" containing about 25 per cent. iron. The ores are products of enrichment of chemically deposited sediments, such as siderite and hydrated iron silicates, for the most part interbedded with normal clastic sediments, such as slate and quartzite. The iron formations range from a few feet up to 1,000 feet in thickness and are sedimentary beds consisting, according to Leith, "mainly of chert, or quartz, and ferric oxide segregated in bands or sheets, or irregularly mingled. Where in bands with the quartz layers colored red and the rock highly crystalline it is called jasper. Where less crystalline and either in bands or irregularly intermingled the rock is known as ferruginous chert. The silica in these rocks varies from 32 to 80 per cent., the ferric oxide from 31 to 66 per cent. Other phases of the iron formation, subordinate in quantity, are (1) ordinary clay slates, showing every possible gradation through ferruginous slates into ferruginous cherts; (2) paint rocks, oxidized equivalents of the slates; (3) cherty iron carbonate (siderite) and hydrous fer- rous silicate (greenalite) ; (4) the iron ores themselves. Almost the entire bulk of the iron formations now consists of iron oxide 362 MINERAL DEPOSITS and silica, with carbonates and alumina present in subordinate quantity." Spurr and Leith found that certain rocks of the Mesabi district contained, in a matrix of chert and iron carbonate, abundant round granules of a green chloritic substance which Leith called greenalite; its composition is approximately 30 to 38 per cent. SiO 2 , 8 to 34 per cent. Fe 2 3 , 25 to 47 per cent. FeO, and 7 to 9 per cent. H 2 O (p. 263). The absence of potassium shows that the mineral is not glauconite. The greenalite rocks contain 50 to 80 per cent, of this mineral, which is soluble in acids. The same mineral occurs in some of the siderite rocks of the more easterly districts. Regional metamorphism and, to a greater degree, contact metamorphism, caused by Keweenawan intrusions of granites and gabbros, have, in places, converted the siderite and the greenalite rocks to magnetite-amphibole schists and the soft hematite to specularite; this is especially well observed in the Marquette and the Mesabi ranges. The Iron Ores. The hematite ores are derived from the fer- ruginous cherts by a process of concentration, and both laterally and in depth gradually change into such rocks. The ores are admixed with enough magnetite to affect the magnetic needle and render possible magnetic surveys of the fields. The hard blue specular ores of the Marquette range contain more magnetite than the others and are accompanied by contact-metamorphic jaspers and magnetite-amphibole (griinerite) rocks. In other ranges, such as the Mesabi, Penokee, and Baraboo, the ore is soft, bluish, red, or brown in color, and partly hydrated. A micaceous or foliated development of the iron ore is not common. The average analysis of Lake Superior ores in 1909 is as follows: Per cent. Moisture (loss at 100 C.) 11.28 Analysis of dried ore : Iron 58.45 Phosphorus 0.091 Silica 7.67 Alumina 2.23 Manganese 0.71 Lime 0.54 Magnesia . 55 Sulphur 0.06 Loss by ignition 4 12 THE HEMATITE DEPOSITS This corresponds to a composition as follows : Hematite (more or less hydrated) 86.45 Quartz 4.89 Kaolin 5.25 Chlorite 1.01 Dolomite '. 0.81 Apatite 0.48 Miscellaneous ... 1.11 100.00 The tenor in iron of the shipped ore has slowly diminished during recent years; in 1905 it was 59.6 per cent. Fe. The phosphorus ranges between 0.008 and 1.29, the bulk of the ore being of Bessemer grade that is, containing less than 0.05 per cent, phosphorus. Small parts of the ore shipped, particu- larly from the Mesabi and Cuyuna ranges contain as much as 7 per cent, manganese. The sulphur varied from 0.003 to 1.87 per cent., but it averages low. Accessory, more or less rare min- erals in the ore, aside from quartz or chert, are apatite, wavellite, adularia, calcite, dolomite, siderite, pyrite, marcasite, chal- copyrite, tourmaline, ottrelite, chlorite, garnet, mica, rhodo- chrosite, barite, gypsum, analcite, goethite, and turgite. The ore reserves of the Mesabi range are estimated to be 1,385,000,000 tons; those of the whole region 1,475,000,000 tons. The total yield of the Lake Superior ores from 1854 to 1916 has been about 770,000,000 long tons, much the greater propor- tion having been extracted in the last three decades. Carbonate ores are now mined in the Miehipicoten district, Canada. Form of Ore Bodies. The ore forms irregular, often very large, but as a rule distinctly bedded or banded masses in the "iron formations;" in places it is entirely embedded in them. The shape is commonly determined by impervious basements like clayey dikes, decomposed amphibolitic rocks, or folded sedi- mentary beds like slate, which have tended to guide the circula- tion of surface water into certain channels; the ores usually occur in pitching troughs caused by any or all of these factors. In some ranges like the Gogebic, Marquette, and Iron River the strata are strongly folded and may dip at high angles; some of the ore-bodies have been followed to a depth of 1,500 or 2,000 feet. Good ore is mined at present in the Newport mine in the Gogebic district at 2,000 feet. In the Mesabi range the rocks lie horizon- tal; the alteration and concentration have extended over a wide 364 MINERAL DEPOSITS area and few of the mines are deeper than 200 feet. The shal- low deposits of this range are mined on an enormous scale by steam-shovels. The annual production, which reached 41,127,- 323 long tons in 1917, is far greater than that of other districts. Marquette Range. The mines of the Marquette range are near Negaunee and Republic, south and southwest of Marquette. The principal "iron formation," the Negaunee, is in the Middle Huronian, and the sedimentary rocks are intruded and meta- morphosed by basic igneous rocks. Extensive folding has taken place and the strata are compressed into a great synclinal basin. FIG. 117. Longitudinal section of the Montreal Mine, Gogebic Range, Michigan, showing dependence of bodies of oxidized iron ore on dikes. \ The ores lie at the base of the Negaunee formation, where the underlying slates have been folded so as to form pitching synclinal basins, or where dikes have guided the concentrating waters. In part they occur also at the contact of the iron formation with basic intrusions for instance, in pitching troughs between igneous masses and dikes branching from them. The surfaces of the igneous rocks are much altered, leached, and changed to clayey masses, called "soapstone" and "paint rock." Menominee Range. The iron-bearing district extends from western Michigan into Wisconsin, the principal mines being THE HEMATITE DEPOSITS 365 located at Iron Mountain, Norway, and Crystal Falls. The iron formation is chiefly in the Upper Huronian and is called the Vulcan formation; it is overlain by Upper Huronian slate and underlain by a Lower Huronian dolomite. Intricate folding characterizes the structure of the range, the ores of the different areas occurring in separate local basins. The deposits are large and consist of soft red hematite, considerably hydrated FIG. 118. Vertical cross-section of the Newport Mine, Gogebic Range, Michigan, showing position of ore-bodies above dikes. Data from H. L. Smyth. in places, and are generally found in pitching synclinal basins bottomed and capped by slate layers. Penokee -Gogebic Range. This range is in northern Michi- gan and Wisconsin, the principal mines being at Hurley, Iron- wood, and Bessemer. The ore appears in the Ironwood forma- tion (Upper Huronian), which is overlain by slate and under- 366 MINERAL DEPOSITS lain by quartzite and black slate. The dip is steep and the sedi- ments are in part metamorphosed by Keweenawan gabbro; for the most part the Ironwood formation is ferruginous chert. The ores are concentrated in large irregular bodies in the angles between the footwall quartzite or black slate and the igneous dikes (Figs. 117 and 118), these rocks making an impervious trough, toward which the meteoric waters converged. Most of the deposits reach depths of 1,000 feet, and some attain 2, 200 feet. Both soft, partly hydrated ore and hard slaty ore occur. Cuyuna Range. The Cuyuna district is situated near Brain- erd about 70 miles southwest of the Mesabi mines. It extends for 65 miles along the strike of the rocks in a northeast direction. The iron ore here is a, partly hydrated hematite, in places accom- panied by an unusual amount of manganese oxide (up to 30 per cent. Mn). It is contained in the usual iron formation of fer- FIG. 119. Generalized cross-section showing relation of iron-bearing formation to associated rocks in the Mesabi Range, Minn. After J. F. Wolff. ruginous jasper which in. depth appears to change to cherty iron carbonate. The enclosing rocks are slates of various kinds com- pressed into steep folds, the details of which are difficult to trace owing to the covering glacial drift. The ore bodies are elongated following the strike and while some cease at shallow depths others have so far been followed down for 300 feet. The phosphorus ranges from 0.1 to 0.5 per cent. The Cuyuna iron ores were dis- covered by means of the magnetic attraction along the range, due to a small quantity of admixed magnetite. Mesabi Range. In northern Minnesota the Mesabi range extends from east to west for a distance of 75 to 100 miles on the south slope of a prominent ridge called the Giants Range. The principal mines are situated near the towns of Biwabik, Eveleth, Virginia, and Hibbing. The Huronian rocks here lie at gentler inclinations than elsewhere, dipping 8 to 10 S.E. so that the iron formation outcrops in a general northeast-southwest belt (Fig. 119). THE HEMATITE DEPOSITS 367 The Biwabik iron formation of the Upper Huronian con- tains the deposits. It is under- lain by the Pokegama quartzite and covered by the thick Vir- ginia slate, a chloritic and aluminous sedimentary rock. Except at the eastern end of the range, where contact-metamor- phic amphibole-magnetite rocks have developed, the iron forma- tion is composed mainly of fer- ruginous chert. The iron ores cover large irregular areas along the outcrop of the Biwabik for- mation, but descend to rela- tively slight depths, few of the mines being more than 200 feet deep (Fig. 120). The deposits are most abundant at the syn- clines of the transverse folds of the formation. They are bedded and along the edges change rather abruptly to the fer- ruginous chert, from which they are derived by leaching of the silica. This relationship is clearly indicated by the slump- ing of the strata near the edges of the ore masses, as shown in Fig. 121. The iron formation is locally called "taconite." The rain water falling on the truncated edges of the beds cir- culates toward the south, its movement being controlled by the' 'slight synclinal basins, by impervious layers of slate, and by fractures. The secondary concentration of the iron ore has evidently taken place under surface con- m 5 S 368 MINERAL DEPOSITS A -Tension. Cracks in Iron -Formation on Axis of an Anticline. -Ore forming by alteration of Tacoflifc along Fissur& Bedding-Piano. 0- Present Condition of Average Trough Orebody. FIG, 121.- Cross-section showing mode of development and slumping of ore-body at Mesabi Bange. After J. F. Wolff. THE HEMATITE DEPOSITS 369 ditions since the remote time of the post-Keweenawan folding, when the deposits first became exposed; it has also taken place below as well as above the present water-level, which is about 75 feet underneath the surface. Analyses show that the present surface water, containing about 20 parts per million of Si0 2 ", is slowly leaching silica, but re- moves little if any iron. The deposits do not appear to continue underneath the edge of the capping Virginia slate, probably be- cause of the ponding of the water below that impervious forma- tion. The amphibole-magnetite rocks in the eastern part of the !" 5 Original Surface Line Open Pit 13th Level !rr:;:: i:^::r;- JSth Level Greenstone FIG. 122. Vertical section through, the Chandler mine, Vermilion range, Minnesota. After J. M. Clements, U. S. Geol. Survey. district are more stable and have not suffered much alteration by oxidation. During the development of the ore-bodies erosion has continu- ally cut down the iron formation and this truncation has been accompanied by slow downward and lateral migration of the iron. Glacial erosion finally removed much material. The ore is a soft and porous hematite, brown, red, or blue in color, averaging 55 to 58 per cent. iron. It contains little mag- netite, but some turgite and goethite. The mineral composition of the ore in 1909 was approximately in per cent.: hematite, 61.81; 370 MINERAL DEPOSITS limonite, 25.95; quartz, 4.10; kaolin, 5.30; manganese dioxide, 1.30; miscellaneous, 1.54. Sulphur is low and phosphorus varies from 0.03 to 0.07 per cent. There is considerable more phosphorus in the ore than in the ferruginous chert; the greenalite and siderite rocks contain scarcely any phosphorus. Vermilion Range. Northeast of Mesabi, near the Canadian boundary, is the Vermilion range, the principal mines being near the towns of Ely and Tower. The country rock is mostly the Keewatin greenstone, but infolded in it in synclinal basins or troughs is the Laurentian iron formation, known as the Soudan. The ores are associated with ferruginous jaspers in these troughs FIG. 123. Ferruginous chert with greenalite granules, in part replaced by ferric oxide (black). Magnified 40 diameters. After C. K. Leith. and generally have a foot wall of greenstone (Fig. 122). The ore is a dense and hard blue or red hematite which contains a little chalcopyrite, an unusual feature in this region. Origin of Lake Superior Iron Ores. It has been shown by Van Hise and Leith and their associates that the ferruginous cherts, jaspers, amphibolite-magnetite schists, and iron ores of the iron formations result from the alteration either of the cherty iron carbonate or of the greenalite. The small amounts of iron carbonate or ferrous silicate now found in the formations represent mere remnants left unaltered where protected by other rocks. The steps of the alteration may be observed and, in THE HEMATITE DEPOSITS 371 the end products, the structures and textures of the original rock are often remarkably well retained. It is held that the ores and the ferruginous cherts or jaspers on one hand and the am- phibole schists on the other hand represent alterations from the same original type. The source of the ore is not, as a rule, in the present ferruginous cherts, but it was developed from origi- nal lean siderite and greenalite ' rocks. It is held that in the largest deposits ores and jaspers may have developed side by side, at the same time, from such original minerals. Iron carbonate prevailed in the Marquette, Gogebic, Vermilion, and Crystal Falls districts; greenalite in the Mesabi district (Fig. 123). The concentration has been effected, according to the Lake Superior geologists, by water coming more or less directly from the surface, especially at places where such waters converge owing to the existence of impervious underlying formations, such as slate or "soapstone," that form pitching troughs, or owing to brecciation and fracturing of the iron formations. The alteration of the iron formations, resulting in the concen- tration of the iron ores or in the development of ferruginous cherts, jaspers, and amphibolite schists, has taken place in dif- ferent geologic periods under varying conditions. So far as the alteration has proceeded continuously under the influence of surface waters, without interruption by igneous activity or orogenic movements, soft ores and ferruginous cherts have re- sulted. So far as these products have been subjected to deep- seated alteration they have become dehydrated into hard red and blue specular ores and brilliant jaspers. So far as the altera- tion of the original iron formations has taken place within the sphere of influence of great intrusive masses, when waters were heated and oxygen not abundant, or under similar conditions, developed by deep submergence or by orogenic movement, fer- rous silicates and magnetite resulted, as shown in the develop- ment of the grtinerite schists. The concentration of the ores was far advanced before Cam- brian time, as shown by the fragments of ores in Cambrian con- glomerates. Most of the deposits were formed between the Keweenawan and the Cambrian deposition. At the close of pre-Cambrian time the ores were largely as we now find them, though some concentration has been going on since. During the Cretaceous period the region of the Mesabi range, at least, was covered by the sea. 372 MINERAL DEPOSITS Regarding the origin of the cherty iron carbonates, Van Hisehas held that they were derived largely from the more ancient basic volcanic rocks of the Lake Superior region. The iron was leached by underground waters and carried to the sea as carbonate, partly also as sulphate solution, and there deposited as limonite, from which through reduction by organic matter ferrous car- bonate was formed. Somewhat different views have lately been expressed by C. K. Leith, 1 who sums up the origin of the ores as follows: The iron was brought to the surface by igneous rocks and either con- tributed directly to the ocean by hot magmatic waters or later brought there by surface waters from weathered rocks. The iron-bearing minerals were then deposited as a chemical sediment in a conformable succession of sedimentary rocks and still later, under conditions of weathering, were locally enriched to ore by percolating surface waters. "It begins also to appear that the iron, copper, nickel, and silver ores of the Lake Superior and Lake Huron districts are related in a great metallographic province in which the characteristics and distribution of the different ores are initially controlled by igneous rocks. As first deposited the iron formation consisted essentially of iron carbonate or ferrous silicate (greenalite) with some ferric oxide, all minutely inter- layered with chert, forming the ferruginous chert. When these were exposed to weathering the ferrous compounds, the siderite and greenalite, oxidized to hematite and limonite, essentially in situ, although some of it was simultaneously carried and rede- posited. The result was ferruginous chert or jasper, averaging less than 30 per cent, of iron. The concentration of the iron to 50 per cent, and over has been accomplished essentially by the leaching of silica bands from the ferruginous chert and jasper. Infiltration of iron has been on a smaller and more variable scale. The leaching of the silica develops pore space and allows the iron layers to slump, thereby enriching the formation suffi- ciently to constitute an ore." Only a small part of the volume of the iron formations less than 2 per cent. has been altered to ore. Resume. The literature of the Lake Superior iron ores is extensive and many different views have been expressed. J. D. Whitney regarded the ores as of igneous origin, and this view has also been advocated by N. H. Winchell. T. B. Brooks and R. Pumpelly at one time considered them as dehydrated bog iron 1 C. K. Leith, Iron ores of Canada, Econ. Geol., vol. 3, 1908, pp. 276-291. THE HEMATITE DEPOSITS 373 ores, and this view has lately been adopted by S. Weidman in his description of the Baraboo ores of Wisconsin, where the ores appear to grade into dolomites. The views of Van Hise and Leith and their associates, which appear to be generally accepted, have been given above in some detail and in part verbatim. The development of their theory of the origin of the iron ores has been gradual; at first the iron formations were considered as purely sedimentary and the re- crystallization to amphibole-magnetite rocks as evidence of regional metamorphism; later the effects of contact metamor- phism were recognized, and finally it is held that the iron of the iron formations was in large part yielded by the extensive erup- tions accompanying their deposition. Although the ferruginous cherts are still thought to be formed by the oxidation of the siderite and greenalite rocks, which now form a small part of the formations, there seems to be a tendency to regard the iron ores as mainly formed directly by the solution of the silica in the ferruginous cherts. The history of any one group of these deposits is probably even more complicated than would appear from the descriptions. In the Mesabi range, for instance, amphi- bole and adularia occur in the ore, but the development of both these minerals is incompatible with descending and oxidizing waters. The age of the concentration of the iron deserves emphasis. The ores were formed mainly before the Cambrian, as indicated by fragments of ore in the Cambrian conglomerate. Indeed, they were in part developed in inter-Huronian time, even in early Huronian time. This is set forth in the publications cited, but is not generally realized. The ores are not the product of the present circulation and oxidation, but of forces acting in ancient periods when conditions were probably widely different from those of to-day. It is stated that the concentration has also pro- ceeded since pre-Cambrian time, but this assertion seems far from being established, even for the surface deposits of the Mesabi range. Weidman (op. tit.} has pointed out that the present ground- waters are entirely similar in composition in the Paleozoic rocks and in the iron formations and has shown that they do not now transport or dissolve iron or notable quantities of silica. A. C. Lane has shown (p. 440) that the depth reached by the potable surface waters is limited and that in some parts of the iron dis- tricts, as well as in the copper districts of the Keweenawan, they are replaced, at depths of 1,000 to 2,000 feet, by scant and appar- 374 MINERAL DEPOSITS ently stagnant water rich in calcium and sodium chlorides. The present ground-water in a region of high water-level is clearly unable to produce the extensive oxidation shown by the iron ores. Undoubtedly special conditions of circulation existed in pre-Cam- brian time which are not paralleled to-day. Oxidizing waters do not penetrate deeply in temperate regions of high water-level, and even where they reach a depth of a few hundred feet the product is a limonite. The hematites appear to result from oxidation only in arid and tropical countries. The only times at which large bodies of rock could be oxidized to hematite by descending waters would seem to be during epochs of great aridity, when the water-level was exceptionally low; possibly just such conditions prevailed in pre-Potsdam time. Similar extraordinary deep oxidation of pre-Cambrian age has been described by L. L. Fermor from the manganese deposits of India. 1 It may be pointed out that deposits of hematite were formed during this period in the Hartville district in eastern Wyoming, 2 and finally in certain recently described areas in western Arizona. 3 At Hartville lenses of hematite occur in schist along a lime- stone foot wall and have been followed to a depth of 900 feet. Ball shows that the deposit antedates the Guernsey formation, the lowest Paleozoic terrane present, and believes that the iron was leached by descending solutions from the upper part of the schist and deposited in its lower part by replacement. Another question of possible importance relates to the per- centage of phosphorus in the Lake Superior ores. It is remark- ably low for sedimentary deposits in the origin of which organic life played a part. It is still more remarkable that the primary greenalite rocks at Mesabi are almost free from phosphorus. The Baraboo deposits of Wisconsin are peculiar in that igneous rocks are there entirely absent, and in that the hematite grades into the overlying dolomite; the argument advanced by S. Weidman in favor of a primary deposition of the ore as limonite or hematite is not without strength. In spite of the great amount of work done the problem of the origin of the Lake Superior hematites still possesses some puzzling features. 1 Mem., Geol. Survey India, vol. 37, 1909. 2 S. H. Ball, Bull. 315, U. S. Geol. Survey, 1907, pp. 190-205. 3 H. Bancroft, Bull. 451, U. S. Geol. Survey, 1911. CHAPTER XX DEPOSITS FORMED BY CONCENTRATION OF SUB- STANCES CONTAINED IN THE SURROUNDING ROCKS, BY MEANS OF CIRCULATING WATERS GENERAL STATEMENT The water which sinks through the soil and effects the weather- ing of rocks becomes charged with small amounts of carbonates of calcium/ sodium, magnesium, potassium, iron, -and other metals, and also with soluble silica. By far the larger part of it, after a short journey through the belt of oxidation, re- turns to the surface as springs and seepage and is carried off in the watercourses to the sea. A smaller part of this water sinks into the ground and either joins the active circulation, descending in smaller fractures and openings to ascend on the larger fissures and other waterways, or becomes a part of the stagnant or al- most stagnant and gradually diminishing ground-water of deeper levels. In places the active circulation may descend to depths of 10,000 feet. In comparison with the depth of the ground- water, the depth of oxidation or rock decay is on the whole insignificant, and that part of the dissolved substance which is carried down is also insignificant in comparison with the vast amount of underlying rocks, so that we cannot expect that the material added from the zone of weathering will produce any far-reaching changes in the composition of these rocks. Nevertheless dissolved salts are carried down from the weath- ered belt and may cause deposits in open cavities or may form more or less complex replacements. In the openings silica may be deposited as chalcedony, chert, or quartz; calcium carbonate may fill fissures and replace silicates, or ferrous carbonate may be substituted for limestone. Chlorite, kaolin, and sericite may develop in igneous rocks. All these changes are, however, accompanied by renewed solution, and it is a debatable question whether the solution does not more than balance deposition. On the other hand, the water returning to the surface after a jour- ney of varying length, more or less heavily loaded with soluble 375 376 MINERAL DEPOSITS salts deposits these by reason of decrease of temperature or by reaction with other surface waters of different composition. Finally, hydration absorbs much water, both from the active cir- culation and from the more stagnant ground-water, and de- posits of valuable minerals may result from this simple process. In a rough way the deposits resulting from the work of under- ground waters of meteoric origin may be divided into (1) those formed from abundant material contained in the surrounding rocks, for instance, magnesite, serpentine, sulphur (by reduction of gypsum), and certain kinds of hematite; and (2) those formed by the deposition of rarer substances dissolved by the water from the surrounding rocks or from rocks that lie deeper. In this second division it is possible to indicate with great confi- dence the derivation of some substances e.g., barite from certain limestones, and copper from certain basic igneous rocks; but the exact derivation of some other substances may be doubtful. Waters of atmospheric origin doubtless have the power to dissolve many of the rarer metals contained in rocks, to carry them for considerable distances, and to concentrate them in places suitable for deposition; but unless it is aided by higher temperatures at considerable depths below the surface this power is probably not strong enough to produce important deposits of these rarer "metals. BARITE 1 Modes of Occurrence and Origin. Barite, the sulphate of barium, also known as barytes or heavy spar, contains when pure 65.7 per cent. BaO and 34.3 per cent. S0 3 . It is usually white and coarsely crystalline with curved cleavage faces but appears also, especially in residual deposits, with granular, earthy or even fibrous texture. Many barites contain from a fraction to several per cent, of strontium sulphate; the material mined is often quite pure except for small amounts of silica, calcite, gypsum, kaolin and iron hydroxide. Witherite, the barium carbonate, is a much rarer mineral and is found in barite veins, associated with galena. It occurs rather abundantly in such veins in Cumberland and Northumberland 1 E. F. Burchard and W. C. Phalen, Mineral Resources, U. S. Geol. Sur- vey, Annual publication. J. M. Hill, Barytes and Strontium. An excellent review in same pub- lication for 1915, pt. 2, pp. 161-187. Mineral Industry, Annual publication. CONCENTRATIONS FROM SURROUNDING ROCKS 377 in England, particularly at the New Brancepeth colliery where it is secondary after barite. Barite is not a mineral of igneous origin, nor does it occur in contact metamorphic deposits. It is common, however, as lenses and veins in almost all kinds of rocks. Most abundantly, however, it occurs in sedimentary rocks and is probably in almost all cases leached from the country rock by circulating waters. All igneous rocks contain at least a trace of barium oxide but rarely more than 0.1 per cent. Leucite and analcite rocks from the Wyoming-Montana province are unusually rich in this metal, some analyses showing from 0.5 to 1.2 per cent, of BaO probably present in the feldspathoids as a silicate. Granites, rhyolites, andesites, and basalts are poor in barium. In more concentrated form barium is often present in sediments. It has been deter- mined in some limestones; analyses quoted by T. L. Watson from the Ordovician of Virginia show from 0.62 to 1.62 per cent. BaSO^ 1 Limestones of the same age from Missouri contain, according to Steel, only 0.001 to 0.005 per cent. BaO. Some sandstones, like the Cambrian Weisner quartzite of Georgia and Alabama contain barite; it has also been found in shales with sedimentary manganese ores (p. 274). Sea water contains traces of barium and strontium and many natural waters particularly salt brines, hold quite a little barium in solution as chloride or carbonate. Many cases of deposition of barite by natural waters are men- tioned on pp. 103-108. Barium sulphate is soluble in water to the extent of 2.9 milli- grams per liter; it is somewhat more soluble, probably - with decomposition, in waters containing alkaline carbonates and chlorides. Barite is a common gangue mineral in many ore deposits but is here scarcely ever of economic importance. In most cases its origin is to be sought in the rocks traversed by the ascending solutions. The barite deposits worked generally contain few other minerals and are found in sedimentary rocks of all ages as veins and lenses whose width in places may be from 10 to 50 feet. Sometimes a little pyrite, galena and sphalerite is asso- 1 C. W. Dickson, The concentration of barium in limestone, School of Mines Quart., vol. 23, 1902, pp. 366-370. This author fails to find it in many limestones. F. W. Clarke does not mention its presence in single and composite analyses of limestone, Geochemistry, Butt. 616, U. S. Geol. Survey, 1916, p. 558. 378 MINERAL DEPOSITS ciated with the barite. The source of the mineral is undoubtedly in the surrounding sediments from which it has been dissolved by meteoric waters. The larger part of the barite mined in the United States is a residual mineral, forming concretions in clay resulting from the decay of limestone. It is sometimes difficult to separate the residual and the strictly epigenetic deposits for it appears that the same solutions which formed the concretions deposited barite in fracture zones in the underlying rock. Deposits in the United States. The barite deposits now worked are mostly contained in the Paleozoic limestones in the southern Appalachian and the central States, the order of impor- tance being Missouri, Georgia, Tennessee, Kentucky and Virginia. In Missouri barite often accompanies the zinc and lead deposits but the important deposits are found in a separate area in Washington County, in southeastern Missouri, 1 not far from the great lead mines in the Bonneterre (Cambrian) dolomite (p. 461). The principal deposits are found in the shattered and dolomitized Gasconade limestone (Ordovician) as filling of irregular veins and other open cavities. The order of precipitation is given by Steel as follows : A thin coating of chalcedony was first deposited ; this was followed by deposition of a little galena; and this in turn was succeeded by barite, which is the main filling. The series of events was closed by the precipitation of marcasite, dolomitization, and the formation of a second generation of barite. and by a much later coating of ruby-red sphalerite on the older barite. In Georgia barite occurs in the Cartersville district 2 as deposits from solution in fractures and cavities in the Weisner quartzite in intimate association with yellow ocher, and also as nodules embedded in residual clays. The barite deposits of Virginia have been described by T. L. Watson, 3 who states that they are probably caused by the leaching of limestones by meteoric waters. Deep rock decay character- 1 A. A. Steel, Trans., Am. Inst. Min. Eng., vol. 40, 1910, pp. 85-117. Arthur Winslow, Missouri Geol. Survey, vol. 7, 1894, p. 678. E. R. Buckley, Missouri Bureau of Geology and Mines, vol. 9, pt. 1, 1908. 2 T. L. Watson and J. S. Grasty, Barite of the Appalachian States, Trans., Am. Inst. Min. Eng., vol. 51, 1916, pp. 514-559. 3 Trans., Am. Inst. Mm. Eng., vol. 38, 1907, pp. 953-976. T. L. Watson and J. S. Grasty, op. cit. CONCENTRATIONS FROM SURROUNDING[ROCKS 379 izes the whole region. The barite in part fills fractures and in part replaces limestone. It occurs: 1. In crystalline Cambrian or pre-Cambrian limestone as irregular, lenticular lodes or pockets replacing the limestone and associated with calcite and chalcopyrite (Fig. 124). 2. In crystalline schists as filling of fractures. 3. In the Shenandoah (Cambro-Ordovician) limestone as filling of fractures or in residual soil. 4. In Triassic shales and limestone as filling of fissures in a crushed zone. FIG. 124. Section of the Bennett mine, Virginia, showing occurrence of barite as residual and as replacement deposit. After T. L. Watson. Large and pure barite veins have been described from the Ket- chikan district 1 and near Wrangell 2 in Alaska. They are con- tained in crystalline limestone and in schists. Foreign Deposits. Barite deposits are common in all countries. Bodies of exceptional size and purity are found in central Ger- many in sedimentary rocks of Permian and Triassic age. 1 T. Chapin and G. H. Canfield, Bull. 642, TJ. S. Geol. Survey, 1916. 2 E. F. Burchard, Bull. 592, U. S. Geol. Survey, 1914, pp. 109-117. 380 MINERAL DEPOSITS Uses and Production. Barite is used extensively as a pigment in the manufacture of mixed paint and to give weight to paper. It is the raw material for other barium salts, such as the nitrate, which is used in pyrotechnics for green fire. For most of the purposes indicated its purity and white color are essential. The crude material is crushed and treated in log washers and jigs. After grinding the pulp is classified and the settled cream-colored mud is finally treated with sulphuric acid to remove the staining ferric hydrate. The domestic production in 1917 was about 207,000 tons, which came from a great number of small operators in Missouri, Virginia, Kentucky, Georgia, North Carolina, and Tennessee. Before the European war began from 16,000 to 30,000 tons of barite and barium salts were imported annually, largely from England and Germany. The removal of competi- tion with high-grade European barite has greatly stimulated the American industry. The average price was $5.66 per ton. CELESTITE AND STRONTIANITE 1 Strontium accompanies barium as a primary constituent of igneous rocks but is present in much smaller quantities. As celestite (SrS0 4 ) and strontianite (SrCO 3 ) it is sometimes found in fissure veins of hydrothermal origin, but the two minerals are much more commonly found as veins, nodules and layers in sedi- mentary rocks, particularly limestone. In the latter case they are undoubtedly leached by cool surface waters from small quan- tities in the sediments and deposited in convenient places. Celestite is found in crystals and granular masses often of bluish color; but sometimes it is dark or brownish. It usually, but not always, contains BaS0 4 . Strontianite is crystalline, fine- grained, fibrous or nodular and has white, brownish or dark color. It has often been mistaken for calcite, and always con- tains a few per cent, of CaC0 3 . In geods, veins, disseminations and replacements celestite is found in Paleozoic dolomite and limestone of Michigan, 2 New York 3 and Ohio. A cave at Put-in Bay is said to have yielded 1 J. M. Hill, Barytes and strontium, Mineral Resources, U. S. Geol. Survey, pt. 2, 1915, pp. 161-187. 2 E. H. Kraus and W. F. Hunt, Am. Jour. Sci., 4th ser., vol. 21, 1906, p. 237. W. A. Sherzer, Am. Jour. Sd., 3d ser., vol. 50, 1895, p. 246; Rept., Michigan Geol. Survey, vol. 7, pt. 1, 1900, p. 208. E. H. Kraus, Am. Jour. Sci . 4th ser., vol. 18, 1904, p. 30; vol. 19, 1905, p. 286. CONCENTRATIONS FROM SURROUNDING ROCKS 381 150 tons of celestite. Deposits in the limestone quarries of north- western Ohio and southeastern Michigan, near Toledo are said to be of possible economic importance. Strontianite is sub- ordinate. Celestite also occurs in Cretaceous limestone in Texas 1 near Austin. Interesting deposits of celestite have lately been discovered in Tertiary lake beds in Arizona, 2 near Gila Bend where the mineral occurs in sandstone and shale with gypsum and salt; similar beds have been described from the Awavatz mountains near the southern end of Death Valley. Strontianite deposits in Miocene lake beds 10 miles north of Barstow, 3 San Bernardino County, have been discussed by A. Knopf who be- lieves that the beds, veins and fibrous concretions at this place are replacements of lacustrine limestone. All these deposits are of low grade but it is planned to utilize them at the present time. In 1916, 250 tons of strontium ore was mined in the United States, the first production for many years. In 1917 the output had increased to 4,035 short tons. The principal supply of strontium was for many years derived from the Strontianite veins in Cretaceous marl and limestone of Westphalia, Germany; the mineral is here accompanied by calcite and a little pyrite. 4 The strontium used in the United States before the war came largely from England. Important celestite deposits are found near Bristol, where the mineral forms lenses and veins in a Triassic marl and in the underlying rocks. 6 Celestite is also concentrated, like barite, during the formation of sulphur from gypsum and as noted below under "Sulphur" it occurs in considerable quantities in the sulphur mines of Sicily. For commercial purposes celestite should contain at least 95 per cent. SrSO 4 . The principal use of strontium is in sugar refining, in the so- called Scheibler process, in which strontium hydroxide is used 1 First report, Geol. Survey Texas, 1889, p. 125. F. L. Hess, Eng. and Min. Jour., July 17, 1909, p. 117. 2 W. C. Phalen, Celestite deposits in California and Arizona, Bull. 540, U. S. Geol Survey, 1914, pp. 526-531. 3 Strontianite deposits near Barstow, California, Bull. 660, U. S. Geol. Survey, 1918, pp. 257-270. 4 Getting, Oesterr. Zeitschr. B. u. H, Wesen, vol. 37, 1889, p. 113. 8 R. L. Sherlock, Mem., Geol. Survey England, Special reports on mineral resources, vol. 3, 1918, pp. 48-61. 382 MINERAL DEPOSITS for the recovery of sugar from beet sugar molasses. The nitrate is used in pyrotechnics for red fire. The domestic production has been stimulated because of the recent embargo on exports from Great Britain. The price of British celestite was about $12 per ton at the eastern sea board. SULPHUR 1 Modes of Occurrence. Native sulphur may be formed by various reactions. The oxidation of pyrite sometimes results in crusts of sulphur coating the cavities once occupied by the dis- solved crystals. In the craters of volcanoes where sulphurous gases ascend on crevices sulphur is often found, as the result of a reaction between sulphur dioxide and hydrogen sulphide (H 2 S+2S0 2 = H 2 SO 4 +2S), or more probably by incomplete oxi- dation of hydrogen sulphide (2H 2 S+0 2 = 2H 2 0+2S) or by the by the reaction 3SO 2 +2H 2 O = 2H 2 S0 4 +S. A large deposit of this kind is worked at the Abosanobori mine, Hokkaido, Japan, and consists of clayey beds in an old crater lake. Considerable quantities are exported from Japan to the United States. It has been proposed to utilize a similar deposit in the crater of Popo- catepetl, Mexico; other deposits are found in the volcanoes of the Chilean and Argentine Andes. Much more commonly sulphur is found at active or extinct hot springs in the tufas or other adjoining porous rocks like vol- canic tuffs. It evidently results from the incomplete oxidation of H 2 S, by the oxygen or by bacterial action. Such deposits have been observed at many places in the Western States for in- stance, at Cuprite, Esmeralda County, Nevada; at the Rabbit Hole mines in Humboldt County, Nevada; 2 at Sulphur Bank, California; at the Cove Creek mine, Beaver County, Utah; 3 and at Cody and Thermopolis, in Wyoming. 4 The three last-named deposits have been worked. In Wyoming the sulphur in part 1 0. Stutzer, Die wichtigsten Lagerstatten der Nicht-Erze, 1911, pp. 185-263. W. C. Phalen, Mineral Resources, U. S. Geol. Survey, Annual publica- tion. S. H. Salisbury, Mineral Industry, Annual publication. 2 G. I. Adams, Bull. 225, U. S. Geol. Survey, 1904, pp. 497-500. 8 W. T. Lee, Bull. 315, U. S. Geol. Survey, 1907, pp. 485-489. 4 E. G. Woodruff, Bull. 340, U. S. Geol. Survey, 1908, pp. 451-456. E. G. Woodruff, Bull. 380, U. S. Geol. Survey, 1909, pp. 373-380. CONCENTRATIONS FROM SURROUNDING ROCKS 383 replaces the limestoue underlying the travertine or tufa. All these deposits are superficial, and, though some are fairly pro- ductive, they play no great part in the world's output. The greater part of the native sulphur is not connected with volcanic processes or hot springs but is obtained from sedimen- tary beds, in close association with gypsum and limestone; calcite, aragonite, barite, celestite, opal, more rarely quartz, together with gaseous and solid hydrocarbons, are found with the sulphur. This ^'associa- tion is constant and char- acteristic and recurs in al- most all the great gypsum beds of the world, though the sulphur is not always present in quantities of economic importance. As an illustration it is interest- ing to scan the boring records in Louisiana, contained in the bulletins of the State Survey, and note the fre- quency with which sulphur accompanies gypsum. The sulphur is in earthy or resirious masses and forms lenticular beds, veinlets, and concretions in marl, lime- stone, and gypsum. Origin of Sulphur Deposits in Gypsum. Sulphur is undoubt- edly derived from gypsum through the reducing action of organic matter, by way of calcium sulphide and hydrogen sulphide. Regarding the details of the transformation the views are not uniform; it is certain that the reaction can take place at low temperature. G. Bischof, in the middle of the last century, first discussed this matter 1 and assumed the following reactions: CaS0 4 +2C = CaS+2CO 2 . CaS+C0 2 +H 2 = CaCO 3 +H 2 S. = H 2 0+S. FIG. 125. Banded sulphur rock irom Sicily, one-half original size. Black, sulphur; white, calcite; stippled, limestone. After 0. Stutzer. 1 G. Bischof, Chemische und physikalische Geologie, vol. 2, 1851, pp. 144-164. 384 MINERAL DEPOSITS ' The objection to this scheme would be that the sulphur is evidently often formed at depths of several thousand feet, and that .the presence of much oxygen at such depths would be im- probable; more likely the hydrogen sulphide generated from the gypsum reacts upon calcium carbonate, resulting in secondary gypsum and sulphur. The deposits of Sicily have been the subject of extended dis- cussion. A. von Lasaulx 1 has regarded them as formed in fresh-water lakes into which springs containing H 2 S were dis- charged. G. Spezia 2 has advanced a similar view, believing, however, that the hot springs deposited the sulphur at the bottom of a sea basin, accounting for the presence of celestite by the same agency. Baldacci 3 held that the deposition of sulphur took place in a partially evaporated marine basin, in or near which numerous mud-volcanoes, like those of the Apsheron peninsula in the Caspian Sea, discharged large volumes of hydrocarbon that effected the reduction of gypsum to calcium sulphide. A theory of the purely sedimentary origin of the sulphur deposits of Sicily was recently advanced by 0. Stutzer. The well- defined stratification of the sulphur beds, with occasional cross- bedding, the occurrence of the sulphur in limestone and its absence in the overlying gypsum, and finally the presence of intercalated clay beds which would prohibit the free circulation of water are cited by Stutzer as proofs of his view. Sedi- mentary sulphur deposits may form, according to him, in any closed basin in which hydrogen sulphide is developed. The gas may be produced by decay of organisms, or by reduction of dissolved calcium sulphate by carbon, or by hydrocarbons. The oxidation of hydrogen sulphide is effected by the oxygen of the air or by the aid of bacteria. In organic decay many bacteria reduce sulphates and develop hydrogen sulphide. Other low organic forms, the so-called sulphur bacteria, oxidize H 2 S and accumulate sulphur in their cells as minute particles. The oxidation of this sulphur supports the life of the organism, the resulting sulphuric acid being converted into sulphates by 1 Neues Jahrbuch, 1879, pp. 490-517. 2 G. Spezia, SulT origine del solfo nei giacimenti solfiferi della Sicilia, Torino, 1892. Reviewed in Neues Jahrbuch, 1893, 1, p. 281. 3 Descrizione geol. dell Isola di Sicilia, Mem. descritt. d. Carta geol. d'ltalia, 1, 1886. CONCENTRATIONS FROM SURROUNDING ROCKS 385 carbonates which are absorbed and are necessary for the growth of the bacteria. The sulphur bacteria are found in sulphur springs and in the mud of seas and lakes, in which hydrogen sulphide is developed. Stutzer also refers to the fact that the water of closed basins, such as the Black Sea, contains H 2 S in quantities increasing with the depth. Stutzer's conclusions are supported by W. F. Hunt 1 who describes the bacterial action in detail. Interesting as these views are, the sedimentary deposition of sulphur cannot be regarded as proved. The presence of epigenetic sulphur throughout large masses of gypsum is so common that its origin through the direct reduction of gypsum, by way of hydrogen sulphide and organic matter, can scarcely be doubted. Conditions in the Black Sea would seem to be favorable for the deposition of sulphur and yet no sulphur appears to have been brought up by deep dredgings in that basin. 2 Examples. Sulphur is widely distributed in the Miocene and Pliocene of the Mediterranean countries, everywhere accom- panied by gypsum. By far the most important deposits are in Sicily, which for years has supplied the bulk of the world's production. The sedimentary rocks of Sicily, in part marine, in part land deposits, consist of basal clays covered by diatomaceous and radiolarian shales. Above these beds the sulphur-bearing gyp- sum formation extends over an area of almost 800 square kilo- meters. This formation is about 300 feet in thickness; gypsum, limestone, salt, clay, and sandstone are the principal rocks. There are three or four beds of sulphur, the substance ramifying through the bluish-gray limestone. Celestite occurs in econom- ically important quantities and with sulphur, gypsum, calcite, and more rarely barite forms beautiful crystals coating the walls of cavities. The crude ores of Sicily contain from 8 to 25 per cent, sulphur. As noted above, sulphur is common in the Tertiary and Cre- 1 The origin of the sulphur deposits of Sicily. Econ. Geol. t vol. ]0, 1915, pp. 543-579. 1 Sir John Murray, The deposits of the Black Sea, Scottish Geog. Mag., 16, 1900, pp. 673-702. Stelzner and Bergeat, Erzlagerstatten, p. 470. A comprehensive review by Doss is given in the Neues Jahrbuch, 1900, 1, pp. 224-228. 386 MINERAL DEPOSITS taceous beds underlying the Louisiana and Texas coast. 1 In 1865 an unusually large deposit was discovered in Calcasieu Parish (230 miles west of New Orleans), Louisiana, at a depth of 443 feet underneath clay, sand, and limestone of Tertiary and Cretaceous age. The borings showed a thickness of 100 feet of almost pure sulphur, underlain by a great thickness of sulphur- bearing gypsum (Fig. 126). The lateral extent is sharply de- fined as a circular area, half a mile in diameter. It is now evident that these deposits have been formed in the upper part of one of the great salt domes of the Gulf coast (Fig. 102, see also p. FIG. 126. Vertical section of sulphur-bearing bed at Calcasieu parish, Louisiana. After Kirby Thomas. 309). The difficulties of sinking a shaft through the quicksands for a long time prevented the utilization of this deposit. Later, however, the difficulties were overcome by the invention of the Frasch process. 2 Through bore holes superheated water is forced down to the sulphur, which is thereby melted; hot air is then supplied under pressure to aerate the molten mass and facilitate its ascent by water pressure to the surface 1 J. F. Kemp, Mineral Industry, 1893, p. 585. W. C. Phalen, Mineral Resources, U. S. Geol. Survey, 1907, pt. 2, p. 674. See also issue for 1911, pt. 2. 2 H. Frasch, The Mining World, December 14, 1907. W. C. Phalen, Op. tit. CONCENTRATIONS FROM SURROUNDING ROCKS 387 Similar sulphur beds occur at several places along the Gulf Coast. A deposit has recently been opened by drilling by the Freeport Sulphur Company near the mouth of the Brazos River, Texas, and is now in successful operation. Here the sulphur beds lie under 760 feet of gravel, sand and clay. Below this are 150 feet of sulphur bearing limestone, gypsum and dolomite, containing from 10 to 50 per cent, sulphur. These beds are underlain by gypsum, limestone, sandstone and rock salt. 1 For some reason the literature on Texas and Louisiana sulphur deposits is extremely scant. Production. Until recently the Sicilian deposits, with an an- nual output of about 450,000 metric tons, supplied the world's demand. In 1901 the Frasch process revolutionized the trade conditions and the production of the United States rose at once to 200,000 or 300,000 long tons, and the importation from Sicily fell off correspondingly. The interesting trade conditions developing from these changes and the struggles of the Italian government to aid the distressed Sicilian operators are described in the articles in Mineral Resources cited above. In 1915 the Italian production was 364, 260 metric tons while that of the United States was in excess of 400,000 tons with rapidly diminishing imports. The price was about $17 per ton. In 1917 the sulphur production of the United States was probably much larger though, on account of the output being divided between two principal producers, the figures are no longer published by the U. S. Geological Survey. Louisiana, Texas, Wyoming and Nevada are the producing states. Uses. The manifold industrial uses of sulphur need not be specified; the larger part is used for the manufacture of sulphuric acid, for bleaching purposes by the development of sulphur dioxide, for the prevention of mildew on grapevines, and for the manufacture of gunpowder, matches, etc. Sulphuric Acid. Just before the war the production of sulphuric acid in the United States was about 3,500,000 short tons. War conditions soon created an enormous demand for sulphuric acid, mainly for explosives and the price has risen rapidly. In 1916 about 6,250,000 tons of acid (50 Baume) was produced and for 1918 the requirements were 8,000,000 tons. Of the production in 1916, 40 per cent, was made from 1 W. C. Phalen, quoting Thomas Kirby in Mineral Resources, IT. S. Geol. Survey, 1912, p. 936. 388 MINERAL DEPOSITS Spanish (Rio Tinto) pyrite, 6 per cent, from Canadian nyrite, 13 per cent, from domestic pyrite, marcasite and pyrrhotite, 22 per cent, from fumes from copper and zinc smelters, leaving about 19 per cent, which had to be supplied from native sulphur 1 which practically did not appear at all in the acid production of normal times. The scarcity of shipping has reduced the importations from Spain so that great efforts have been made to develop our native supplies of pyrite and sulphur. It appears then that pyritic ores with or without other metals are the principal source of sulphuric acid. Many countries, particularly Spain, Norway, Portugal, France, United States, Italy and Germany in the order of importance stated, produce annually over 200,000 tons of pyrite. In the United States pyrite, marcasite and pyrrhotite with, respectively, 53.3 and 38.4 per cent, of sulphur are the prin- cipal sulphur ores. They are obtained : 1. From pyrite deposits along the Appalachian mountains from Alabama to Vermont. 2. From pyrite deposits in California. 3. From pyrrhotite deposits in Virginia, Tennessee and Maine, 4. From marcasite as a by-product of coal mines in Illinois, Ohio, Indiana and Pennsylvania. 5. From marcasite as a by-product in zinc-lead mines of Wis- consin and Illinois. The domestic production of pyrite (including marcasite and pyrrhotite, was about 400,000 long tons of which the larger part came from Virginia and California. Pyrite was also imported from Quebec and Ontario. The "pyritic deposits" comprise many types (p. 635), but aside from the minor supplies mentioned under 4 and 5, they are mainly products of high or intermediate temperature under intrusive conditions and most of them may be considered as copper deposits of very low grade. Many among those along the Appalachian belt are of early Paleozoic age and more or less strongly dynamo metamorphosed. We may mention the pyrrhotite deposits of Ducktown, Tennessee, and of the " Great Gossan lead," Virginia, further the pyritic deposits of Louisa County, Virginia, which form long lenses in a Cambrian sericite schist and northward change into lead and zinc deposits. Other 1 W. Y. Westervelt and A. G. White, Bull 130, Trans., Am. Inst. Min. Eng., October, 1917, pp. 5-15. CONCENTRATIONS FROM SURROUNDING ROCKS 389 deposits of pre-Cambrian age are found in St. Lawrence County, New York. The ores contain from 30 to 50 per cent, of sulphur with up to 10 per cent, of insoluble. Lead, zinc, antimony and arsenic are objectionable constituents. The residues from sulphuric acid manufacture are often used as copper and iron ores. THE MAGNESIAN DEPOSITS The magnesian silicate rocks lend themselves easily to trans- formation and yield a number of economically valuable products, among them serpentine, magnesite, meerschaum, talc, soapstone, and asbestos. All of these result from the action of water, in most cases doubtless of atmospheric origin, on peridotites, pyroxenites, or gabbros, either near the surface or with the co-operation of stress at greater depths. Talc, soapstone, and asbestos belong, in part, to the latter class. SERPENTINE 1 Serpentine forms by simple hydration from a rock consisting of enstatite and olivine according to the following equation : Mg 2 Si0 4 + MgSi0 3 +2H 2 O = H 4 Mg3Si 2 09. (Olivine) (Enstatite) (Serpentine) ' It may also develop from olivine alone, with the removal of some magnesium as carbonate: 2Mg 2 Si0 4 +C0 2 +2H 2 = H4Mg 3 Si 2 9 +MgC03. The latter equation probably represents the usual process of serpentinization a short distance below the surface. Under oxidizing conditions serpentine is unstable, though of course the change takes place very slowly and erosion may work far ahead of decomposition. Serpentine is, however, also formed on a large scale at greater ^depths, where quantities of CO 2 could not very well be assumed for the reason that such alteration would result in a mixture of serpentine and carbonates, whereas the large serpentine masses rarely contain admixed carbonates. The deep canons of the Sierra Nevada, in California, show clearly that the serpentines of this range are not superficial, but descend to the depth of several thousand feet. The modus operandi of such^ exten- 1 H. Leitmeier in Doelter's Mineralchemie, vol. 1, pt. 1, 1914, 'pp. -385- 428. 390 MINERAL DEPOSITS sive hydration is not fully explained. Some have held that it might have been effected by ascending waters, shortly after the intrusions. Serpentine is generally rich in iron, for the original rocks are not of the purity indicated by the equation given above; the iron is present both as silicate and magnetite, and also in the chromite which forms a characteristic accessory. Rock that is not too much broken by joints finds fairly extensive use as building and ornamental stone. For the latter purpose the oily green translucent varieties formed in crystalline limestone by serpentinization of the contained pyroxene are particularly valued. MAGNESITE 1 Origin. Magnesite (MgCOs) appears in two modifications: (1) As 'an amorphous, earthy, hard and compact mineral, which probably is a hardened colloid precipitate. It is often concre- tionary and has a conchoidal fracture like that of unglazed porcelain. In this form it is an alteration product of ser- pentine or allied magnesian rocks as illustrated by the equation- H4Mg3Si 2 O9+3CO 3 = 3MgC0 3 +2H 2 O+2SiO 2 . (2) As a crys: talline mineral, isomorphous with calcite and usually holo- crystaUine granular. In this form it is generally a replacement of dolomite produced by magnesian solutions in connection with intrusions. 2 1 Robert Scheerer, Der magnesit, Vienna und Leipzig, 1908, p. 256. M. Dittrich, H. Leitmeier, K. A. Redlich in Doelter's Handbuch der Mineralchemie, Dresden and Leipzig, 1912, vol. 1, pp. 212-267. F. L. Hess, The magnesite deposits of California, Bull. 335, U. S. Geol. Survey, 1908. C. G. Yale and H. S. Gale, Mineral Resources, TJ. S. Geol. Survey, Annual publication. S. H. Dolbear, Mineral Industry, Annual publication. 2 Certain minor occurrences are of interest: Magnesite of the amorphous type is found as sedimentary beds and lenses in clays of Miocene lake beds near Bissell, San Bernardino County, California. See H. S. Gale, Bull. 540; U. S. Geol. Survey, 1914, p. 512. Crystalline magnesite occurs in many crystalline schists of the Austrian Alps. Hydromagnesite (3MgC0 3 .Mg(OH) 2 + 3H 2 O) is reported from Atlin, British Columbia, as a deposit of fine white powder several feet deep and appearing like a spring deposit. In connection with this it is recalled that H. Leitmeier found that a magnesian hydrocarbonate was deposited by the mineral waters of Rohitch in Styria. Zeitschr. Kryst. Min., vol. 47, 1909, p. 118. CONCENTRATIONS FROM SURROUNDING ROCKS 391 Occurrence. The amorphous magnesite is not uncommon in areas of serpentine, and it occurs in fissures or crush-zones or irregular masses, often mixed with more or less serpentine and some opal or chalcedony. It is often very pure with slight admixtures of iron, alumina and lime; a few per cent, of free silica are often present. Magnesite occurs abundantly but generally in small deposits in the California 1 Coast Ranges; the best deposits are near Porterville, Tulare County. Until recently the production has been small owing to distance from the eastern market. Similar deposits of great extent producing annually 134,000 tons are worked on the coast of Euboea, in Greece. Other localities are found at Salem, near Madras, India, in the Transvaal, and many other lands where serpen- tinoid rocks abound. Magnesite does not always accompany serpentine, however, and it may be surmised that ascending springs with much C0 2 , as are so common in California, may have some connection with its genesis. The largest magnesite deposits in the world are at Veitsch, in Styria, 2 Austria, where in 1914 about 200,000 metric tons were mined in open quarries. This crystalline magnesite is a replacement of dolomite formed under the influences of intru- sions of porphyry and other acidic and basic rocks. The Austrian magnesite, though otherwise pure, contains like that from Greece a few per cent, of iron which makes it desirable for basic linings and bricks. In 1914, 110,000 tons of this material were imported and the war created a shortage which stimulated local production and search. In 1916 a large mag- nesite deposit was discovered near Chewelah, Washington, which bids fair to supply the demand. Like the Styrian de- posits it is a replacement of dolomite of Carboniferous age, near granite; in places the material contains a few per cent, of silica, lime and iron. Very similar are also the deposits found some years ago in the Grenville township, 3 Quebec, and which are 1 F. L. Hess, Op. tit. H. S. Gale, Bull. 540, U. S. Geol. Survey, 1914, pp. 483-520. 2 K. A. Redlich, Die Genese der Pinolitmagnesite, Siderite und Ankerite der Ostalpen, Tsch. m. und petr. Mitt., vol. 26, 1907, pp. 499-505. K. A. Redlich, Genesis der kristallinen Magnesite, Zeitschr. prakt. Geol, vol. 21, 1913, pp. 90-101. 3 H. J. Roast, The development of Canadian magnesite, Trans., Canad. Min. Inst., vol. 20, 1917, pp. 237-255. M. E. Wilson, Magnesite deposits of Grenville districts, etc. Mem. 98, Canada Geol. Survey, 1917. 392 MINERAL DEPOSITS now being worked. At this locality the magnesite contains several per cent, of lime and very little iron which to some degree has made difficult its use for refractories. Production and Use. The domestic production advanced sharply in 1916 to 154,000 tons, and in 1917 reached 316,000 tons. The price was about $10 per ton. Magnesite gives oft" its carbon dioxide at 800 C., and is, therefore, preferred to calcite in the production of this gas. After calcining, the sub- stance is used for the manufacture of various magnesium salts, and in the paper and sugar industries. It is employed exten- sively with magnesium chloride for the so-called Sorel cement, used for flooring, etc. Its most important use is for basic furnace lining in the Thomas process. Until recently the Styrian magnesite was imported mainly for this purpose. Mag- nesite for bricks should contain a few per cent, of FeO and little CaO; 8 per cent. CaO, being the allowable limit. Magnesite serves also as an ore for the production of metallic magnesium, which on account of its low specific gravity (1.74) is now used in alloys with aluminum and other metals. The reduction is effected by treating the chloride made from mag- nesite in an electric furnace. The best ore for the purpose is naturally the carnallite (KCl.MgCl 2 +6H 2 O) from the Stassfurt salt beds (p. 312). The production of magnesium in United States in 1917 was 116,000 pounds, the price being about $2.00 per pound. MEERSCHAUM 1 Meerschaum or sepiolite (H 4 Mg 2 Si3Oio, containing SiO2, 60.8 per cent.; MgO, 27.1 per cent.; H^O, 12.1 per cent.) is a hydra ted silicate of magnesia of tough, compact texture, white or cream color, and smooth feel. As is well known, it finds a rather exten- sive use in the manufacture of pipes and cigar holders. Its analysis usually shows a little iron, alumina, and lime. It is prob- ably derived from serpentine by slow hydration and is in most cases a colloid precipitate. The principal occurrence is in Asia Minor at Eski-Shehr, where it is found as nodular masses near the surface; at this and several other localities in Crimea and Bosnia serpentine rocks are found in the vicinity, although the material itself is embedded in Quaternary or Tertiary beds. 1 G. P. Merrill, Non-metallic minerals, 1910, pp. 218-221. C. Doelter, Mineralchemie, vol. 2, pt. 1, 1914, pp. 374-383. CONCENTRATIONS FROM SURROUNDING ROCKS 393 A different occurrence is that recently discovered in New Mexico, 1 on the upper Gila River, where the substance forms veins and balls in a Paleozoic cherty limestone. Here it is probably derived from a dolomitic carbonate. TALC AND SOAPSTONE 2 General Occurrence and Origin. Talc (HjMgaSi^^, or 3Mg0.4Si0 2 .H 2 O; 65.5 per cent. SiO 2 , 31.7 per cent. MgO, 4.8 per cent. H 2 0), is a hydrated magnesium silicate, but holds much less water than serpentine. It is a soft, crystalline, foliated or compact mineral of white, gray, or pale-green color and a greasy feel. The more compact, as well as some impure varieties, are usually called soapstone; they may contain shreds of chlorite and other ferromagnesian minerals, like enstatite or amphibole. Soapstone is easily worked and is of great resistance to acids and high temperatures. Talc and soapstone are products of the hydration of magnesian rocks, either of distinctly igneous origin, like gabbro, pyroxenite, or peridotite, or crystalline schists rich in such minerals as enstatite and tremolite or other pyroxenes and amphiboles. These schists may result from the shearing of igneous or contact- metamorphic rocks, the latter derived from the igneous alteration of limestone and dolomite. The purest talc deposits are asso- ciated with crystalline carbonate rocks containing amphibole. In general serpentine forms from olivine and talc from pyroxene and amphibole, but this rule does not always hold. Talc often contains 1 or 2 per cent, of iron and aluminum, as well as a little calcium; according to the analyses given by Merrill (op. tit.} the soapstones contain, in addition to silica and magnesia, from 5 to 11 per cent, alumina, 7 to 13 per cent, ferrous oxide, and 1 to 4 per cent, lime; some of them contain so much water that a strong admixture of serpentine must be assumed. The formulas show that talc may be obtained from enstatite or tremolite by the addition of water and carbon dioxide, with sepa- ration of magnesium or calcium carbonate, which is probably carried away in solution; or, in case of deficiency of CO2, the 1 D. B. Sterrett, Bull 340, U. S. Geol. Survey, 1908. 2 J. S. Diller, Mineral Resources, U. S. Geol. Survey, Annual issues. G. P. Merrill, op. ctt., pp. 208-216. C. Doelter, op. cit., pp. 356-374. 394 MINERAL DEPOSITS magnesia may combine with silica, possibly set free from other minerals, to form additional talcose material. The exact conditions and temperature needed for the formation of talc are not known, but it seems certain that dynamic stress, together with a limited supply of water not over rich in CO2, is favorable to its development; it also undoubtedly forms from magnesian minerals by the aid of a scant supply of surface water under static conditions. It is also known that talc may develop along fissures under the influence of ascending hot waters, whenever magnesian silicate rocks are traversed. E. Weinschenk/ in his description of the talc deposits of the Austrian Alps, holds that the mineral develops by replace- ment of schist composed of quartz, chlorite, chloritoid, and graphite along its contact with limestone and believes this trans- formation due to waters following the irruption of large igneous bodies. Occurrences. The crystalline schists of all countries yield talc. Some occurrences are known from the Pacific coast, but the production in the United States is limited exclusively to the belt of ancient crystalline rocks which forms the axis of the Appalachian Mountain system from Canada to Alabama. North Carolina is rich in talc, and one belt of Cambrian marble along the Nantahala Valley and Nottely River 2 yields many lenses as much as 200 feet long and 50 feet thick. The mineral is mined in open cuts and by shafts and tunnels. New York and Vermont easily outrank all other States in the production of talc. The output of New York comes from a small district about 12 miles southeast of Gouverneur, 3 which has been worked for many years by underground methods. One mine at Talcville has attained a depth of 550 feet. The mineral occurs in schistose layers of enstatite and tremolite, gradually merging into the surrounding crystalline limestone. The deposit forms a persistent layer averaging 20 feet in width, within the enstatite-tremolite rock. Virginia yields most of the soapstone produced in the United 1 Abhandl. Bayer. Akad. d. Wiss., vol. 21, pt. 2, 1901, p. 270. 2 Arthur Keith, Bull. 213, U. S. Geol. Survey, 1903, p. 443. J. H. Pratt, North Carolina Geol. Survey, Economic Paper 3, 1900, p. 99. J. H. Pratt, Mineral Resources, U. S. Geol. Survey, 1905, p. 1361. * C. H. Smyth, Jr., School of Mines Quarterly, vol. 17, 1896, pp. 333-341. CONCENTRATIONS FROM SURROUNDING ROCKS 395 States. It is derived from a belt nearly 30 miles long and less than one mile wide. According to T. L. Watson's description 1 the soapstone occurs as sheets or dike-like masses, 100 feet or more in thickness, conformably interbedded with quartzitic schists, but is probably derived from an igneous rock. Production and Uses.- The rapidly expanding production of talc in the United States was about 198,600 tons in 1917. The larger part was sold in powdered form. The value of powdered talc is about $10 per ton. Talc is used as a filler for paper, including wall paper; also for admixture or adulteration of pigment, as a heat insulator, lubricant, polishing powder of glass, for toilet powders, and as an absorbent for nitroglycerine. The compact talc or soapstone is used for fire-bricks, laboratory tables, gas burners, crayons, etc. Pyrophyllite. 2 Pyrophyllite is a hydrous silicate of alumina (H 2 Al 3 Si4Oi 2 ), containing 66.7 per cent. SiO 2 , 28.3 per cent. A1 2 3 , and 5.0 per cent. H 2 O. In composition and physical qualities it is similar to talc, though it does not command as high a price as the best talc. It is mined in Moore and Chatham counties, North Carolina, where it occurs in thick beds asso- ciated with slate. ASBESTOS 3 Amphibole Asbestos. The asbestos of mineralogy is a mono- clinic amphibole which develops in seams and slips in normal amphibolitic rocks, especially where the rocks have been sub- jected to pressure and movement. Chemically it is a calcium- magnesium metasilicate. According to the series of analyses given by Merrill the silica varies from 52 to 58 per cent., the lime from 12 to 16 per cent., the magnesia from 20 to 30 per cent. Other constituents are alumina, varying from 1 to 6 per cent., and ferrous oxide, usually from 1 to 6 per cent., though in some cases considerably higher. Water is always present, the amount generally varying between 2 and 5 per cent. Although contrary to the views of some authorities, the conclusion can 1 T. L. Watson, Mineral Resources of Virginia, 1907, p. '293. 2 J. H. Pratt, Op. tit. 3 G. P. Merrill, Proc. U. S. Nat. Mus., vol. 18, 1895, p. 181. Bull. Geol. Soc. Am., vol. 16, 1905, p. 113. Non-metallic minerals, 1910, pp. 183-197. F. Cirkel, Chrysotile-asbestos, Canada Dept. of Mines, Mines Branch, 1910; 316 pp. J. S. Diller, Mineral Resources, U. S. Geol. Survey, Annual publication. 396 MINERAL DEPOSITS hardly be avoided that the water is an essential constituent and that the mineral is really a hydrated form of tremolite or actino- lite. The extinction angle appears, however, to be that char- acteristic of these amphiboles, or about 18. No experiments appear to have been made as to the temperatures at which the water is driven off. The normal varieties of amphibole also hold a little water, but in far smaller quantities than asbestos. Anthophyllite (Mg,Fe) SiO 3 , and crocidolite, NaFeSi 2 6 .- FeSiO 3 , a dark blue sodium amphibole, also yield asbestiform varieties. Merrill has shown that the fibers are polygonal in outline and run out into needle-like points; down to a diameter of 0.002 or 0.001 millimeter the fibers retain their uniform diameter and polygonal outlines. The color of amphibole asbestos is usually white to greenish white. Only the finer kinds are utilized, but even these are less valued than the serpentine asbestos. They are apt to be less flexible and somewhat brittle. Most of the small quantity of asbestos mined in the United States is of the tremolite or actinolite variety, and it often occurs in limestones which have been partly metamorphosed to amphibolitic rocks. The mineral is classed as slip-fiber or cross-fiber, according to the position of the fibers in the veinlets. The radial or divergent structures are designated as mass-fiber. There are many occurrences, mainly in pre-Cambrian rocks along the Appalachian Mountain system, from Vermont to Ala- bama. One of the most important localities worked is at Sail Mountain, Georgia, where, according to Diller, the asbestos occurs in large lenticular masses in gneiss and is believed to be an altered igneous rock. Almost the entire domestic production is derived from Georgia. Serpentine Asbestos (Chrysotile). Chrysotile asbestos is green or yellowish-green and is easily reduced to a white fluffy state. The fiber is short, but of very uniform diameter and great divisibility and flexibility; the decomposing effect of hydrochloric acid also distinguishes it from amphibole asbestos. In composition it is practically identical with the purer kinds of serpentine. A typical analysis of the Canadian material yielded per cents, as follows: 42 Si0 2 , 42 MgO, 14 H 2 O, 1 FeO, and 1.7 A1 2 O 3 . Fig. 127 shows the appearance of the two kinds of asbestos. CONCENTRATIONS FROM SURROUNDING ROCKS 397 This variety is found as veinlets, rarely over 6 inches thick, in serpentine or peridotite, and has almost always a cross-fiber that is, the silky fibers lie perpendicularly to the plane of the veinlet. The pure yellowish-green serpentine which occurs in conta metamorphic limestone and which is an alteration product of diopside sometimes contains chrysotile of exceptionally high FIG. 127.- -Chrysotile (a) and amphibole (b) asbestos. Photograph by J. S. Diller. grade. A deposit of such aterial is now worked in Arizona, n or theast of G lobe . l Chrysotile veinlets may be found in almost any serpentine area, but they are rarely so abundant and large as to be of economic importance. The views regarding their origin differ. Dresser shows that serpentinization in the Canadian deposits proceeded along irregular cracks in the peridotite, and the chryso- tile veinlets are found in the center of the serpentinized bands. 1 J. S. Diller, Mineral Resources, U. S Geol. Survey, 1917, pt. 2, p. 197. 398 MINERAL DEPOSITS These veinlets were interpreted by Pratt and Merrill as fillings of contraction cracks, but other authors are probably correct in considering them the result of a recrystallization of the serpen- tine, proceeding inward from the cracks. S. Taber 1 believes that all cross-fiber veins are formed by a process of lateral secretion, the growing veins pushing aside the enclosing walls. Since, however, the material in the veins is derived from the serpentine itself it is not apparent why there is any need of increase of volume. Until about 1895 the small quantity of asbestos used in the United States came from Italy. Since that date the development of the asbestos industry in Canada has been extremely rapid, and the Canadian mines now supply this country. The Canadian deposits 2 center in Asbestos Hill at Thetford, in the eastern town- ships of Quebec. As stated, the mineral occurs as irregular veinlets in serpentine and peridotite. These rocks are in places accompanied by somewhat later gabbro and granite and all of them are intrusive into Ordovician sediments. The mineral is mined in open pits, one of which, for instance, is 700 feet long, 200 feet wide, and 165 feet in greatest depth. A small percentage is obtained by hand cobbing, but the larger part 30 to 60 per cent. of the crude material quarried is crushed and screened, and the fibers are separated by air currents. 3 The extraction of fiber of the milled rock is from 6 to 10 per cent. Of late years the Russian chrysotile from the Ural Mountains and the deposits in southern Rhodesia as well as the crocidolite asbestos from Griqualand West, Cape Colony, are becoming important. The large deposits of crocidolite occur in thin layers interbedded with jaspers and iron stones of the Pretoria series. In the United States 4 chrysotile of economic importance is worked in Vermont, near Casper, Wyoming, and in Arizona. Thus far, the production is small. Uses. "The fundamental property of asbestos, upon which its use depends, is its flexible, fibrous structure, but coupled with this are the scarcely less important qualities of incombustibility 1 Bull. 120, Am. Inst. Min. Eng., Nov., 1916, pp. 1973-1998. 2 J. A. Dresser, Econ. Geol., vol. 4, 1909, pp. 130-140. J. A. Dresser, Preliminary report on the serpentine, etc., of southern Quebec, Mem. 22, Canada Geol. Survey, 1913, p. 103. 3 W. J. Woolsey, Jour. Can. Min. Inst., vol. 13, 1910, pp. 408-413. 4 J. S. Diller, Bull. 470, U. S. Geol. Survey, 1910, pp. 506-524. CONCENTRATIONS FROM SURROUNDING ROCKS 399 and slow conduction of heat and electricity when the mass is fiberized and porous." The spinning and weaving of fire-proof cloth form an important part of the asbestos industry carried on in the United States with Canadian raw material. The highest grade of the crude mineral is expensive, costing $275 to $350 per ton; the fines cost $25 to $125 per ton, while the lowest grade a mixture of serpentine and asbestos is sold at less than $1 per ton. Amphibole asbestos is much cheaper, costing about $18 per ton. Crocidolite is more easily fusible but is more resistant than chrysotile to acids and sea water. The London price is about $125 per ton. 1 The Canadian production in 1917 was 144,185 tons; almost the whole production was exported to the United States. In the same year the output in the United States was 1,683 tons. ORES OF COPPER, LEAD, VANADIUM, AND URANIUM IN SAND- STONE AND SHALE General Features. Ores of copper, lead, vanadium, and ura- nium are often found disseminated in sandstones and shales far from igneous rocks. The sedimentary strata containing the ores are usually parts of thick series of terrigenous or shallow- water beds, commonly of reddish color. The ores are of low tenor and can be utilized only in exceptional cases. Never- theless this class of deposits presents many interesting features. The ore minerals are chalcocite, galena, roscoelite (a vana- dium mica), various copper and lead vanadates, carnotite (a vanadate of uranium), etc. Bornite, chalcopyrite, and pyrite are less common. The ores frequently carry small amounts of silver, nickel, cobalt, molybdenum, and selenium. Gangue minerals occur sparingly and are usually confined to a little barite, calcite, and gypsum. The outcrops are likely to be bril- liantly colored by malachite and azurite. While the deposits are confined to certain formations or members, they do not continu- ously follow a particular horizon and give no evidence of being of sedimentary origin. They often appear in fractured and brec- ciated beds or in strata rich in carbonaceous matter and plant remains. More rarely the ores follow distinct fissures in the sedimentary rocks. They do not seem to have any genetic relation with thermal springs. The copper, lead, and vanadium *P. A. Wagner, South African Journal of Industries, Nov., 1917. 400 MINERAL DEPOSITS deposits form three groups in this class, but each group is likely to contain more or less of the other metals. There is no reason why the deposits should be confined to any particular geological age, but as a matter of fact almost all of them are in the upper Carboniferous, Permian, Triassic, or Jurassic. Origin. In considering the class as a whole it appears that igneous agencies had no part in the genesis. The ores are assuredly epigenetic and their universal appearance in land or shallow-water beds is significant. In all probability these ores have been concentrated by meteoric waters which leached the small quantities of metals disseminated in the strata. The sediments were rapidly accumulated, under arid conditions, from adjacent land areas and the metals were probably carried down in fine detritus and in solutions from older ore deposits in these continental areas. The waters which concentrated the ores are believed to have been mainly sodium chloride and calcium sulphate solutions containing sulphates and perhaps chlorides of copper and lead. The mineral association and geological features indicate deposi- tion at low temperature, probably well below 100 C., and at shallow depths but below the zone of direct oxidation. Very likely these ores have been forming continuously since the estab- lishment of active water circulation in the beds; in favorable places below the surface concentration may now be in progress. COPPER AND LEAD DEPOSITS IN SANDSTONE European Occurrences. 1 The European occurrences are con- fined to the Permian and the Triassic, both, generally speaking, ages of arid climate and saline deposits. The Russian Permian, extending far west from the Urals, con- sists in its lower division of sandstones, marls (in part marine), and conglomerates. The sandstones are rich in vegetable re- mains. Copper ores are found over wide areas, but have not been worked extensively of late. The average tenor is said to be 0.9 per cent, metallic copper. The chalcocite ores replace plant remains and tree trunks or form the cement of the sand- stones. The minerals mentioned from this locality are (besides 1 For an excellent review of European localities, as well as complete index of literature, in part difficultly accessible, see Stelzner and Bergeat, Die Erzlagerstatten, 1904, pp. 388-439. CONCENTRATIONS FROM SURROUNDING ROCKS 401 secondary malachite and azurite) chalcocite, chalcopyrite, barite, vanadinite, and volborthite (vanadate of copper and calcium.) Recently much interest has been taken in the copper deposits of the Khirgiz Steppes, 1 between the Urals and the Altai, in the Karkaralinsk and Akmolinsk districts. Very rich copper ores, consisting of malachite, azurite, and bornite, have been found here in sandstones reported to be of Paleozoic age. At Nankat, west of Kokand in Turkestan, deposits of metallic copper have been discovered in sandstones and gypsiferous marls of Tertiary age; fossil wood and chalcocite are also found. 2 The lower Permian (Rothliegende) of Bohemia, 3 along the south slope of the Rieserigebirge, contains similar ores. Over a large part of western Europe the Triassic is copper- bearing, and together with the copper more or less lead is found. In England, at Alderley Edge and Mottram St. Andrews, 4 south of Manchester, copper ores have been mined. They occur in the cement of Triassic sandstones and conglomerates and consist of copper carbonates, galena, pyromorphite, and vana- dinite; also some barite, manganese, and cobalt. The ores are said to contain at most 1.4 per cent, copper. The mineral mott- ramite, a vanadate of copper and lead, was discovered at this place and vanadium was extracted from the ores. In Germany the Triassic is divided into three parts the lower Variegated Sandstone ("Buntsandstein"), the middle Shell Limestone (Muschelkalk) ; and the upper marls and sandstones (Keuper); of these the lower and upper divisions contain lead and copper ores. In Bavaria the Keuper contains galena and chalcopyrite in certain gypsiferous beds, and these minerals are associated with a little zinc blende and barite. In Wurttemberg galena with barite and some oxidized copper ores is generally distributed in the Corbula bed of the lower, gypsiferous Keuper. In the Palatinate, near Freihung, the littoral characteristics of the formation are plainly indicated and there is an abundance of fossil wood; at two horizons the sand- stones contain galena and cerussite and were formerly worked. 1 A. Addiassewich, A journey to central Asia, Trans., Inst. Min. and Met., vol. 17, 1907-1908, pp. 498-522. 2 R. Beck, Lehre von den Erzlagerstatten, 1909, vol. 2, p. 172. 3 F. Gurich, Zeitschr. prakt. Geol, 1893, pp. 370-371. 4 Phillips and Louis, Ore deposits, 1896, pp. 266-269. 402 MINERAL DEPOSITS In the " Buntsandstein " in Prussia and Lorraine, near Saar- louis and other places, a formation known as the Voltzia sand- stone is particularly rich in lead and copper ores, which at times have been mined. The bed contains abundant plant remains. The minerals are cerussite, galena, chalcocite (?), and copper carbonates. The best-known deposits of the European Triassic are those of Commern and Mechernich, not far from Aix-la-Chapelle, in Prussia. Lead ores have been mined here for several hundred years, but it is reported that the mines may soon be closed. The ores are of low grade and are mined in open cuts by removing about 130 feet of overburden. In 1903 the ores averaged 1.5 per cent. lead. The ore minerals are galena and cerussite, with a little chalcopyrite and barite, the latter filling veins and veinlets in the sandstone. Small amounts of silver, nickel, and cobalt are present. The thickness of the ore-bearing sandstone is about 20 meters. The general occurrence of the galena in so-called "Knoten" or knotty concretions is very remarkable. They often enclose several sand grains and some of them are bounded by the crystal faces of the galena. The epigenetic character of the ore is beyond doubt. American Occurrences. 1 On the North American continent copper ores are widely distributed in the "Red Beds" of the Southwest, in Texas, Oklahoma, New Mexico, Arizona, Colorado, 1 E. T. Dumble, First Ann. Rept., Geol. Survey Texas, 1889, p. 186. E. J. Schmitz, Copper ores in the Permian of Texas, Trans., Am. Inst. Min. Eng., vol. 26, 1896, pp. 1051-1052. S. F. Emmons, Copper in the Red Beds, Bull. 260, U. S. Geol. Survey, 1905, pp. 221-232. W. H. Emmons, The Cashin mine, Bull. 285, U. S. Geol. Survey, 1906, pp. 125-128. E. P. Jennings, Trans., Am. Inst. Min. Eng., vol. 34, 1904, p. 839. H. W. Turner, Trans., Am. Inst. Min. Eng., vol. 33, 1903, p. 678. W. Lindgren, L. C. Graton, and C. H. Gordon, The ore deposits of New Mexico, Prof. Paper 68, U. S. Geol. Survey, 1910. H. S. Gale (Idaho), Bull. 430, U. S. Geol. Survey, 1909, pp. 112-121. W. Lindgren (Colorado), Bull. 340, U. S. Geol. Survey, 1907, pp. 170-174. W. A. Tarr (Oklahoma), Econ. Geol, vol. 5, 1910, pp. 221-226. A. E. Fath (Oklahoma), Econ. Geol., vol. 10, 1915, pp. 140-150. L. M. Richard (Texas), Econ. Geol, vol. 10, 1915, pp. 634-650. A. F. Rogers, Origin of copper ores of the "Red Bed" type, Econ. Geol, vol. 11, 1916, pp. 366-380. CONCENTRATIONS FROM SURROUNDING ROCKS 403 Wyoming, Utah, and Idaho and, though conspicuous by green and blue colors, rarely prove of economic importance. The ore occurs in arkose sandstone, conglomerate, or clay shale and is usually associated with plant remains and fossil wood. These strata were accumulated in shallow seas or as subaerial deposits by a process of rapid degradation of adjacent land areas of the Rocky Mountain region, and they have been referred to the upper Carboniferous, Permian, Triassic, and Jurassic. In Texas the copper-bearing beds appear over large areas in Permian sandstones and shales. They lie at several horizons, in strata rich in plant remains; covellite, chalcocite, chalcopyrite, and pyrite are the minerals mentioned. The area is said to ex- tend from 33 to 34 in latitude and from 98 to 100 in longitude. FIG. 128. Chalcocite nodules replacing fossil wood and carbonaceous shale of "Red Beds," Red Gulch, Colorado. Natural size. In Oklahoma nodules of chalcocite are found in red shales and sandstones of the same age. Fossil wood is often converted to chalcocite, sometimes with a shell of chalcopyrite; unusually high silver assays of 31 ounces per ton and traces of gold are reported. In Colorado these ores have been recorded at several places, notably at Red Gulch, Fremont County, where Lindgren ob- served nodules of chalcocite with barite in black carbonaceous shale; sections (Fig. 128) show that the copper sulphide actually replaces the coal and shale. The horizon is probably the upper- most Carboniferous. In northeastern Arizona, according to Greg- ory, 1 small quantities of oxidized copper ores are frequently seen 1 H. E. Gregory, Prof. Paper 93, U. S. Geol. Survey, 1917, p 140 J. M. Hill, Bull. 540, U. S. Geol. Survey, 1913, p. 163. 404 MINERAL DEPOSITS in the La Plata sandstone. North of the Colorado River they appear in the Carboniferous of the Kaibab Plateau. S. F. Emmons believed that the copper in the oxidized ore and chal- cocite in the Aubrey limestone near Grandview, Arizona, was leached from the "Red Beds" and carried down into the limestone. In southwestern Colorado copper, often accompanied by vanadium ores, is widely distributed in the Jurassic La Plata sandstone. W. H. Emmons has described the Cashin vein in this formation near Placerville. The ores are here argentiferous chalcocite, covellite, and bornite, with some calcite. No igneous rocks are present and Emmons believes that the ores were leached from the " Red Beds." There is an active circulation of water in the formation, and springs with salt, sulphates, and hydrogen sulphide abound. A production of about 300,000 ounces of silver and 700,000 pounds of copper is recorded from this mine. The greatest development of the copper-bearing sandstones is in New Mexico; considerable production from picked ore has been achieved at the Nacimiento deposits, in the northern part of the State, where the "Red Beds" rest on pre-Cambrian gra- nitic rocks which contain much older copper deposits. The beds have been referred to the Triassic on the evidence of fossil plants. According to Schrader 1 most of the copper ores occur in the basal beds and are confined within a thickness of 25 feet in a reddish- white sandstone rich in fossil wood, which is largely chalcocitized. A tree trunk 60 feet long with a basal diameter of 2% feet is men- tioned, which was almost wholly converted to copper glance. Besides malachite, azurite, and chrysocolla, there is some barite and, at one place, cerussite. The low-grade ores have not been utilized. According to the same geologist the copper-bearing beds of the Zuni Mountains, in northeastern New Mexico, 2 lie at the base of the "Red Beds," resting on pre-Cambrian gneisses which contain older copper veins. The sandstones, shales, and marls for 30 to 60 feet just above the base of the beds contain oxidized ores and chalcocite replacing wood. Graton describes in detail the ores from the Tecolote district, San Miguel County, which are partly in the "Red Beds" of the upper Carboniferous (Abo formation), and partly at a higher 1 F. C. Schrader, Prof. Paper 68, U. S. Geol. Survey, pp. 141-149. * Idem, p. 134. CONCENTRATIONS FROM SURROUNDING ROCKS 405 horizon, perhaps in the Dakota sandstone. The calcareous cement of the arkose is replaced by chalcocite, bornite, chal- copyrite, and pyrite, the replacement extending into the feld- spar grains. In the Oscura Range, also in New Mexico, red sandstones, probably Carboniferous "Red Beds," contain chalcocite, bornite, and chalcopyrite, in part as replacement of fossil wood. Turner mentions the occurrence of plant remains, said to have been identified as the Triassic Podozamites crassifolia, the same cycad which is characteristic of the deposits at Abiquiu first studied by Newberry. Graton believes, contrary to Turner, that the copper ores have been introduced into the strata by waters ascending along a number of dislocations in the sandstone. Finally, H. S. Gale describes copper ores from southern Idaho which occur in the Ankareh maroon shales and sandstones of the Triassic or Carboniferous (equivalent to the Permo-Carbonif- erous of the Fortieth Parallel Survey). (See Fig. 10.) A thick limestone (including the Meekoceras beds) underlying these shales is believed by some geologists to be Triassic. There are then at least two main cupriferous formations in the Southwest (1) the upper Carboniferous "Red Beds," equivalent to the Permo-Carboniferous, or the Abo formation; (2) the undoubtedly Jurassic La Plata sandstone. The silver deposits in the supposedly Triassic sandstones of Silver Reef, 1 in southern Utah (Harrisburg district), which created a boom about 1880, are now worked only on a small scale. The ores were silver chloride above the water level and native silver and argentite in depth; copper was also present, and selenium is reported. Plant remains were abundant. A sec- ondary concentration from a primary argentiferous chalcocite is the probable genesis. In Nova Scotia, Cumberland County, chalcocite nodules, with remains of pyrite, and also chalcocitized wood, are found in the Permian sandstone. 1 C. M. Rolker, The silver sandstone district of Utah, Trans. Am. Inst. Min. Eng., vol. 9, 1881, pp. 21-33. J. P. Rothwell, The silver sandstone formation of Silver Reef, Eng. and Min. Jour., vol. 29, 1880, pp. 25, 48, 79. J. S. Newberry, Report on the property of the Stormont Silver Mining Company, Eng. and Min. Jour., vol. 30, 1880, p. 269; vol. 31, 1881, pp. 4-5. J. F. Kemp, Ore deposits of the United States, 1900, p. 334. 406 MINERAL DEPOSITS South America. The well-known and long-worked copper deposits of Coro-Coro, 1 in Bolivia, are contained in a series of sandstones, believed to be of Permian age. There are several cupriferous beds with disseminated native copper, in places den- dritic, and much gypsum, also some native silver, domeykite, and chalcocite. The copper-bearing beds are much lighter in color than the prevailing deep-red sandstones. According to Steinmann the strata are of Cretaceous age and the copper was introduced by hot waters derived from an intru- sion of diorite. Nevertheless the descriptions suggest strongly that the deposits belong in a different class. Africa. Sufficient information is not at hand to decide whether the recently opened Katanga ores 2 of southeastern Belgian Kongo, near Rhodesia, belong to the class of deposits described in this chapter. Large masses of high-grade oxidized copper ores are contained in sandstones, shale, and limestone, probably of Paleozoic age. The ores are of high grade (8 to 12 per cent, copper) and are said to contain a little gold and silver; some manganese, cobalt, and nickel are present. Barite and quartz appear as gangue minerals. Genesis of Sedimentary Copper Ores. The epigenetic char- acter of the copper deposits in sandstone is proved beyond rea- sonable doubt. The replacement of coal, carbonaceous shale, and calcareous or kaolinic sandstone cement by chalcocite is also proved. The gangue minerals are few and quartz is con- spicuously absent. Barite in small amounts is rather common. Irregularity in dissemination is typical, though the ores often follow certain horizons rather persistently. The entire independ- ence of the occurrence of igneous rocks is marked. The occurrences are mainly on the flanks of older continental areas containing pre-Cambrian copper deposits; the sandstones were rapidly laid down as arkoses, indicating a long epoch of rock decay, the products of which were swept away during a following arid epoch. Considering the evidence as a whole the sedimen- 1 Older literature: See Stelzner and Bergeat, Die Erzlagerstatten, vol. 1, 1904, p. 419. G. Steinmann, Rosenbusch Festschrift, 1906, pp. 335-368. F. C. Lincoln, Min. and Sd. Press, Sept. 29, 1917. Lester W. Strauss, Min. Mag., vol. 7, 1912, p. 207. 2 S. H. Ball and M. K. Shaler, Mining conditions in the Belgian Congo, Trans., Am. Inst. Min. Eng., vol. 41, 1911, pp. 189-219; also Econ. Geol, vol. 9, 1914, pp. 617-632, with literature. CONCENTRATIONS FROM SURROUNDING ROCKS 407 tary deposits must have contained finely divided copper ores, in part from solutions derived from the land area, in part as cupriferous detritus. When atmospheric waters charged with salt and gypsum searched these beds they must have taken this copper into solution and concentrated it at certain horizons when reducing substances like coaly vegetable matter were available. In most cases the solution probably contained the copper as sulphate, though where much salt was present it might well have been transformed into chloride. Genesis by ascending thermal solutions of meteoric origin is a possible cause of some deposits. Graton, Fath and Rogers have noted pyrite, bornite and chal- copyrite in the chalcocite nodules and the latter two authors have shown that the chalcocite replaces earlier pyrite or marcasite (Fig. 128). No doubt this is true in places but this view may not be universally applicable. Rogers 1 believes that the fossil wood was successively replaced by hematite, pyrite, bornite and chalcocite, a rather improbable series of events, considering that the wood structure is preserved even in the fourth and last re- placement. Graton finds sharp cubes of pyrite in the chalcocite and no evidence of replacement of pyrite. In the precipitation the most important chemical reactions were those between the hydrocarbons of plant remains and the calcareous cement or the kaolin in the sandstone on one hand and the cupriferous solutions on the other hand. How the metallic copper in these ores was precipitated is not known. The Bolivian occurrences show distinct bleaching of the reddish sandstone around the copper aggregates, from which it may be inferred that the solutions were reducing in character. VANADIUM AND URANIUM ORES IN SANDSTONES' Composition. Many of the copper deposits described above carry some vanadium as vanadinite or volborthite. Lately vanadium with some uranium and a trace of radium has been 1 Op. tit. 2 G. P. Merrill, Non-metallic minerals, 1904, pp. 299-320. W. F. Hillebrand and F. L. Ransome, Carnotite, etc., in western Colorado, Am. Jour. Sci., 4th ser., vol. 10, 1900, pp. 120-144. Bull. 262, U. S. Geol. Survey, 1905, pp. 9-13. H. Fleck and W. G. Haldane, Rept. State Bureau of Mines, Colorado, 1907, pp. 47-115. 408 MINERAL DEPOSITS shown to be common in certain Jurassic sandstones in Colorado and Utah. The deposits in western Colorado are now worked and a reduction plant is located at Vanadium, near Placerville. A number of unusual minerals are contained in these deposits. One of the most conspicuous is carnotite ^UsOs.V^Os.I^O.SHaO) a crystalline potassium-uranium vanadate, first named by Fuchs and Cumenge, which forms a bright yellow powder occurring in seams and on fossil wood. An analysis by W. F. Hillebrand gave : Per cent. Per cent. UO,.... 54.89 CuO 0.15 V Z O 8 18.49 MoO 3 0.18 CaO 3.34 H 2 4.54 BaO 0.90 CO- 0.56 K 2 6.52 Insoluble 7.10 PbO 0.13 In calcio-carnotite potassium is replaced by calcium. Associated with carnotite are a number of other obscure vanadium 'minerals, which appear as earthy black, brown and red coatings or fissure fillings. They are crystalline and highly hydrous vanadates. Metahewettite 1 (CaO.SV^Os.OH^O) is a dark red calcium vanadate containing, in per cent., 70V 2 3 , 7.25CaO and 21.30H 2 0. Pintadoite is another mineral of similar composition but with only 42.4 per cent. V 2 O 5 ; it forms green efflorescences and occurs in Utah. 2 Uvanite, a brownish-yellow uranium vanadate (2U0 3 .- J. M. Boutwell, Bull. 260, U. S. Geol. Survey, 1905, p. 205. H. S. Gale (Carnotite in Colorado), Bull. 340, U. S. Geol. Survey, 1908; Idem, Bull. 315, 1906, pp. 110-117. F. L. Hess, Vanadium deposits in Colorado, Utah, and New Mexico, Bull 530, U. S. Geol. Survey, 1912. Idem, Mineral Resources, Annual issue, 1912, pp. 1003-1036. F. L. Hess, A hypothesis for the origin of the carnotites, Econ. Geol., vol. 9, 1914, pp. 675-688. R. B. Moore and K. L. Kitthil, A preliminary report on uranium, radium and vanadium, Bull. 70, U. S. Bureau of Mines, 1913. i K. L. Kitthil and John A. Davis, Mining and concentration of carnotite ore, Bull 103, U. S. Bureau of Mines, 1917. Parsons, Moore, Lind and Schaefer, Extraction and recovery of radium, uranium and vanadium, etc., Bull. 104, U. S. Bureau of Mines, 1915. 1 W. F. Hillebrand, H. E. Merwin and F. E. Wright, Proc., Am. Philos. Soc., vol. 53, 1914, pp. 31-54. 2 F. L. Hess and W. T. Schaller, Jour., Washington Acad. Sci., vol. 4, pp. 576-579. CONCENTRATIONS FROM SURROUNDING ROCKS 409 3V 2 5 .15H 2 0) with 39.60 per cent. UO 3 and 37.70 per cent. V 2 8 is mined in Emery County, Utah. Volborthite and calcio-volborthite both vanadates of copper, and a uranium sulphate have also been identified. Roscoelite, a dark green vanadium mica is abundant as veins and as replacements in the cement of some sandstones (Fig. 129) at Placerville, Colorado. About two-thirds of the aluminum is replaced by vanadium so that it contains from 20 to 29 per cent. V 2 O3. Some sandstones contain as much as 20 per cent, of roscoelite. 1 Fia. 129. Vanadium ores in sandstone. White areas, quartz; shaded areas, roscoelite, partly radial. Magnified 25 diameters. After F. L. Hess, U. S. Geol. Survey. Chromium is also present in these remarkable ores, probably as mariposite, or chromium mica. Barium, copper, lead, molyb- denum and arsenic are contained in the ores as shown by the analysis of carnotite. Molybdenum is quite abundant in some places and appears to occur as a soluble sulphate the dark blue ilsemannite. A similar occurrence of the latter mineral is re- ported from South Africa. Native selenium has been identified by Hillebrand, associated with metahewettite. Some of these minerals are quite certainly secondary products. 1 Roscoelite is also known from some gold-quartz veins and is often intimately associated with native gold. Mariposite is a characteristic mineral of many gold quartz veins. 410 MINERAL DEPOSITS The carnotite, which is the most abundant uranium mineral, contains a small trace of radium 1 which is recovered. Gypsum is about the only gangue mineral associated with the ores. The ores are not rich. The carnotite ores contain about 1.5 to 3 per cent. U 6 8 and 3 to 5 per cent. V 2 3 . Concentration has been attempted in some cases. The roscoelite ore at Placerville contains about 3.50 per cent. V 2 3 and 0.05 per cent. U0 3 . Occurrence. The ores are found in the Plateau province of horizontal or gently inclined strata in southwestern Colorado and eastern Utah. The best known localities are at Placerville, Col- Apparent nconformity Seam . Coarse . .'sandstone.' j? Feet FIG. 130. Sketch of vanadium-bearing sandstone at mine of Primes Chemical Company, on the east side of Bear Creek, Newmire, Colo. After F. L. Hess, U. S. Geol. Survey. orado, and in the LaSal, Paradox and Sindbad valleys somewhat farther west. The ores are mainly confined to the McElmo and LaPlata formations of white, often cross bedded Jurassic sandstone which frequently contain much transported partly carbonized wood. They follow certain horizons or appear in fissures of flat veins or in brecciated zones (Fig. 130) and are often associated with the fossil wood. Some observers have thought the ores merely superficial but it now seems certain that they may be found in depth. At Placerville the workings are said to have penetrated 2,000 feet under ground in horizontal direction. 1 At the rate of one gram of radium in 3, 000 kilograms of metallic uranium. CONCENTRATIONS FROM SURROUNDING ROCKS 411 Not similar to those deposits, yet perhaps of a similar origin, are the recently discovered important vanadium deposits' at Minasragra, 1 near Quisque, Province of Pasco, Peru, described by D. F. Hewett and W. F. Hillebrand. The vanadium sulphide, . patronite, occurs here on a large scale as greenish-black masses associated with a hydrocarbon and a peculiar nickel-bearing py- rite. An analysis of the patronite gave 58.79 per cent, sulphur, 19.53 per cent, vanadium, 0.18 per cent, molybdenum, 1.87 per cent, nickel, and 3.47 per cent, carbon. The deposit is said to be a vein with much bitumen and clay in gently dipping Cretaceous strata. Igneous rocks are present in abundance. Genesis. Traces of vanadium are found, according to Hille- brand, in most igneous rocks, and some varieties of augites carry as much as 2 per cent, of the metal. Titanic iron ores usually contain a fraction of a percent. Smaller quantities occur as oxidation products in many ore deposits, mostly as vanadinite or descloizite. Vanadium tends to concentrate in clays and shales; it is also concentrated in coal; the ashes of many varieties are rich in vanadium. 2 A coal from the copper deposits in sandstone of Red Gulch, Colorado, contained, according to Hillebrand, 3 0.18 per cent, vanadium. The processes of weather- ing and vegetation appear to be favorable to its concentration; and, to some extent, uranium shares in this behavior. F. L. Hess believes that adjoining land areas may have con- tained uranium- and vanadium-bearing veins at a certain level which would account for the ores being practically confined to one stratigraphic horizon, but such an assumption is scarcely sup- ported by any facts. At present the localization of the ores can- not be satisfactorily explained. The deposits are probably products of concentration, by meteoric waters, of small quantities of the metals distributed in h'ttoral beds or in land deposits and derived from older deposits of some kind in ancient land areas of igneous or metamorphic rocks. Production and Use. In 1915, 47,000 tons of ore were mined in Colorado containing, according to U. S. Geological Survey, 1 D. F. Hewett, Vanadium deposits of Peru, Trans., Am. Inst. Min. Eng., vol. 40, 1909, pp. 274-299. W. F. Hillebrand, The vanadium sulphide, patronite, Jour., Am. Chem. Soc., vol. 29, 1907. W. F. Hillebrand, Am. Jour. Sci., 4th ser., vol. 24, 1907, p. 141. 2 F. W. Clarke, Geochemistry, Butt. 616, U. S. Geol. Survey, 1916, p. 705. Bull. 340, U. S. Geol. Survey, 1908, p. 172. 412 MINERAL DEPOSITS 19.9 tons of metallic uranium, 627 tons of vanadium (mainly from roscoelite ores) and 6.1 grams of radium, the value of the metals being about $700,000. Some years ago most of the uranium ores were exported and radium extracted abroad. In 1914, the ores mined contained 87.2 tons of uranium and 22.3 grams of radium. The pure vanadium ores of Placerville are roasted with sodium chloride, the resulting sodium vanadate extracted with water and precipitated with ferrous sulphate as iron vanadate which is shipped east for reduction to ferro- vanadium. About 1 per cent, vanadium added to steel increases its toughness and resistance to torsion and high tem- perature. It is, however, less essential to the steel industry than tungsten, and the principal supply is obtained from the Peruvian patronite mine. The value of ferro-vanadium alloy is about $1,000 per ton. Minor amounts of vanadium salts are used as mordants for dyeing and cloth printing and for other chemical purposes. Various ore deposits yield small quantities of vanadium ores such as vanadinite and descloizite. In the car- notite ores there is difficulty in separating uranium from vana- dium and only a small price is paid for the latter. Uranium salts have a limited use for a yellowish-green glass and for pottery; also as a mordant in dyeing. Ferro-uranium is at present not used in the stc-el industry. Radium is separated from uranium by a complicated process explained in Bull. 104, U. S. Bureau of Mines. It is produced as a chloride or bro- mide and its principal use is in medical science,, various diseases yielding more or less to its emanations. It is said that it can be extracted at a cost of $37 per milligram. Radium in corresponding quantities is also contained in urani- nite (crystalline) and pitchblende (amorphous), both essentially UOz.UOs with 80 + per cent, uranium oxides. 1 It is obtained in small quantities from gold-pyrite veins of Gilpin County, Cali- fornia, from cobalt arsenide veins of Joachimsthal, Bohemia, and from tin veins of Cornwall, all of which are deposits formed at higher temperatures. Various uranium minerals also occur in pegmatite dikes. !. S. Bastin and J. M. Hill, Prof. Paper 94, U. S. Geol. Survey, 1917, pp. 121-128. CONCENTRATIONS FROM SURROUNDING ROCKS 413 THE COPPER-BEARING SHALES OF MANSFELD' It seems proper to consider at this place the celebrated cu- priferous shale (Kupferschiefer) of Mansfeld, in central Germany, for, though not identical with the deposits described in this chapter, it presents most interesting analogies to them. A flourishing mining industry is still based on the Kupfer- schiefer, the annual production being approximately 700,000 metric tons of ore containing between 2 and 3 per cent, of copper. The first stratum deposited in the subsiding basin of the upper Permian in central Germany was a marine conglomerate of slight thickness. ''Above it extends like a black shroud the thin bed of cupriferous shale, one of the most remarkable products Prospecting . Shaft /^n^v * FIG. 131. Section of a part of the copper-bearing shale bed at Mansfeld, Germany. After Schroder. of the geologic ages. Characterized by its fauna as a shallow-sea deposit, full of plant remains carried in from adjacent coasts, the formation bears the stamp of an organic mud deposit admixed with organic precipitates. 2 Above the cupriferous shale, which is less than 1 meter thick, extends a marine limestone (Zechstein), 6 to 10 meters thick, and above that lie the great gypsum and salt masses of the upper- most Permian. Folding and faulting have since affected the beds, and the mining now follows the inclined strata to a depth of 500 meters. The cupriferous bed averages 50 centimeters in thickness, but only the lower 20 to 30 centimeters is utilized as ore (Fig. 131). 1 Best description with literature in Stelzner and Bergeat, Die Erzlager- statten, vol. 1, 1904, pp. 391-417. 2 F. Beyschlag, in Deutschlands Kalibergbau, 1907, p. 4. 414 MINERAL DEPOSITS The ores are sulphides, in minute distribution through the shale, giving it a bronzy appearance. Chalcopyrite predominates, but there are also bornite, pyrite, chalcocite, and rarely galena and tetrahedrite. Small quantities of nickel, cobalt, selenium, vanadium, and molybdenum have been recognized; there is also about 0.015 per cent, of silver. Zinc is present, and in the upper part of the bed, not mined, there is more zinc than copper. An average analysis is as follows: 1 AVERAGE ANALYSIS OF CUPRIFEROUS SHALE Dr. Haase, Analyst Per cent. Per cent. SiO 2 33.15 Ag 0.014 A1 2 O 3 17.3 Ni 0.018 CaO 10.4 S 2.31 MgO 1.0 CO 2 9.24 Fe 2.6 H 2 1.7 Zn 1.276 Bitumen 9.06 Cu 2.75 There is about 3 per cent. K 2 O and 1 per cent. Na 2 0; lead amounts to 1.50 per cent., manganese to about 0.25 per cent. There is practically no gangue, except veinlets of gypsum and barite. The bed is cut by faults, along which in places occur barite, anhydrite, calcite, niccolite, pyrite, and chalcopyrite, and near these breaks (called "Riicken") the metal content is subject to enrichment, impoverishment, or removal upward in adjacent beds. Bergeat declares that these changes take place on secondary fissures and cracks. There has been much controversy about the Mansfeld deposits. The majority of geologists regard them as sedimentary and syngenetic: von Groddeck, Stelzner, Freiesleben, and von Cotta held this view, and it is shared by Bergeat. Posepny and Beck believe them epigenetic and think that the metals were probably introduced in the shale from the Rucken. The Kupferschiefer is certainly not an ordinary marine deposit precipitated from the sea water. 2 It was laid down in a shallow sea which was full of decaying vegetable and animal remains and into which were probably discharged cupriferous 1 Stelzner and Bergeat, Die Erzlagerstatten, vol. 1, 1904, p. 396. 2 Sea water contains a trace of copper, as shown by Dieulafait (Ann. chim. phys., 5th ser., vol. 18, 1879, p. 359; C. R., 90, p. 1573; 96, p. 70; 101, p. 1297) and others, but the amount present seems utterly insufficient to account for the Mansfeld deposits. CONCENTRATIONS FROM SURRO UNDING ROCKS 415 waters from the surrounding littoral, most likely sulphate solu- tions derived from the eruptives and the ore deposits of the early Permian epochs. No one can read the description of the great uniformity of distribution without being impressed with the very strong arguments for a syngenetic origin. The characteristic presence of nickel, cobalt, vanadium, and selenium recalls the epigenetic deposits in sandstone so abundant around the shores of the Permian sea, in Bohemia and Russia, for instance. The Mansfeld basin was simply, then, the final collecting place of the solutions derived from adjacent desert shores. COPPER SULPHIDE VEINS IN BASIC LAVAS General Features. All basic lavas contain copper, but in many cases conditions were evidently unfavorable for the con- centration of copper immediately after the eruption, and the rocks retained their copper until later opportunities for ore forma- tion were offered. The existence of vast masses of such basic lavas near the surface, without any indication of copper con- centration (e.g., the Columbia River lava or the basalts of the Hawaiian volcanoes), shows plainly that the ordinary surface waters at slight depth are not competent to dissolve and concen- trate accessory metals contained in these rocks. A depth of perhaps a few thousand feet seems to be necessary, under the most favorable conditions, for waters of meteoric origin to ex- tract the copper; though it is, of course, possible that such waters, when ascending in suitable channels, may deposit the dissolved copper at higher horizons. In some of the veins here discussed epidote is present, but more frequently it is absent, and the veins then assume the well-known type of chalcopyrite in a quartz- calcite-siderite gangue. Such veins, deposited by ascending surface waters of the deeper circulation, may not be easy to distinguish from those whose development is a phase of intrusive after-effects. Nor can it be denied that in these veins may be concentrated some gold and silver from the igneous rock; in general, however, they will be found much poorer in gold and silver than deposits connected with the intrusive processes. jWhether native copper, bornite, or chalcopyrite will form seems to be dependent upon the quantity of sulphur which the lavas were able to retain at their eruption. 416 MINERAL DEPOSITS The Nikolai Greenstone. The copper deposits in the Nikolai greenstones of the Copper River region, described by F. C. Schrader, W. C. Mendenhall, A. C. Spencer, and lately again by F. H. Moffit, 1 are of special interest. Flows of Triassic or Car- boniferous' basalts about 4,000 feet in thickness are covered by 2,000 feet of Triassic limestone, which in turn is overlain by a thick series of Jurassic strata. The latter are cut by monzonitic intrusives, which are accompanied by a different kind of metalli- zation characterized by prominent gold deposits. The Nikolai greenstones are amygdaloid flows of basalt; the amygdules contain scarcely any zeolites, but are filled with chlorite, chalcedony, and quartz and carry no copper. Copper sulphides are extremely common in the flows, but occur in slips, brecciated zones, and faults. The minerals are chalcopyrite, pyrite, and bornite, with calcite and a little quartz; there is some epidote, not always present. One of the fissure zones extends up into the Triassic limestone above the greenstone. In the latter a little bornite and chalco- cite appears and the zone cuts across an older series of quartz- epidote veins carrying the same two sulphides with a little native copper. In the limestone the fissure zone develops into the remarkable and valuable deposit worked in the Bonanza mine. It is an almost solid body of massive chalcocite with conchoidal fracture, traced for 400 feet and with a greatest width of 25 feet; its depth is apparently limited. There are no gangue minerals and the limestone adjoining the chalcocite is not altered. No intrusive rocks are present. It is probable that the ores characteristic of the Nikolai green- stones are derived from the rock itself. The copper deposits seemed to be formed mainly after the Triassic limestone had been laid down, and it is likely that meteoric waters did the work. The waters must have descended through limestones and shales in which they would have acquired chlorides, sulphates, car- bonates, carbon dioxide, and hydrogen sulphide, and they would therefore be competent to dissolve copper from the greenstones which they traversed. The chalcocite alters superficially to covellite and copper carbonates; there is no evidence that it has replaced pyrite and it may have been deposited JD its present form. 1 F. H. Moffit and S. R. Capps, Bull. 448, U. S. Geol. Survey, 1911. F. H. Moffit, Bull 662, U. S. Geol. Survey. 1917. Dp. 155-182. CONCENTRATIONS FROM SURROUNDING ROCKS 417 C. F. Tolman 1 has recently found remnants and structures indicating bornite and believes that the chalcocite is secondary and mainly derived from bornite. The final word about this interesting deposit has not been said. COPPER SULPHIDE VEINS IN INTRUSIVE BASIC ROCKS Veins containing pyrite and chalcopyrite, occasionally with other sulphides, in a gangue of quartz, calcite, dolomite, and siderite, more commonly quartz alone, occur abundantly in intrusive basic rocks, such as diabase or gabbro. Here, however, the distinction between the results effected by water of atmos- pheric origin and those effected by magmatic solutions becomes increasingly difficult. F. E. Wright 2 has pointed out the fact that the intrusive Keweenawan gabbro of Mount Bohemia contains veins with chalcopyrite, bornite, chalcocite, calcite, and quartz, while in the surface lavas of the same series native copper is the principal ore mineral. This seems an excellent illustration of the reten- tion of volatile sulphur by intrusives, contrasted to its escape from the extrusive flows. The origin of the water which was the vehicle of deposition in these veins may remain an open question. Along the foot-hills of the Sierra Nevada of California extends a belt of andesitic rocks of Jurassic age collectively called "green- stones." They consist of massive and schistose rocks, including surface flows, tuffs, and intrusions mixed. Within this belt, for instance in Yuba and Nevada counties, short and irregular quartz veins with pyrite and chalcopyrite are common. Proba- bly these veins derived their copper from the greenstones, and undoubtedly they were formed at a time when the rocks now exposed were covered by several thousand feet of overlying and now eroded igneous rocks. Other deposits, such as those at the Dairy Farm in Placer County and at Campo Seco in Calaveras County, are, according to A. Knopf, 3 replacement deposits along shear zones in the same belt of amphibolites and other greenstones. The minerals are pyrite and chalcopyrite, with a trifle of galena and zinc blende, 1 Trans., Am. Inst. Min. Eng., vol. 54, 1917, pp. 402-441. 2 F. E. Wright, The intrusive rocks of Mount Bohemia, Michigan, Seventh Ann. Rept., Geol. Survey Michigan, 1908. 3 A. Knopf, Notes on the foot-hill copper belt of the Sierra Nevada, Bull. 17, California Univ. Dept. Geology, vol. 4, 1906, pp. 411-421. 418 MINERAL DEPOSITS associated with quartz, calcite, epidote, chalcedony, and some- times zeolites. The sulphides contain a little silver and a trace of gold. Similar are the so-called "iron belts" of pyrite and chalcopyrite in the Ophir mining district of gold-silver quartz veins in Placer County. 1 They are contained in amphibolite, but do not always extend parallel to the schistosity; the width varies from a few feet to 400 feet, the length is in places half a mile. No shear planes are visible along these zones. They enrich the distinctly later gold-quartz veins which cross them. The sul- phides are associated with the amphibole, epidote, feldspar, and quartz of the amphibolites and are often intergrown with mag- netite or ilmenite. In the paper cited they were interpreted as concentrations of copper from the surrounding rocks, formed by chemical action during the progress of the dynamic meta- morphism which produced amphibolites from primary diabasic rocks. Other and much larger copper deposits are found in the same region, on the north in Shasta County and on the south in Cala- veras County, but at both places the evidence points clearly to an origin by solutions derived directly from the magma. In the Encampment district, Wyoming, A. C. Spencer 2 studied deposits of primary chalcopyrite enriched by chalco- citization, and probably of pre-Cambrian age. The metalliza- tion is localized in shattered zones in quartzite, or between quartzite and schist, close to intrusive gabbro or diorite which contains copper (p. 9), sometimes visible as chalcopyrite. .Spencer gives several good structural reasons indicating that the depositing waters were ascending and believes that the copper was leached from the cupriferous gabbro. The minerals present do not indicate especially high temperatures. Although the deposits were formed at considerable depth, as shown by the flexing of the schist bands, the quartzite was decidedly in its zone of fracture. OTHER VEINS DEPOSITED BY WATERS OF THE UPPER CIRCULATION In the preceding pages it has often been pointed out that the competency of the circulation of certain kinds of atmospheric 1 W. Lindgren, Fourteenth Ann. Rept., U. S. Geol. Survey, 1895, pp. 262-264. 2 The copper deposits of the Encampment district, Wyoming, Prof. Paper 25, U. S. Geol. Survey, 1904. CONCENTRATIONS FROM SURROUNDING ROCKS 419 waters to form many mineral deposits cannot be questioned and that it may be difficult or impossible to determine the origin of certain occurrences. Nevertheless, the fact stands firm that surface waters of the ordinary type, even in slightly heated ascending currents, do not form mineral deposits even in localities where the condi- tions are such that they might be expected to do so, as in the Alps, for instance. There are, however, other localities, particu- larly in the region of the saline Paleozoic and Mesozoic beds of central Germany, where such deposition appears to have taken place. Veins of this origin are likely to contain an abundant gangue of calcite, dolomite, or barite; with some quartz and a scant amount of sulphides. 1 The sweeping generalizations of F. Hornung and his interpre- tation of all the mineral veins of the Harz Mountains as being formed by inter- Permian brines cannot be accepted, but it is not improbable that he is correct in believing that many barite and hematite veins have had this origin. 2 In connection with this K. Ochsenius 3 showed that solutions containing 2.59 per cent. NaCl, 3.16 per cent. MgCl 2 , and 1.85 per cent. MgSO 4 decom- pose chalcopyrite and chalcocite at room temperature. This action is slow and is noticeable only after several years. Galena was not dissolved. Similar examples of ore deposition by saline waters also exist in the western part of the United States; one, the Cashin mine of Colorado, is mentioned above (p. 404). The prevailing influence of igneous intrusions on ore deposition is, however, so strong that it is difficult to establish the proofs of the less con- spicuous deposition by purely meteoric water. That the ordinary surface waters are in most cases quite incom- petent to effect concentration is plainly shown by the lack of important mineralization in fissures and joints cutting the rocks of mining districts. In the Globe district, Arizona, for instance, the Paleozoic rocks are intersected by a network of dislocations which would offer excellent paths for these waters; and yet the important deposits are in no way connected with these fractures. 1 P. Krusch, Ueber die Zusammensetzung der Westfalischen Spalten- wasser, Zeitschr. prakt. Geol, vol. 12, 1904, p. 252. 2 F. Hornung, Ursprang und Alter des Schwerspates und der Erze im Harze, Zeitschr. Deutsch. geol. Gesell., vol. 57, 1905, pp. 291-360. 3 Idem, p. 567. 420 MINERAL DEPOSITS Similarly "cross courses" often fault the gold-quartz veins of California and yet they are, as a rule, absolutely barren, often open fissures. Similar post-mineral fissures traverse lead and zinc veins in the Coeur d' Alene district, Idaho, but generally show no trace of mineralization. The Cordilleran region contains many great ranges of pre-Cam- brian rocks capped in places by Paleozoic and Mesozoic strata. Among them may be mentioned the Front Ranges of the Rocky Mountains in Colorado, the Wind River Range in Wyoming, and the Mission Range in Montana. Uplift, folding, and faulting have in each of these ranges intensified the circulation of me- teoric waters, but in spite of this the ranges are remarkably poor in mineral deposits, which appear only in the vicinity of later intrusives. These relations show very plainly the slight concentrating power of ordinary cool surface waters and even of the waters of atmospheric origin that have become a part of the deeper circulation. CHAPTER XXI DEPOSITS RESULTING FROM REGIONAL METAMOR- PHISM Rocks subjected to stress at moderate depths within the zone" of fracture may rupture in closely spaced breaks, producing the appearance of a schistose structure. In such rocks no great chemical changes would occur, except perhaps by subsequent deposition along the tight fissures. At greater depth deformation may take place by granulation and recrystallization, accompanied by chemical changes that are effected by the aid of the scant rock moisture. Uralite and chlorite may form from pyroxene, the soda-lime feldspars may recrystallize to zoisite and albite, the quartz crystals may be crushed and elongated, new mica, particularly muscovite, may develop; also crystals of aluminum garnet. The chemical composition of the rock will, however, change but little; although the various transformations involve transportation of substance, this movement is not free, but is limited and hindered in all directions. Under these circumstances it is improbable that processes of concentration could have much opportunity to assert themselves; the minute quantities of useful metals contained in the original rock could not easily assemble to form larger masses. In amphibolite schist small grains of chalcopyrite, often intergrown with pyrrhotite and magnetite, appear to be more common than in the primary igneous rock from which the schist was derived. If even the slightest and slowest circulation of water was established during the deformation, some concentra- tion of chalcopyrite could well take place, as it does in fissures traversing similar rocks. When the deformation takes place at higher temperatures a number of minerals are developed which are similar or identical with those of contact metamorphism. It is often difficult, indeed, to draw the line between regional and igneous metamorphism, especially in intensely metamorphic regions where intrusive masses are abundant. There is reason to believe that at tem- peratures of, say, several hundred degrees some of the rocks, par- 421 422 MINERAL DEPOSITS ticularly limestones, become permeable to the gaseous emanations of water and metallic compounds yielded by intrusive masses, and thus an opportunity is afforded for the introduction of new substances which in places may become concentrated into ore deposits. To such a permeation in the deep zone of anamor- phism many of the most enigmatic ore deposits of the crystalline schists may owe their origin. These deposits would then differ in some respects from the ordinary contact-metamorphic ores, which have, as a rule, developed only close to intrusive contacts, in most cases also actually within the zone of fracture. Dissemination of sulphides is a phenomenon often encountered in almost any area of crystalline schists. In the majority of occurrences pyrite, pyrrhotite, and chalcopyrite are prominent; the sulphides of lead and zinc are far less common. Such dis- seminations are also particularly connected with amphibolitic or chloritic rocks. As indicated above, these ore minerals may have various modes of origin. In the first place the dissemination may be caused by mineralization along both sides of a fissure, parallel with the schistosity that is, by the formation of a "bedded vein." Such mineralization is later than metamor- phism, and the metamorphic minerals will probably be found to be altered sericitized, carbonatized, or more rarely silicified. If, on the other hand, the sulphide minerals were contained in the rock previous to metamorphism, or if they were devel- oped during that process, they will be found intergrown with the metamorphic minerals, such as amphibole, epidote, chlorite, garnet, and albite, and are usually accompanied by some mag- netite or ilmenite. Larger pyritic masses of this kind are, in most cases, probably original products of magmatic concentration; or they may be old fissure veins or replacement veins which have been rendered unrecognizable by deformation; or, finally, they may be of con- tact-metamorphic origin. Sparser disseminations, often following certain lines along the strike of the schist, are often called "fahlbands" (the German "fahl" meaning rusty brown and referring to the oxidized out- crops). Such fahlbands, first noted in Kongsberg, Norway, 1 where they enrich the silver veins, may be several miles long and vary in thickness between a fraction of a foot and several hundred 1 C. A. Miinster, ref. in Zeitschr. prakt. Geol, 1896, p. 93. J. H. L. Vogt, idem, 1899, pp. 177-181. DEPOSITS FROM REGIONAL METAMORPHISM 423 feet. The enclosing rocks vary from gneiss to mica schist, diorite, and amphibolite. The ore minerals are pyrite, pyrrho- tite, zinc blende, chalcopyrite, molybdenite, and sometimes cobalt minerals. They are often intergrown with amphibole or garnet. The fahlbands are rarely of. economic importance, but many of them characteristically enrich intersecting veins, causing native silver and gold as well as cobalt and nickel, ores to appear at the intersections. This is probably only a special case of the general law that veins are enriched where they cut across belts of pyritic impregnation. Fahlbands rich in cobaltite, with pyrite, chalcopyrite, pyrrhotite, and molybdenite, were worked at Skutterud and Snarum, in the Modum parish, Norway, from 1776 to 1899. For a long time these deposits were among the principal sources of cobalt oxide, which is used to impart a deep blue color to glass and porcelain. According to the older litera- ture quoted by Stelzner and Bergeat 1 the fahlbands at Skutterud lie between gneiss or quartz schist and amphibolite. Other minerals mentioned are malacolite, antophyllite, and rarely graphite and tourmaline. The ores were poor, containing, even when sorted, less than 1 per cent, cobalt. A parallel belt at Snarum is said to be enclosed in amphibolite and contains more copper. The fahlbands have been variously interpreted. At a time when the crystalline schists were generally considered as altered sediments, they were held to be sedimentary deposits. Ball and Kjerulf, 2 in 1880, held them to be impregnations related to gabbro intrusions. Vogt considered the gray gneiss of Kongs- berg as a pressed granite and held that it had been impregnated with sulphides at the same time as the surrounding schists. That the dissemination of sulphides in its present form is dependent upon dynamo-chemical metamorphism is clearly shown by the minerals with which the sulphides are now inter- grown. Sulphide emanations from intrusive magmas at a con- siderable distance from their source do not usually crystallize with amphibole, pyroxene, and garnet, but rather with calcite, sericite, and quartz as gangue minerals. Still, the recrystalliza- tion under pressure does not necessarily explain the ultimate origin of the minerals and it is probably hopeless to speculate on this subject until the metamorphic series at the location of typical 1 Die Erzlagerstatten, 1, 1904, pp. 269-271. 2 Die Geologie des siidlichen und mittleren Norwegens, 1880. 424 MINERAL DEPOSITS fahlbands has been more carefully examined as to the original character of its rocks. Somewhat similar fahlbands in amphibolite and gneiss are reported in the older literature from Schladming, in Styria, where they enrich intersecting cobalt-nickel veins, and from Les Challanches, in France, where similar relations exist. 1 Recent descriptions from both places show that the so-called fahlbands are in reality narrow veins accompanied by alteration of the wall rocks. 2 1 Stelzner and Bergeat, Die Erzlagerstatten, 1, 1904, pp. 268-269. 2 C. Schmidt and J. H. Verloop (Schladming), Zeitschr. prakt. Geol, vol. 17, 1909, pp. 271-276. T. A. Rickard (Challanches), Trans. Am. Inst. Min. Eng., vol. 24, 1894, pp. 689-705. CHAPTER XXII DEPOSITS OF NATIVE COPPER WITH ZEOLITES IN BASIC LAVAS GENERAL STATEMENT Native copper, chalcocite, bornite, much more rarely chalco- pyrite, and their products of oxidation are often found in flows of basic lavas, particularly in basalts, associated with minerals of the zeolite group, such as analcite, natrolite, stilbite, chabazite, and laumontite, and the minerals prehnite and datolite; together with these calcite, quartz, chalcedony, chlorite, epidote, and adularia may be present, sometimes in predominating quantity. These gangue minerals, together with the copper minerals, fill vacuoles or blowholes in the basic rocks or replace the rock. Pyrite and sulphides of metals other than copper rarely occur. These deposits have been formed near the surface under condi- tions which will be discussed in a following paragraph. The mineral association does not indicate a deep-seated origin. Instances of native copper occurring in this manner are plenti- ful, though the occurrences are not always of economic impor- tance. Among the numerous localities the following may be mentioned: The Faeroer, 1 north of Scotland; Sterling, 2 in Scot- land; Oberstein a. d. Nahe, Germany; Sao Paulo, 3 Brazil; the Kristiania region, 4 Norway; the Triassic "traps" of New Jersey 5 and Connecticut; New Guinea; 6 the Transbaikalian prov- inces 7 on the Dochida River; the Bay of Fundy, 8 Nova Scotia. 1 F. Cornu, Zeitschr. prakt. Geol, vol. 15, 1907, p. 321. 2 Carl Hintze, Handbuch der Mineralogie, 1898. 3 E. Hussak, Centralblatt f. Min., 1906, p. 333. (No zeolites; copper between the peripheral covering of the amygdules, consisting of an iron silicate and the filling of chalcedony.) 4 J. H. L. Vogt, Zeitschr. prakt. Geol, vol. 7, 1899, p. 12. 8 Volney Lewis, Ann. Rept., Geol. Survey New Jersey, 1907, pp. 157 and 165. R. Beck, Lehre von den Erzlagerstatten, I, 1909, p. 345. 7 Idem, p. 346; also Zeitschr. prakt. Geol, vol. 9, 1901, p. 391. (With opal, chalcedony, calcite, epidote, and prehnite.) 8 R. W. Ells, Copper in the provinces of Nova Scotia, New Brunswick, and Quebec, Min. Res. Can., Geol. Survey Canada, 1904, 58 pp. 425 426 MINERAL DEPOSITS The first six of these occurrences have not been worked as copper deposits. The last four are of some economic importance. A pre-Cambrian series of basaltic rocks in the Lake Superior region contains the most prominent example of this class of deposits in the world, which will be described in more detail below. In eastern Oregon, 1 about 20 miles east of Baker City, and along the Snake River canon, opposite the Seven Devils Moun- tains in Idaho, are extensive areas covered by a basaltic amygda- loid flow of Jurassic or Triassic age. This rock contains native copper and chalcocite, sparsely disseminated or along obscure fracture zones, in association with epidote, chlorite, calcite, and zeolites. The ores are of low grade and have not yet been worked with profit. Another occurrence of interest in the White River region in Alaska has recently been described by Adolph Knopf. 2 The ba- saltic amygdaloids, with tuffs and breccias, are interbedded with sediments of Carboniferous age and have probably been erupted in part under submarine conditions. Placer copper is common in the creeks, and some large masses have been found. The copper minerals are chalcocite, chalcopyrite, and native metal in stringers and seams, with prehnite, laumontite, thomsonite, and calcite; also native copper with zeolites filling blowholes in reddish, highly amygdaloid lava. These statements will serve to show that the zeolitic copper deposits in basaltic lavas represent a type of world-wide distribu- tion; the same processes of concentration are evidently applicable to all cases. ORIGIN OF THE ZEOLITIC COPPER ORES Probable Source of Copper. Basic igneous rocks such as gab- bro, diabase, basalt, some andesites, and basaltic flows designated melaphyres or amygdaloids probably always contain copper, in some cases as much as 0.1 or 0.2 per cent., more commonly about 0.02 per cent, of the metal (p. 8). According to Volney Lewis and F. F. Grout, the copper is present as a silicate, pos- sibly in part as a sulphide, such as bornite or chalcocite. It is likely that the copper is present mainly as a silicate 1 W. Landgren, The gold belt of the Blue Mountains of Oregon, Twenty- second Ann. RepL, IT. S. Geol. Survey, pt. 2, 1901, pp. 551-776. * Ec2 would escape and the remaining H 2 S precipitated the metals as sulphides. The chlorides of lead and zinc are far more soluble than the bicarbonates and strong brines of sodium chloride are undoubt- edly effective in the transportation of the metals. R. C. Wells found that weak salt solutions decomposed but little zinc sul- phide. 1 Stronger solutions might be more active. The theory explained is then based on the leaching of lead and zinc occur- ring as minutely disseminated sulphides in limestone and shale. Moresnet. 2 The Moresnet district in Belgium, Luxembourg, and Prussia, is situated in a region of folded Devonian and Car- boniferous limestones and slates cut by several large faults and covered unconformably by Cretaceous beds. In the main the ore follows these dislocations, in part as filled veins, in part as large replacement deposits in limestone at the slate contacts or at the intersection of faults. Dolomitization of the limestone is often mentioned. The ore contains zinc blende, galena, iron sulphides, and calcite, and the galena and zinc blende are often intimately intergrown. Nickel is occasionally present. Masses of calamine appeared near the surface and extended to depths of 160 feet; in some cases, notably at Vieille Montagne, they were of enormous size and reached a depth of 330 feet; the sul- phides appeared at depths of 170 to 330 feet, much below the water level. 3 Galena is in general the oldest, pyrite or marcasite the youngest of the minerals; concentric intergrowths, of wurt- zite, zinc blende, and galena (schalenblende) are not uncommon. The quantity of ore is said to diminish in depth, and large amounts of water are found. A considerable part of the world's production of zinc has been obtained from these deposits. Silesia. 4 Silesia, a province of Prussia, remains one of the world's most important zinc-producing regions. The ore occurs in Triassic sandstone and limestone, which lie in flat syn- 1 Bull. 606, U. S. Geol. Survey, 1915, p. 58. 2 Ch. Timmerhans, Les gttes m6talliferes de la region de Moresnet, Liege, 1905, p. 28. F. Klockmann, Die Erzlagerstatten der Gegend von Aachen, Berlin, 1910. See also text -books of Stelzner and Bergeat and R. Beck. 3 The oxidation of this deposit may be of pre-Cretaceous age. 4 G. Gurich, Zur Genesis der oberschlesischen Erzlagerstatten, Zetischr. prakt. Geol, 1903, pp. 202-205. A. Sachs, Die Bildung der schlesischen Erzlagerstatten, Centralblattf. Min., 1904, pp. 40-49; Zeitschr. Deutsch. geol. Gesell., vol. 56, 1904, pp. 269-272 See also text-books of Stelzner and Bergeat and R. Beck. LEAD AND ZINC DEPOSITS 449 clines (Fig. 135). In the lower part of the " Muschelkalk " extensive dolomitization has taken place, mainly along fissures, and the dolomite is underlain by an impermeable " Sohlenstein " or clay rock. The ground-water circulation, dolomitization, and mineralization are all apparently closely connected. The replacement ore occurs along two horizons the lower in a bed of dolomite several meters thick, carrying galena, zinc blende, and abundant marcasite, and the upper in a bed of smithsonite (zinc carbonate) of considerable thickness. The smithsonite and calamine are considered products of oxidation. The zinc blende is in part intergrown with wurtzite. The galena contains a little silver (0.02 to 0.03 per cent.) ; manganese as psilomelane is sometimes present. The ores are said to contain on the average about 17 per cent, zinc and 5 per cent. lead. The marcasite contains a little arsenic and a trace of nickel. The succession is marcasite (oldest), zinc blende and galena. x p.st. FIG. 135. Section through the synclines of Tarnowitz and Beuthen, Silesia. P. St., Carboniferous; B, Triassic sandstone; s, Triassic limestone; Do, dolomite; b, galena deposits; z, zinc blende deposits; o, oxidized zinc ores; e, limonite; T, Tertiary beds; Dt, Quaternary beds. After Gurich. There has been much discussion concerning the origin of these ores. Beyschlag and Michael 2 have shown that some ore-bearing fissures descend into the Carboniferous and hence believe that ascending waters did the work; others, like Sachs, believe that the ores resulted from descending waters and that organic matter caused the precipitation. Alpine Trias. The Alpine Trias in Austria contains a number of deposits of this type. At Bleiberg, in Carinthia, the ores 1 E. Schulz, Geol. Rundschau, vol. 4, 1913, pp. 126-136 2 Beyschlag, Zeitschr. prakt. Geol, 1902, p. 143. Michael, Zeitschr. Deutsch. geol. Gesell., voL 56, 1904, Protocol, pp. 127-139. 450 MINERAL DEPOSITS occupy filled flats and gash veins; they consist of light-colored zinc blende and marcasite, with calcite and barite gangue, and a little anhydrite and fluorite, but no quartz. No silver, antimony, copper, or arsenic is present. At Raibl, made famous by Posepny 's investigations, 1 the ores form fillings and replacements along three dislocations. The minerals are sphalerite, occasionally with wurtzite, and galena, with a little marcasite and chalcopyrite, and their deposi- tion was accompanied by extensive dolomitization. Posepny describes stalactites of galena, pyrite, and zinc blende, but such occurrences are exceptional. Other European Localities.- The great deposits of Santander, Spain, are contained in Carboniferous limestone and are said to be replacements connected with dislocations. The light-yellow zinc blende from these deposits is famous. Some cinnabar is present. At Monteponi, 2 Sardinia, large "stocks" of galena with zinc blende and pyrite are contained in Paleozoic limestones. There is much dolomitization, and a little quartz and barite also occur. Cinnabar is reported and the ores contain silver in part. Igneous rocks are represented only by a diabase. The genetic relations of both of these deposits are as yet uncertain. The Lead -Zinc Ores of the Mississippi Valley. One of the most remarkable metallogenetic provinces characterized by lead and zinc ores extends over the valley of the Mississippi in the generally flat-lying limestones of the Paleozoic, ranging from the Ordovi- cian to the lower Carboniferous (Mississippian) inclusive. These ores are found in Arkansas, Missouri, Oklahoma, Kansas, Illinois, Wisconsin, and Iowa and reach eastward as far as western Virginia and Tennessee. The ores are mined on a large scale in comparatively few regions. Small deposits of lead and zinc are widely spread and are even found in Pennsylvania, New^York and Ontario. Igneous rocks are absent. There are, however, a few small deposits in southern Arkansas, Kentucky, and southern 1 F. Posepny, Jahrb. K. k. geol. Reichsanstalt, vol. 23, 1873, pp. 315-420. The genesis of ore deposits, Trans. Am. Inst. Min. Eng., vol. 23, 1894, pp. 197-369. An elaborate series of illustrations of this deposit was recently published by the Department of Agriculture of Austria. 2 R. Beck, Lehre von den Erzlagerstatten, vol. 2, 1909, p. 257. LEAD AND ZINC DEPOSITS ' 451 Illinois in which gold, silver, antimony, or fluorite is present and which appear to be genetically related to local intrusions of igneous rocks. The main characteristics of the predominating type are sufficiently described in the introduction to this chap- ter. In details they differ considerably. In point of production 1 the deposits in Missouri easily pre- dominate. The zinc-mining industry centers in the southwestern part of that State, about Joplin, and in 1917 yielded 132,730 short tons of spelter, Jof a value of about $27,000,000, making about 20 per cent, of the production of the United States. The lead mining in the southeastern part of the State in the same year produced 204,545 short tons of lead, to which should be added 29,611 tons from the Joplin region, making a total value of $40,000,000. This is 37 per cent, of the lead production of the United States. The ore mined is generally referred to as " dirt; " the concentrates are spoken of as "ore." The total quantity of crude ore raised annually in Missouri is now about 19,000,000 tons, consequently it is of low grade. Practically all of it is treated in concentrating works, to yield high-grade material suitable for the reduction plants. Southwestern Missouri. 2 The Joplin region includes, outside of Missouri, adjacent parts of Kansas]and Oklahoma. The prin- cipal camps are at Aurora, Granby, Webb City, Alba, Neck, Joplin, Galena, Badger, Quapaw, and Miami. In the early days lead was the only metal won, but since 1870 zinc ores have been mined and now predominate entirely. From the districts in Missouri, near Joplin, the quantity of lead recovered is about one-fourth as much as zinc. The yield of lead and zinc con- centrates from the crude ore averages, according to Siebenthal, about 3.7 per cent, corresponding to 1.9 per cent. zinc. The 1 C. E. Siebenthal and J. P. Dunlop, in Mines Report of Missouri, Mineral Resources, U. S. Geol. Survey. Annual publication. 2 W. P. Jenney, Trans. Am. Inst. Min. Eng., vol. 22, 1894. A. Winslow, Missouri Geol. Survey, vols. 6 and 7. 1895. E. Haworth, Relations between the Ozark uplift and ore deposits, Bull., Geol. Soc. Am., vol. 11, 1900, pp. 231-240. H. F. Bain (with C. R. Van Hise), Preliminary report on the lead and zinc deposits of the Ozark region, Twenty-second Ann. Rept. U. S. Geol. Survey, pt. 2, 1901. W. S. T. Smith and C. E. Siebenthal, U. S. Geol Atlas, Folici 148 (Joplin). E. R. Buckley and H. A. Buehler, The geology of the Granby area, Missouri Bur. Geol. and Mines, vol. 4, 1909. MINERAL DEPOSITS concentrates, averaging 58 per cent, zinc, are smelted in Kansas and Oklahoma. The districts are situated on the flanks of the Ozark uplift (Fig. 136). The ore occurs in the Boone formation and in rocks of Kinderhook age, both belonging to the lower Carboniferous (Mississippian) . The beds form a very flat anticline pitch- ing gradually northwest and are displaced slightly by the Seneca fault in Missouri and Oklahoma as well as by the Miami fault in Oklahoma and Kansas. All three structural features appear to be of importance in the ore deposition. The Boone formation con- tains much light-colored chert, especially in the Grand Falls chert member, which con- tains the so-called "sheet ground" deposits. The surface of the Boone contained numerous sink-holes and caves, perhaps also drainage channels, and over this "Karst" topography were deposited the sandstones and shales, in part carbonaceous, of the Coal Measures (Pennsylvanian) ; there is, then, an uncon- formity by erosion. Post-Carboniferous erosion has now removed much of these rocks, but near Joplin the Pennsylvanian shale still remains in many of the old de- pressions (Fig. 137). The succession of the ore minerals is given by Smith and Siebenthal as follows: Dolomite (oldest), chalcopyrite, galena, sphalerite, galena, chalcopyrite, marcasite, pyrite, calcite, barite, and marcasite, the whole series of course being seldom found in one locality. All the minerals are fre- quently well crystallized. There is general agreement among the investigators that the mineralization began by dolomitization, and Bain sees in it a result of the more rapid diffusion of the magnesia in the ore- forming solutions than of the zinc. The LEAD AND ZINC DEPOSITS 453 sphalerite occurs as crystals and grains in the secondary chert which forms the gangue of the ore, the primary chert containing no metasomatic sphalerite (Fig. 138). This secondary chert largely made up of cryptocrystalline or microcrystalline quartz, contains much organic matter with minutely disseminated sul- phides, 1 and is darker than the primary chert which antedated Cherokee formation: Shale, sandstone, and thin coal beds. Unconformity. Carterville formation : Shale and sandstone. Unconformity. Short Creek oolite member. Boone formation : Limestone with chert beds. Irand Falls chert member. FIG. 137. Generalized section for the Joplin district, Missouri. After Smith and Siebenthal, U. S. Geol. Survey. ore deposition. The dark chert is probably in part a replace- ment of limestone, in part, where cementing breccias, a silicified mud. 1 Cox, Dean and Gottschalb, Studies on the origin of Missouri cherts and zinc ores, Bull., School of Mines and Met., Nov., 1916. 454 MINERAL DEPOSITS A composite sample of the zinc concentrate representing 3,800 lots has the composition given below. 1 ANALYSIS OF CONCENTRATED ZINC BLENDE FROM THE JOPLIN REGION Zinc Cadmium Lead Iron Manganese Copper 58.260 0.304 0.700 2.230 0.010 0.049 Sulphur 30.720 Calcium carbonate .... 1 . 880 Magnesium carbonate. 0.850 Barium sulphate . 820 Silica... . 3.950 >.773 FIG. 138. Thin section of "black chert" showing matrix of fine-grained quartz with grains of zinc blende (shaded) and crystals of dolomite. Note quartz crystals developing in dolomite. Magnified 53 diameters. After Smith and Siebenthal, U. S. Geol. Survey. The galena contains only a trace of silver. The ores are found as irregular deposits in the "broken ground " near the surface and as a flat "blanket deposit" or "sheet ground" in a chert member of the Boone formation at depths of 150 to 300 feet. Below this horizon there are, as yet, un worked 1 W. G. Waring, The zinc ores of the Joplin district, Trans., Am. Inst. Min. Eng., vol. 57. 1918, pp. 657-670. Waring has also found thallium, indium, gallium and germanium in the flue dust and in the zinc metal. LEAD AND ZINC DEPOSITS 455 deposits of disseminated ore of doubtful value. The ore occurs mainly as fillings of cavities, the fillings of distinct veins or crevices being subordinate. The ore minerals with secondary chert fill spaces of brecciation or solution cavities along the stratification, perhaps also spaces of discission in limestone caused by stretch- ing and adjustments. In the "broken ground," which extends for 100 or 150 feet below the surface, the ores occur in clayey chert breccias in old sink holes filled with Pennsylvanian sediments, or along the out- side of such sink holes, forming "circles" where the slipping and settling provided open ground (Fig. 139). In these occur- "Circle"'with Broken Ground Pennsylvaniai! Shale FIG. 139. Diagram of zinc-lead deposits at Joplin showing "broken ground" around "circle" near surface and "sheet ground" deposit in Grand Falls chert member below. Black areas represent ore. Scale 100 feet to one inch. rences the galena predominates, partly because of solution and oxidation of zinc blende, and partly because the deposition of galena prevailed at these upper levels, below the shale. Large masses of galena are found here, in contrast to the conditions in the sheet ground. Both at Joplin and at Aurora (Fig. 140), as well as in camps in Oklahoma, the "runs" are also a characteristic form of the upper deposits; these sometimes extend for 1 or 2 miles, following 456 MINERAL DEPOSITS the same horizon at depths less than 150 feet usually much less. At Granby the width of the run is rarely more than 50 to 150 feet. Each run has usually several "openings" (brecciated ground filled with ore), each opening being rarely more than 5 or 6 feet thick. These runs appear to be solution cavities controlled by joints in the rocks. While in places the brecciation and mineralization continue down to the blanket deposits of the sheet ground, the latter extends in the main independently of the old pre-Pennsylvanian surface. In this sheet ground, which is from 6 to 15 feet thick, the galena and zinc blende occur in dark chert, filling brecciated L r * '^ i v~- SSSi, f -< sa^Sf-^^ Sec.6 U 8ec.5 Sec. 4 *3 m&*& **3&&" Sec.7 Sec.8 8ec.9 ~4 a FIG. 140. Plan showing shafts and workings along run of galena ore, north of Aurora, Missouri. After Arthur Win&low. old chert, and in elongated, narrow solution cavities due to dis- solved streaks of limestone in the prevailing chert. The sheet ground is the most important source of ore (Fig. 141). The newly discovered deposits in Oklahoma 1 at Miami and Quapaw have assumed great importance. The ore bodies form "runs" which in the main extend N.E. approximately parallel to the Miami fault. The ore which is richer than at Joplin occurs at a depth of 200 feet or less in Mississippian rocks under- neath a rather thick cover of Pennsylvanian shales. The mine water appears to form part of an artesian circulation and contains much H 2 S. Bitumen is in part so abundant as to become an objectionable constituent. The whole occurrence appears to support the theory of origin by ascending waters. 1 E. C. Siebenthal, Bull 340, U. S. Geol. Survey, 1908, pp. 187-228. Maps by H. A. Buehler in "War minerals of the Joplin district," Am. Inst. Mtn. Eng., Joplin-Miami meeting, Oct., 1917. LEAD AND ZINC DEPOSITS 457 Arkansas. In northern Arkansas, 1 a short distance southeast from the Joplin region, the zinc blende, generally without galena, occurs in fissures or crevices, in fault breccias, and in solution breccias, accompanied by secondary chert or dolomite, sometimes FIG. 141. Sketch illustrating the occurrence of galena and sphalerite in cavities in the sheet ground, Joplin, Mo. After C. E. Siebenthal, U. S. Geol. Survey. also by crystallized quartz; the ores are found in Ordovician limestone and also in the Boone (Mississippian) formation. Upper Mississippi Valley.' 2 ' The districts of the upper valley 1 G. I. Adams, Pro/. Paper 24, U. S. Geol. Survey, 1904. .1. C. Branner, Arkansas Geol. Survey, vol. 5, 1900. 2 J. D. Whitney, Geology of Wisconsin, vol. 1, 1862. T. C. Chamberlin, Geology of Wisconsin, vol. 4, 1882. W. P. Jenney, Trans., Am. Inst. Min. Eng., vol. 22, 1894, pp. 208-209 C. R. Van Hise, Some principles controlling the deposition of ores, Trans., Am. Inst. Min. Eng.. vol. 30, 1901. H. F. Bain, Bull. 294, U. S. Geol. Survey, 1906. U. S. Grant, Bull. 14, Wisconsin Geol. Survey, 1906. G. H. Cox, Econ. Geol., vol. 6, 1911, pp. 427-448; 582-603. H. C. George, Bull. 132, Am. Inst. Min. Eng., 1917, pp. 2045-2074. 458 MINERAL DEPOSITS lie in Wisconsin, Iowa, and Illinois. The most important dis- tricts are in Wisconsin and yielded in 1917 about 4,100 short tons of lead and 59,700 tons of zinc, with a total value of about $13,000,000. The ore deposits are found in Ordovician strata of almost horizontal position. The following formations are recognized : Feet 160 240 55 100 Lower magnesian limestone 350 Below the magnesian limestone is 700 feet of the Cambrian Potsdam sandstone. The so-called "oil rock," a thin bed of Cincinnati or Maquoketa shale . . Galena dolomite Platteville limestone (Trenton) . . St. Peter sandstone . . . FIG. 142. Section showing occurrence of lead and zinc in vertical crevices, flats, and pitches; also of disseminated ores in the rocks, gd, Galena dolomite; tk, Trenton limestone; of, upper flat; uf, lower flat; k, con- necting flats, pitches, and verticals. After T. C. Chamberlin. bituminous shale, is found at the base of the Galena formation or at the top of the Trenton. The rocks dip gently southwest and are flexed into very shallow troughs. The ores are confined to the Galena limestone and the upper part of the Platteville limestone; the minerals consist of mar- casite, sphalerite, and galena, deposited in the order noted. The gangue is crystallized calcite, rarely barite. Cadmium is absent, but a trace of silver is found. The abundance of marcasite causes metallurgical difficulties and necessitates treatment of the concentrates in electrostatic or magnetic separators, in the latter case preceded by partial roasting. The ores occur as LEAD AND ZINC DEPOSITS 459 fillings of open spaces, vertical crevices, or "gash veins" con- nected with "pitches" or "flats," all probably due to solution along joint planes (Fig. 142). Stalactites of sulphides are some- times found indicating that the spaces were not always filled by solutions. In part there are also flat bodies of disseminated ores. The galena predominates near the surface, probably largely because the zinc blende has been dissolved as sulphate and transformed to silicate or carbonate in the lower levels (p. 455). In depth zinc blende with a little galena is the prin- cipal ore. Mining operations extend to a depth of at most 200 feet. The distribution of the oil shale, according to Bain, seems to coincide with the extent of the deposits. Cox, however, holds that the metal was derived from the overlying Maquoketa shale and carried down to be concentrated in the Galena limestone. Virginia and Tennessee.'* In western Virginia, and near Knoxville, Tennessee, lead and zinc ores occur in the Cambro- Ordovician (Shenandoah) limestones, mostly where the rocks are faulted or brecciated or where they carry much organic matter. The gangue consists of calcite, dolomite, and rarely barite. There is little quartz or pyrite and no definite order of crystallization. Southeastern Missouri. 2 In eastern Missouri not far from the Mississippi River and south of St. Louis lead mining has been carried on more or less extensively since the early part of the eighteenth century, but in the last ten years the industry has assumed very large proportions. In 1917 the yield of lead from this region was 204,545 short tons, worth more than $35,000,000. The crude ores, which yield on the average 5.5 per cent. of. lead concentrates, are treated at the rate of 20,000 tons per day in local concentrating works and a part of the galena 1 T. L. Watson, Lead and zinc deposits of Virginia, Geol. Survey Virginia, vol. 57, 1905. Frank L. Xason, Characteristics of zinc deposits in North America, Trans. Am. Inst. Min. Eng., vol. 57, 1918, pp. 830-855. H. A. Coy and H. B. Henegar, Mining methods of the American Zinc Co. of Tenn., idem, vol. 58, 1918, pp. 36-47. 2 A. Winslow, Missouri Geol. Survey, vols. 6 and 7, 1894. A. Winslow, Bull. 132, U. S. Geol. Survey, 1896. C. R. Keyes, Missouri Geol. Survey, vol. 9, 1896. E. R. Buckley, Geology of the disseminated lead deposits, Missouri Bur. Geol. and Mines, vol. 9, pts. 1 and 2, 1909. A. P. Watt, Concentration practice in southeastern Missouri, Trans., Am. Inst, Min. Eng., vol. 57, 1918, pp. 322-419 460 MINERAL DEPOSITS is smelted in the district. Practically no zinc is contained in the ore. The geological position of the deposits is in the Cambrian and therefore lower than those of the other Mississippi Valley ores. On an irregular surface of pre-Cambrian granite and porphyry rests the basal La Motte sandstone, about 200 feet thick (Fig. 143). Above this lies the arenaceous dolomite of the Bonne- terre formation, often chloritic, with beds of shale having in all a thickness of 300 to 400 feet. Covering the Bonneterre are the Davis, Derby, Doe Run, and Potosi formations, which soo' Surface <>f Ground Uunucterre Formation Horizontal and Vertical Scale FIG. 143. Vertical section showing workings in mine No. 4, Federal Lead Company, southeastern Missouri. Horizontal bodies of disseminated ore, following bedding of shaly dolomite of Bonneterre formation. After E. R. Buckley. are mainly dolomites and shales and all of which belong to the Upper Cambrian. The principal ore horizon is in the lower part of the Bonne- terre dolomite, though some galena occurs throughout that for- mation. A second, less important ore horizon is in the Potosi dolomite, where the galena is accompanied by barite. The strata are horizontal or have very gentle dip. The ore minerals are mainly galena accompanied by calcite, a little pyrite, and sometimes chalcopyrite. In places for in- stance, at Mine La Motte and Fredericktown the ores contain LEAD AND ZINC DEPOSITS 461 nickel and cobalt, as linnseite (Co,Ni) s S4; some of the ores have been worked for these metals. Watt quotes a representative analysis of the crude ore of the southeastern district. 'ANALYSIS OF DISSEMINATED ORE-FROM SOUTHEASTERN MISSOURI Pb Cu '. Zn S Si0 2 Fe 2 3 4.32 0.03 0.50 0.97 4.83 6.64 A1 2 3 1.16 CaO.... 30.80 MgO 17.96 CO 2 32.79 100.00 Silver 0.12 oz. per ton. Trace Ni, CO, Mn. Watt states that the silver follows the zinc. Concentrates of zinc blende contain up to 10 ounces of silver per ton. The ores are often called disseminated, for the galena usually occurs as grain or crystals disseminated in the greenish-gray dolomite (Fig. 144); sometimes these crystals are several centi- meters in diameter. According to Buckley the ores of the lower part of the Bonneterre occur as follows : 1. As horizontal sheets along bedding planes, generally along the upper side of thin shale beds. 2. Disseminated in dolomite. 3. Filling or lining joints. 4. In cavities or vugs. The galena is persistently asso- ciated with dark dolomite and black shale. The ores are mined from ver- tical shafts, 100 to 550 feet deep. The ore does not extend in all directions like a coal bed, but the flat shoots or "runs" follow rather persistently one or two direc- tions, undoubtedly controlled by joints and small faults. Some of these rune have been followed for miles and may be several hundred feet wide; some of the mine workings in the Bonne- terre district are 100 feet high. Genesis of the Mississippi Valley Deposits. An unusually extensive literature, full of controversy and divergent views, covers the question of the genesis of these ores. A majority of FIG. 144. Crystals of galena developing in shaly dolomite. Black, galena; shaded and stip- pled, shaly dolomite; white, quartz. Magnified about 10 diam- eters. After E. R. Buckley. 462 MINERAL DEPOSITS the authors agree that the source of the ores was in the Paleozoic sedimentary beds, also that the deposition was effected by atmos- pheric waters, and finally that the metals were in solution mainly as sulphates. Summaries of the various views are found in the text-books of Kemp and Hies and in the reports of A. Winslow and E. R. Buckley. At. present there are two strongly contrast- ing opinions regarding the Missouri deposits. The descensionists are represented by Whitney, Chamberlin, Blake, Robertson, Winslow, Buckley, and Buehler and the ascensionists by Jenney, Nason, Van Hise, Bain, W. S. T. Smith, and Siebenthal. From the latter we may separate Jenney and Nason, who see in the ore deposits the result of fissuring extending into the underlying pre-Cambrian rocks, through which thermal waters ascended. Buckley and Buehler hold that the source of the lead and zinc was in the Pennsylvanian sediments, which, however, contain no important deposits and only in places small amounts of galena and zinc blende. The finely distributed sulphides were dissolved as sulphates and carried downward in acid solutions which finally mingled with neutral or alkaline solutions from the unoxidized parts of Pennsylvanian sediments. These mingled waters de- posited galena and zinc blende in the sink holes and drainage channels of the underlying Mississippian limestone and chert. Unlike Winslow, Buckley and Buehler do not believe that the deposition was effected at the time of deposition of the Pennsyl- vanian shale, but later, after the erosion of a part of those beds. The obvious difficulty in their theory appears to be that it re- quires the waters to have descended through an impervious shale cover. A similar theory is advocated by Buckley for the lead deposits of southeastern Missouri. The pre-Cambrian rocks are held to be the original source of the metals. The water flowing into the Cambrian sea contained lead, which was deposited with the Bonneterre dolomite as small particles. After the subsequent formations were laid down the concentration of lead by surface waters began. Finally the Pennsylvanian shales were laid down over this area and from them the greatest amounts of metals were derived. The solutions were thus in the main descending, though in part they may have ascended in artesian circulation through the La Motte sandstone. Buckley states, indeed, that even at present there are strong indications of arte- sian conditions in the mines. On the whole the ore deposition is LEAD AND ZINC DEPOSITS 463 post-Pennsylvanian. Here again the impermeable character of the Pennsylvanian may be advanced as an argument against Buckley's view, as well as the improbability of a strong descend- ing flow through the great thickness of Cambrian, Ordovician, and Mississippian beds. A satisfactory explanation of the southeastern Missouri deposits is as yet lacking. On the other hand, Van Hise, Bain, Tangier Smith, and Siebenthal, who have studied the Joplin district, believe that the source of the ores is in the various formations below the Pennsylvanian, particularly in the Cambro-Ordovician, and that atmospheric waters penetrating these rocks were carried up against the impervious beds of the Pennsylvanian and here de- posited in the pre-Pennsylvanian breccias and sink-holes. Smith and Siebenthal hold that the ores were formed much later than the Pennsylvanian, after the Ozark uplift (Fig. 1 36) had established an artesian circulation. The surface waters entered the older Paleozoic outcrops to the south and east of the Joplin region. After following these beds they passed upward through the jointed and brecciated Mississippian limestone until they reached the vicinity of the impermeable shales. There is, indeed, in the deep wells of Joplin good evidence of the existence of artesian pressure. Siebenthal's recent contribution in which very strong arguments are advanced in favor of the artesian theory has already been reviewed in the general part of this chapter. For the deposits of Wisconsin and Illinois, Van Hise and Bain assume that the metals were minutely disseminated as sulphides through the Galena dolomite and concentrated, probably in late Tertiary or post-Tertiary time, by the action of surface waters descending in shallow troughs through the fractured and slightly inclined Galena limestone, and that the reduction was effected by the organic matter of the oil rock. In the Galena limestone the solutions were practically confined between two beds of almost impermeable 'shale. According to Cox the metals were derived from the overlying Maquoketa shale, in which he finds some evidence of the presence of sulphides, particularly sphalerite. He points out, with good reason, that the shales are better suited as receptacles for metallic ores than the limestones. The latter are deposited in deeper water, while the shales are shore formations in which the metallic substances derived from adjacent continents would most easily be deposited as detritus or precipitated from solutions. 464 MINERAL DEPOSITS We find here the same conflict of opinion, as in the case of the Missouri deposits and those of Silesia, between the ascensionists and the descensionists. The problem is not yet solved, but looking beyond these controversies, we cannot deny that in many countries transition types appear which seem to connect these apparently distinct non-igneous deposits with deposits of igneous affiliations. L. V. Pirsson 1 in 1915 expressed the view that the zinc-lead deposits of the Mississippi valley type might well result from "the quiet upward movement of volatile magmatic material" thus ranging himself with W. P. Jenney and F. L. Nason. It will be incumbent upon the supporters of this theory to controvert the strong arguments offered by Siebenthal in favor of leaching of limestones by ascending saline meteoric waters. Origin of certain ore deposits, Econ. Geol, vol. 10, 1915, pp. 180-186. CHAPTER XXIV METALLIFEROUS DEPOSITS FORMED NEAR THE SUR- FACE BY ASCENDING THERMAL WATERS AND IN GENETIC CONNECTION WITH IGNEOUS ROCKS CHARACTER AND ORIGIN General Features. The deposits at the orifices of hot ascend- ing springs have been described in Chapter VII. It has been shown that they consist of opal, chalcedony, quartz, calcite, aragonite, barite, and fluorite, with a number of other gangue minerals, and that they also contain in places metallic gold and certain sulphides, such as cinnabar, stibnite, and pyrite, but not the other' common ore minerals such as chalcopyrite, galena, zinc blende, and arsenopyrite. The ore deposits described in the present chapter present some striking analogies to those products of the hot springs. In regions of comparatively recent volcanic activity where the measure of erosion since the eruptions ceased is in hundreds rather than in thousands of feet we find a group of important ore deposits, most commonly in the form of fissure veins. They generally occur in igneous flow rocks 1 and also 'cut the un- derlying or adjacent formations. They constitute the source of a large part of the world's production of gold, silver, and quicksilver, and they contain the spectacular bonanzas of the Cordilleran region, of which examples are found at Tuscarora, Virginia City, Goldfield, Cripple Creek, Pachuca, Guanajuato, and many other districts. Following the Tertiary outbursts of effusive rocks, these deposits accompany the "circle of fire" 1 We are accustomed to consider as intrusive rocks those which have congealed with granular texture far below the surface. Intrusions are, however, not confined to any particular depth or texture. Intrusive bodies may be found in any series of rocks even near the surface and may then have fine-grained, trachytic or even glassy texture. The distinction between flows and intrusions may in such cases become difficult and, as shown in case of the Tonopah, Waihi and other districts, the relations may have far reaching bearing upon the richness and continuation of the deposits con- tained in such a series of rocks. 465 466 MINERAL DEPOSITS that encompasses the Pacific Ocean. We find them in Japan, in the East Indian Islands, and in New Zealand. They are characteristically developed in that classical mining region of the Old World, in Hungary and Transylvania, where one of the elements tellurium which so often accompanies them was first found. Though most of these ore-deposits are found in the Tertiary flow rocks they are not confined to rocks of this period. There is good reason to believe that veins are developing now in some regions of recent volcanism, and also that similar veins have been formed during pre-Tertiary outbreaks, although erosion has removed most of the older representatives of this type. These deposits have certain well-marked characteristics which are partly of a mechanical, partly of a chemical origin. Because the fissuring of the rocks took place near the surface, under slight load, open cavities were abundant, and filling, crustification, and comb structure are conspicuous. The walls are likely to be irregular, and the vein matter is often "frozen" to the walls. Splitting, chambering, and brecciation are features of the veins. While metasomatic processes have been active in the surrounding rocks, the ore is usually confined to the open fissures. Short and irregular veins are more frequent than the regularly developed conjugated fractures resulting from strong compressive stress. Divergent systems of fractures or several parallel systems with little apparent relationship are thought to be due to the gravitative settling of volcanic piles. Banding caused by crustification is common, as illustrated in Figs. 160, 166, 172, 176 and 180. It is much more delicate and frequent than in deposits formed at greater depth and higher temperature. The occurrence of the ore in "stock works," or in pipes, or below impervious beds is often observed. In superimposed lava flows of different kinds, some are usually better adapted to the deposition of ore than others and this difference may result in the development of ore-shoots which are approximately horizontal. Among the metals contained in these deposits gold and silver are by far the most important. Base metals are present, plenti- fully enough in places, but the mines are rarely worked for these. Large bodies of galena and zinc blende occur in some places, but it is decidedly rare to find important copper deposits. The "pyritic" deposits are not represented; they are confined to the DEPOSITS FORMED NEAR THE SURFACE 467 deeper zones or to those of higher temperatures. Arsenic and antimony, bismuth, tellurium, and selenium are common but are rarely of economic importance; quicksilver is present in some deposits and indeed the typical quicksilver deposits belong to this class. Cobalt and nickel, tungsten, and molybdenum are not unknown, but are entirely subordinate. Their home is in the deeper deposits. The pure gold deposits are relatively scarce. Those carrying silver only are common in certain regions, like Mexico. The usual metals are gold and silver occurring together in varying proportions. Among the ore minerals native gold should be mentioned first. It contains silver, as a rule, and is of pale yellow color; a propor- tion sometimes occurring is ounce for ounce when the mineral is of very pale grayish-yellow color (electrum). Deep yellow gold is not unknown, however. The gold is often present in very fine mechanical distribution, being sometimes so closely intergrown with ore minerals and gangue that no colors can be obtained by panning. When derived by oxidation of tellurides it is of dull brown color and is difficult to recognize even in rich specimens. The whole series of tellurides is present. As the gold generally occurs in minute particles rich placers below the croppings of these deposits are rather unusual. Native silver is ordinarily a product of oxidation. The primary and most abundant silver mineral is argentite; com- plex silver sulphantimonides and sulpharsenides are also charac- teristic ; it is often difficult to say which are secondary and which are primary. Among them are proustite, pyrargyrite, miargy- rite, stephanite, polybasite, tetrahedrite, and more rarely enargite. Stibnite is plentiful in deposits of certain types. Among the base minerals pyrite is always present, but in small quantity and fine distribution; marcasite, a mineral typical of deposition near the surface, is not unusual; often it is secondary. There are also galena, zinc blende, chalcopyrite, and sometimes ala- bandite; rarely arsenopyrite ; never pyrrhotite or magnetite. Of gangue minerals quartz is the most abundant, and crystals of it are plentiful but rarely large; an amethyst color is often noticeable. The quartz aggregates are not glassy or milky but usually fine-grained (hornstone) and often chalcedonic, with banding a-nd rapidly changing grain. Chalcedony and opal 468 MINERAL DEPOSITS are usually later than the quartz. Calcite, dolomite, barite, and fluorite are locally the dominant gangue minerals, while siderite is rare. Manganese minerals like rhodochrosite and sometimes rhodonite are typical of certain groups. Kaolin accompanies the veins, sometimes in large amounts,, but is probably in most cases a product of secondary changes by descending waters. Sericite and chlorite appear in the altered country rock. Zeolites are present in some deposits, but are certainly of exceptional occurrence. One of the most widespread and characteristic gangue minerals and the most difficult to explain is adularia (or valencianite). 1 Discovered by Breithaupt in specimens from the Valenciana mine at Guanajuato, this mineral has since been found in numer- ous other places, mainly in the Cordilleran region, as part of the filling, and as a metasomatic product in the country rock. Among the places where this feldspar plays an important part may be mentioned Silver City (Idaho), Tuscarora, Tonopah, and Rawhide (Nevada), Gold Road (Arizona), Republic (Wash- ington), and Cripple Creek and Creede (Colorado). It does not occur at Goldfield, Nevada, where solutions of acid reaction appear to have deposited the ore. The orthoclase mineral is usually a pure potassium feldspar, although varieties with several per cent, of sodium have been found at Waihi, New Zealand, and in the Gold Spring district, Utah. 2 Sometimes the adularia replaces orthoclase, biotite, and other rock minerals (Fig. 145); it is also found in the form of well-developed crystals of prism and dome intergrown with vein quartz (Fig. 146). The cross- 1 A. Breithaupt, Ueber die Felsite und einige neue Specien ihres Gesch- lechts, Schweigg. Jour., Bd. 60, p. 322, 1830. W. Lindgren, Orthoclase as a gangue mineral in fissure veins, Am. Jour. Sci., 4th ser., vol. 5, 1898, p. 418. W. Lindgren, Twentieth Ann. Rept. U. S. Geol. Survey, pt. 3, 1900, p. 167. J. E. Spurr, Prof. Paper 42, U. S. Geol. Survey, 1905, p. 86. W. Lindgren and F. L. Ransome, Prof. Paper 54, U. S. Geol. Survey, 1906, p. 187. A. F. Rogers, Orthoclase-bearing veins from Rawhide, Nevada, Econ. Geol, vol. 6, 1911, p. 790. F. C. Schrader, Mineral deposits of the Cerbat Range, Black Mountains and Grand Wash Cliffs, Mohave County, Arizona, Bull. 397, U. S. Geol. Survey, 1909. F. C. Schrader, A reconnaissance of the Jarbidge district, Nevada. Bull. 497, U. S. Geol. Survey, 1912. 2 B. S. Butler, U. S. Geol. Survey, oral communication. DEPOSITS FORMED NEAR THE SURFACE 469 sections of the adularia crystals are usually of rhombic shape. The mineral also occurs abundantly in some veins that had originally a calcite gangue, now replaced by an intimate inter- growth of adularia and quartz. The high-temperature minerals, such as augite, amphibole, olivine, biotite, tourmaline, topaz, garnet, magnetite, ilmenite, and chromite, are conspicuously absent. Successive Phases of Mineralization. Veins formed near the surface in volcanic regions are sometimes subject to peculiar changes, which are rarely observed in deposits of more deep- ^^V^^^^^^B^^ .FiG. 1 45. Adularia (Ad) replacing soda-lime feldspar (An) in andesito from Tonopah, Nevada. Magnified 17 diameters. After J. E. Spurr, U. S. Geol. Survey. seated origin. An earlier gangue mineral, such as calcite or barite, may be wholly wiped out and replaced by a new gangue of quartz and adularia. This alteration has nothing to do with surface waters though the latter may sometimes produce a simi- lar cellular or lamellar structure; it is plainly caused by a change in the composition of the ascending currents. Indications of this process may be seen even where it has not been carried to completion. In many veins at Cripple Creek deposition be- gan by the growth of slender crystals of celestite from the walls, 470 MINERAL DEPOSITS and these crystals were subsequently replaced by quartz, in which the pseudomorphs are now embedded. In the Trade Dollar vein at Silver City, Idaho, the filling consists of quartz and adu- laria, but casts of barite or calcite covered with minute crystals of adularia indicate that here also there was a preliminary carbon- ate or sulphate stage. In many instances the vein was completely filled by calcite, each grain separated by a slender partition of quartz; at the beginning of the second stage this calcite was dissolved, leaving a skeleton of thin silica walls; secondary quartz and often also FIG. 146. Intergrowth of quartz (q) and adularia (a), Fraction vein, Tonopah, Nevada. Magnified 38 diameters. After J. E. Spurr, U. S. Geol. Survey. adularia were deposited upon these walls, giving them more strength, but the ore remains a delicate aggregate of "hackly" or lamellar quartz, such as'is exceedingly characteristic of some mining district. At De Lamar, Idaho, this ore consists only of quartz (Figs. 147 and 148). In the veins at Gold Road, Arizona, and many other veins in the same district, the original gangue material consisted of calcite and fluorite and the "pseudo- morphic" ore consists of quartz and large amounts of adularia. Similar ore may be seen in the Mount Baldy district, southern DEPOSITS FORMED NEAR THE SURFACE 471 Utah, at Jarbidge, Nevada (Fig. 149), and at many other places. This important development of adularia, involving FIG. 147. Lamellar quartz, replacing calcite gangue, De Lamar, Idaho. One-half natural size. FIG. 148. Section of lamellar ore, De Lamar, Idaho. Natural size. transportation of alumina by siliceous solutions, remains without full explanation. The composition of the ore may be similar to that of a pegmatite dike, but the structure is wholly different. 472 MINERAL DEPOSITS There is reason to believe that this ''pseudomorphism" is accompanied by a change in the metal content of the vein. At least it seems as if the original filling of barite, calcite, and fluorite carried more silver and as if the silicification and feld- spathization was accompanied by a concentration of the gold. Similar processes may be traced in some quartz veins of the Republic district, Washington. Here quartz with some adularia replaces a slender acicular or thin tabular mineral, probably calcite, developed parallel to c and r, which seems to have been deposited only along the walls of the vein. Zeolitic Replacement. Zeolites are foreign to veins of the deeper zones ;'in the'veins formed near the surface they_are occa- FIG. 149. Thin section of lamellar quartz and adularia, pseudomorphic after calcite, Jarbidge, Nevada. Magnified 12 diameters. After F. C. Schroder, U. S. Geol. Survey. sionally found, but they are rare. At a few places zeolites are reported in the altered country rock (Tonopah, the Comstock, and Waihi). At Guanajuato zeolites are found in the filling of the veins, but here they always belong to the latest phases of vein formation. Apophyllite, laumontite, and stilbite are the species reported. Few of these occurrences in the vein filling have been carefully studied. In the Southern Republic mine at Republic, Washington, laumontite, associated with calcite, occurs on a fairly large scale. 1 At this place the ordinary fine- 1 W. Lindgren, Trans., Canadian Min. Inst., vol. 15, 1912, pp. 187-191. DEPOSITS FORMED NEAR THE SURFACE 473 grained banded quartz filling had evidently been dissolved and the laumontite-calcite aggregate was deposited in its place. The ore in this zeolitic zone or shoot contains mainly silver, whereas elsewhere in the mine gold predominates in the quartz gangue. It seems to be worth while to draw attention to this occurrence with a view to ascertaining whether the development of zeolites is not favorable to silver enrichment. Many facts noted in veins of other classes, like those of Kongsberg and Andreasberg, point in this direction. Primary Ore Shoots, Oxidation, and Sulphide Enrichment. Rich oxidized ores are often encountered in these deposits at the surface and down to the water-level. Whether the primary ore is greatly enriched in this zone depends really more on the texture and composition of the ore than on its original tenor. In veins of hard fine-grained quartz oxidation often fails to pro- duce an ore of higher grade. There are many districts in which the oxidized ores are little, if any, richer than those below the oxidized zone. Such conditions exist, for instance, at Cripple Creek and at Tonopah. The largest and richest masses of ore are often found just below the oxidized zone and in general contain sulphides, sulph- antimonides, and sulpharsenides. It will suffice to call atten- tion to the great silver bonanzas of Guanajuato and Pachuca; to the Comstock, where in one month silver-gold ores valued at $6,000,000 were extracted; to Tonopah, Nevada, where in three months ore yielding over $3,000,000 was extracted; to the Caledonian mine at Thames, New Zealand, which in one year from a small ore-shoot produced $6,000,000 in gold; to Cripple Creek, Colorado, where in one year from a small area but a considerable number of mines $18,000,000 in gold was pro- duced; to Goldfield, Nevada, where during a recent year over $10,000,000 in gold was recovered from one property, the ore averaging $38.50 per ton, and where, of the phenomenally rich ore shipped in 1907, one carload of 47 tons yielded $600,000 in gold. At the same time it is well to emphasize the fact that most of these high yields proved ephemeral. The bonanzas were ex- tracted, poorer ore was found in depth, and the mine was aban- doned or continued in feeble existence. These great bonanzas are due in part to primary deposition in large degree probably to the later, reconcentrating phases of 474 MINERAL DEPOSITS primary deposition; and in part to sulphide concentration by descending waters charged with precious metals from the upper parts of the veins. It should not be overlooked that ore-shoots of primary origin are common enough. Take, for instance, the Cripple Creek gold ores, in which evidence of enrichment is conspicuously lacking; these primary shoots are usually of a markedly irregular form; many of the smaller ore-bodies are likely to follow intersections of fissures. In depth the rich shoots show a tendency to contract, to feather out, or simply to become impoverished. In many cases zinc blende in a quartz gangue appears in depth. On the other hand sulphide enrichment is conspicuously marked, especially in silver veins. In gold-bearing veins the enrichment in gold is likely to be localized in the lower part of the zone of oxidation. The secondary silver minerals are native silver, argentite, ruby silver, stephanite, and polyba- site that is, the same minerals (except the native silver) as occur in primary ore. As a consequence it is often exceedingly difficult to distinguish primary ore and enriched ore, and when in addition to this we recognize that the later effects of primary mineraliza- tion may sometimes closely simulate the products of descending surface waters, the difficulties of correct interpretation will be fully realized. The distinction is made easier when the secondary sulphides form a distinct zone immediately below the oxidized part of the lode. The conditions for the deposition of gold and silver are appar- ently much more favorable near the surface than at greater depths; on the other hand, deposition took place rapidly and the gold and silver contents of the solutions were doubtless ex- hausted before they reached the actual surface. Types of Deposits. The merging of the various types makes it difficult to establish a rigid classification. One type, namely, the zeolitic copper deposits in amygdaloid rocks, has been left out of consideration at this time, for it really represents a mineralization of the lava derived from its own body. To gain a general orientation the deposits here described are classified as follows: 1. Cinnabar Deposits. Cinnabar, marcasite, stibnite, hydro- carbons, quartz, opal, calcite. 2. Stibnite Deposits. Stibnite, pyrite, and some other sul- phides; also quartz % 3. Base Metal Deposits. Chalcopyrite, galena, zinc blende, DEPOSITS FORMED NEAR THE SURFACE 475 tetrahedrite, in an abundant gangue of quartz, carbonate, or barite. Principal values usually in gold and silver. 4. Gold Deposits. Native gold, alloyed with silver. Sub- ordinate argentite, ruby silver, etc. Quartz. 5. Argerdite-Gold Deposits. Argentite, ruby silver, tetra- hedrite, etc.; native gold, quartz, calcite. 6. Argentite Deposits. Argentite, ruby silver, tetrahedrite, etc.; quartz or calcite, barite, and fluorite. 7. Gold Telluride Deposits. Gold tellurides, quartz, or quartz and fluorite. 8. Gold Telluride Deposits with Alunite. Gold tellurides, pyrite, alunite, kaolin. 9. Gold Selenide Deposits. Gold selenides, pyrite, quartz, calcite. Older Representatives of this Class. The types just enum- erated almost wholly represent veins or similar deposits in Terti- ary lavas of the Cordilleran or Pacific or Hungarian regions, but a close examination will easily discover examples of similar de- posits of a greater geological age. Beck 1 described relatively unimportant deposits in the Mesozoic melaphyres and quartz porphyries at Imsbach, in the German Palatinate, that carry chalcopyrite, galena, and tetrahedrite in a gangue of calcite, ba- rite, and rhodochrosite and are probably ancient representatives of this class. The celebrated veins of Freiberg, or at least three types of them, namely, the "noble quartz formation," the "noble carbonate formation" and especially the "barytic lead forma- tion," should be mentioned in this connection. There seems to be good evidence that these are Carboniferous representatives of veins formed near the surface, although the lavas in which they probably reached the surface are now eroded. Their structure' and composition point clearly to a shallow deposition, and were the physiographic conditions in the Erzgebirge fully analyzed the results would probably confirm this view. The "barytic lead formation," for instance, carries barite, fluorite, quartz, and dense quartz as gangue minerals with beautiful crustification, while the ore minerals are argentiferous galena, tetrahedrite, bournonite, and chalcopyrite. Another occurrence that might well be cited comprises the insignificant veins in the Triassic diabase flows at Bergen Hill, New Jersey, which contain pyrite and galena in a gangue quartz and adularia, with secondary zeolites. 1 R. Beck, Lehre von den Erzlagerstatten, vol. 1, 1909, p. 334. i 476 MINERAL DEPOSITS Genesis. In the preceding pages attention has been called to the strong evidence connecting the class of deposits here dis- cussed with igneous action and pointing to ascending hot waters as the agents of deposition. The best proof that the ores were not formed by the ordinary circulation of surface waters is the fact that deposition has not proceeded uniformly, but that the vein-forming epochs were of brief duration and followed closely after each considerable eruption. Evidence of this relation is available from many important districts. At Tonopah the prin- cipal mineralization followed the eruption of the earlier andesite and the veins are truncated by the flow of the later andesite and the later rhyolite. At Jarbidge, Nevada, the veins are con- tained in the early rhyolite, while the later rhyolite is barren. At Waihi, New Zealand, the rich veins are sharply truncated by erosion and capped by a rhyolite of later age. The occurrence of these deposits in lavas really counts for but little; there are vast areas of lava flows absolutely barren of mineral deposits. On the other hand, several of the Hungarian authors have pointed out the fact that the veins are confined mainly to the vicinity of volcanic necks or centers of eruption. Exactly the same conclusions have been reached in the United States. This feature serves to connect the veins formed near the surface with those of greater depths. The deposits in the surface lavas are, then, simply the tops of veins, the roots of which are to be found in the intrusive masses of the depths. No matter whence all the water or part of the water came, the deposition of the substance of the veins their valuable content appears to be a phenomenon connected with intrusive activity and not merely dependent upon the heat furnished by the lava flows to circulating surface waters. The metals, as well as the sulphur, carbon dioxide, and fluorine, were in all probability derived from intrusive underlying masses. Proof of Depth below Surface. Physiography furnishes the data on the original surface during deposition. We may be able to trace the old surface of the volcanic slope or plateau and ascertain the relation of the outcrops to the uppermost flow, or in dissected volcanic piles it may be possible to reconstruct approximately the old surface of the volcanic cone. Of this latter possibility Cripple Creek is an instance (Fig. 167); the present surface was probably less than 1,500 feet below the origi- DEPOSITS FORMED NEAR THE SURFACE 477 nal surface of the volcanic cone. Ransome estimates that at Goldfield, Nevada, the surface has been degraded but a few hundred feet below the original contours of the flows. A fine example showing the connection of deposits formed near the surface with those of more deep-seated type is offered by the San Juan region, in Colorado, where erosion has not only inter- sected the flows but laid bare the intrusive masses forced into them all within a vertical interval of 6,000 feet. Proof of Temperature. The similarity to hot-spring deposits is least marked in deep-seated veins, but becomes striking in the veins here under consideration. The fine-grained chalcedonic and banded quartz of spring deposits (Fig. 5, p. 101) is entirely similar to the often delicate and beautifully banded and crustified portions of these veins. The evidence indicates deposition by waters that held in solution large quantities of substances not easily soluble that is, by hot waters which at the surface could not have had a temperature of more than 100 C. The minerals present are those which we have reason to believe were developed at a temperature less than 200 C. What the actual temperatures were in each case is of course scarcely possible to ascertain. The hot springs are volcanic "after effects" and usually ascend through the cooled lavas. In some cases the waters rise through bodies of hot rocks and then the pressure may become so high that the solutions issue at the surface as gases and form "fumaroles" and "soffioni" which sometimes, at their orifices, have a temperature of as much as 200 C. In these rarer instances the high temperature deposits, marked for in- stance by tourmaline or cassiterite, may develop close to the surface. Relation to Other Veins. The question naturally arises as to the character of these veins in depth. Do they actually change to assume the aspect of the veins of the deeper zones? The evidence, scant as it is, indicates that this is probably true. In regions of deeply eroded volcanic flows, like the San Juan country in Colorado, the veins in the lower exposures show an approach to the types of deep-seated origin. During the long ascent there was no doubt a progressive change in the nature of the depositing waters; some of their constituents were deposited and others were acquired from the rocks they traversed. 478 MINERAL DEPOSITS METASOMATIC PROCESSES Extent of Alteration. At considerable depths the ore-forming solutions move in the paths prescribed by fissuring and breccia- tion; they rarely penetrate great masses of rocks. Near the surface, especially in the great volcanic piles, different conditions prevail. There are here thick beds of tuffs and agglomerates with great porosity, and the stresses may irregularly shatter large volumes of rocks. The solutions and gases of meteoric or telluric origin move far more freely and alteration is effected on the largest scale. Here, too, we find most emphasized the peculiar effects of the mingling of ascending and descending solutions. Any one who has visited an active or recently extinct volcano has undoubtedly observed the areas of discolored red- dish, brown, and yellow rocks which indicate alteration. Erosion of older volcanoes discloses similar zones of alteration on a large scale and exposes metalliferous deposits formed in their interior. Types of Alteration. Gases given off by the ascending lavas penetrate the volcanic cones. They are admixed with water vapor, which may or may not be of intratelluric origin, and, near the point of issue, oxidation and interaction produce compounds like sulphur dioxide, sulphuric acid, hydrochloric acid, and sulphur. Of these reagents carbon dioxide and sulphuric acid are most effective in rock alteration. The volcanic rocks are converted to kaolin. Alunite, jarosite, and other sulphates are often mixed with these minerals. These masses of altered rocks, which are formed on the slopes of volcanoes, scarcely ever carry valuable ores, probably because the metallic load of ascending waters is usually deposited before the cool surface zone is reached. Hot springs begin to issue after the explosive igneous action has declined and the rocks cooled so that the fumarolic action is supplanted by rising aqueous solutions. These waters contain no strong acids, but probably mainly carbon dioxide, silica/and hydrogen sulphide, and are of alkaline rather than acid reaction. Some of these waters move slowly, percolating through great mas- ses of rocks; others move rapidly in prescribed channels and effect extensive changes in the immediately adjoining rock. One of the most common types o"f alteration is that resulting in DEPOSITS FORMED NEAR THE SURFACE 479 the "propylitic facies;" it affects mainly andesites and basalts, more rarely rhyolites, often spreading over wide areas in mineral- ized districts. Its mineralogical characteristics consist in the abundant development of chlorite and pyrite, sometimes also epidote; in places it is accompanied by the development of carbon- ates and a little sericite. The rock assumes a dull green color. The chemical changes consist of a moderate leaching of both potassium and sodium; the silica is usually decreased, as are also calcium and magnesium, except when carbonates of these metals are formed. The composition of the rock changes but Little and the additions consist only of sulphur and some water of hydration. Still another type of alteration, seen mostly in siliceous rocks like rhyolite, but also in other kinds, consist in a general silicifica- tion of the groundmass and phenocrysts, with aureoles of quartz developing around quartz phenocrysts. More rarely sericite develops in abundance and the effusive rocks are converted over large areas to a mixture of quartz and sericite, with more or less pyritization. Near the veins the alteration is usually most intense, although here, too, simply chloritized rock may often adjoin the fissure filling. In sericitization sodium is almost entirely carried away and potassium is accumulated in a marked degree in sericite and adularia; the latter mineral has a wide distribution, both in the altered country rock and in the fissure filling. Unless carbonates are formed, calcium and, to a less degree, magnesium are carried away; much pyrite is introduced which usually contains at least traces of precious metals. The percentage of silica is reduced. Close to the vein, silicification often assumes the ascendancy and a quartzose mass of silica, adularia, and sericite, with more or less sulphides, develops and may form part of the ore. In rare cases hydrargillite and zeolites may appear in the altered rock. Rutile appears to be the only stable titanium mineral. Manganese, titanium, and phosphorus are partly removed from the rock. Nearer the surface another potassium-aluminum mineral alunite appears in considerable quantities. This hydrous sul- phate is characteristic of large altered areas in volcanic regions, 1 but, being inconspicuous, is easily overlooked. That it often occurs together with pyrite and sericite is clearly proved, and its development in this phase is probably confined to the zone ' B. S. Butler and H. S. Gale, Alunite, Bull. 511, TJ. S. Geol. Survey, 1912. 480 MINERAL DEPOSITS where the descending waters carrying free sulphuric acid meet the ascending currents of alkaline waters. It appears to belong to a distinctly higher horizon than the sericite and adularia. In some alunites the potassium is in part replaced by sodium. In eroded and mineralized volcanic regions there is finally another type of alteration, the effects of which were super- imposed upon the earlier changes and tend to confuse the true history. As soon as the mineralized rocks become exposed to the air oxidation begins and sulphuric acid is generated by the action of oxygen on sulphides. This sulphuric acid descends with the surface water and converts the sericitized rocks into kaolin mixed with alunite and other oxidation products. Where waters exceptionally rich in sulphuric acid have acted on the rocks almost everything but quartz is carried away and the final result is a loose quartz aggregate. Descending still farther these sulphuric acid solutions may lose their oxygen, and, under certain circumstances, secondary sulphides, with alunite and sericite, may again develop. This brief sketch indicates how complicated the series of reac- tions may be and how the same minerals may form at different steps of the process. It is assumed in the above discussion that ascending alkaline waters do not form kaolin. This is undoubtedly true in general, but it is possible that kaolin may be formed in places by such waters close to the surface. The processes of alteration by hot ascending waters seem to result in minerals of only moderate hydration; zeolites, kaolin, and other strongly hydrated minerals are conspicuously absent. The zeolites appear to require quies- cent, stagnant conditions, such as do not exist close to strong ascending currents. Metasomatic Processes at Thames and Waihi. Extensive pro- pylitization has been described by several authors from observa- tions in the Hauraki Peninsula, in the northern island of New Zealand, 1 where rich gold-bearing veins appear in volcanic rocks like pyroxene andesite or dacite. The extreme phase close to the veins is a grayish-white rock, but a more widespread type is a chloritized andesite which corresponds to the propylitic facies as defined on a previous page. In this second type the ferric 1 A. M. Finlayson, Problems in the geology of the Hauraki gold fields, Econ. Geol, vol. 4, 1909, pp. 632-645. DEPOSITS FORMED NEAR THE SURFACE 481 minerals are chloritized, the pyroxene often passing first through a uralitic stage, while the plagioclase remains comparatively fresh, but contains some calcite and sericite. Finlayson has presented two extremely valuable series of analyses, which are given below in full. They represent rocks from Thames and Waihi, the two most important fields in the peninsula. The first column in each table gives the composition of the fresh rock, the second that of the propylitic or chloritic facies, and the rest are analyses of the more altered forms in which sericite and adularia are the predominating metasomatic products. The chemical changes during propylitization are not great. ANALYSES OF FRESH AND ALTERED ROCKS IN THE THAMES DISTRICT SiO 2 57.42 52.69 57.99 55.38 58.98 Ti0 2 0.68 0.53 0.51 0.24 0.11 AlA 17.61 18.33 17.59 15.63 11.21 FeA 2.34 2.32 1.56 1.88 1.45 FeO 3.77 2.98 2.37 2.95 2.42 MnO 0.43 0.25 0.21 0.23 0.11 MgO 2.19 3.09 2.01 1.88 1.43 CaO 5.69 7.87 5.45 6.01 8.11 Na 2 O 3.22 2.62 1.98 0.83 0.61 K 2 1.94 0.98 1.65 3.28 3.93 H 2 O- 0.85 0.73 1.56 2.41 2.54 H 2 O+ 2.62 3.71 1.89 1.92 1.15 CO 2 0.95 3.59 3.89 4.58 4.69 0.31 0.42 0.35 0.11 0.06 FeS 2 1.42 2.35 3.13 Total 100.02 100.11 100.43 99.68 99.93 1. Fresh hornblende andesite, Thames. 2. Chloritized hornblende andesite, Halcyon mine. 3. Altered andesite, 14 feet from Ophir vein, Halcyon mine. 4. Altered andesite, 5 feet from Ophir vein, Halcyon mine. 5. Altered andesite, adjoining Ophir vein, I Halcyon mine. J Sericite-py rite - carbonate rock from the 386-foot level. 482 MINERAL DEPOSITS ANALYSES OF FRESH AND ALTERED ROCKS AT THE WAIHI MINE 2 3 4 5 6 1. SiO 2 '. 63.45 58.39 61.78 69.35 76.61 85.65 TiO 2 0.75 0.68 0.69 0.43 0.28 tr. A1 2 O 3 ! lu.26 16.51 14.89 11.66 8.31 1.35 Fe 2 O 3 ' 2.28 2.46 2.08 1.53 1.08 0.43 FeO 3.01 2.98 2.51 1.66 0.59 0.21 MnO i 0.36 0.32 0.28 0.11 0.11 0.12 MgO 1.29 1.66 1.08 0.46 0.51 0.31 CaO 3.44 4.08 3.16 2.09 3.61 2.56 Na 2 O 2.21 2.08 2.18 1.06 0.29 0.28 K 2 1.78 2.89 3.68 3.31 1.98 1.41 H 2 O- 1.10 2.41 1.89 1.61 0.43 0.24 H 2 O + 2.90 2.87 3.05 2.12 1.08 1.33 CO 2 1.08 1.56 2.01 2.24 1.87 2.04 P 2 O 5 0.29 0.31 0.30 0.26 0.11 tr. FeS 2 .. 0.65 1.88 3.59 4.69 Total 99.20 99.20 100.23 99.77 100.45 100.62 1. Fresh hornblende dacite, Waihi. 2. Chloritized hornblende dacite, 45 feet from Empire vein. 3. Altered dacite, 30 feet from Empire vein. ) 4. Altered dacite. 15 feet from Empire vein. . . . . > 850-foot level. 5. Altered dacite, adjoining Empire vein. 6. Replacement ore, Empire vein. Propylitization involves a distinct hydration, caused by the de- velopment of chlorite. Where carbonates are formed, magnesia and lime, especially the latter, are somewhat increased. The per- centages of alkali metals decreases, but only in moderate degree. If sericite has formed, the potassium may be somewhat higher in the altered rock. Within the influence of the vein-form- ing solutions the normal alteration to sericite and adularia asserts itself. The two excellent series of analyses quoted above show a slightly differing trend. At Thames the altered rocks contain 10 or 11 per cent, of carbonates, while at Waihi the carbonates form only one-half of that amount. As in the California gold-quartz veins, this development of carbonates results at Thames in the fixing of calcium, while magnesium shows slight changes. At Waihi there is little change in calcium, while the magnesium has been somewhat reduced. In both DEPOSITS FORMED NEAR THE SURFACE 483 places there is strong leaching of sodium and progressive accumu- lation of potassium, except that at Waihi the potassium finally diminishes in the highly quartzose vein material. Iron in ferric and ferrous state is converted to pyrite, but the total iron is not much increased. At Thames, where carbonates are abundant, the silica tends to decrease; at Waihi the opposite is true. In both places there is an apparent decrease in alumina, and also a remarkable and unmistakable leaching of titanium, phosphorus, and manganese, as has also been noted by Spurr at Tonopah. Mineralogically the alteration near the vein results in sericite, calcite, siderite, pyrite, quartz, and adularia, the last mineral in places forming pseudomorphs after soda-lime feldspars, while it also occurs in small fissures. The adularia (valencianite) from Waihi was analyzed by Finlayson and found to contain 11.25 per cent. K 2 O and 4.11 per cent. Na 2 O, while the material from Silver City, Idaho, and Tonopah, Nevada, previously ex- amined yielded only a very small quantity of Na2O. Stilbite and laumontite have been identified in the altered rocks of Waihi, 1 and analyses 4 (Waihi) and 5 (Thames) suggest the possibility of their presence. Finlayson does not accept Spurr's view that the vein-forming waters, filtered through rock masses, caused propylitization, but thinks that this alteration is due to solutions or gases rich in CO2, which permeated the rocks immediately after solidification; the sericite-pyrite carbonate rock along the veins, according to Finlayson, is caused by ascending solutions of a different class. Metasomatic Processes at Tonopah. During the alteration of the trachyte near the veins at Tonopah, Nevada, 2 biotite and horn- blende, have usually been completely destroyed; their outlines are marked by aggregates of sericite, quartz, pyrite, and siderite, the latter two often crystallizing together. The primary an- desine-oligoclase has changed to quartz and sericite or to adularia; the latter two are not often associated in the same specimens. The microlitic groundmass is largely altered to fine-grained quartz with fibers of sericite; pyrite and siderite are disseminated. Apatite and zircon are residual minerals. Kaolin, when present, is believed to result from the alteration of sericite by descending solutions. 1 P. G. Morgan, Trans., Aust. Inst. Min. Eng., vol. 8, 1902, p. 186. 2 J. E. Spurr, Prof. Paper 42, U. S. Geol. Survey, 1905; Econ. Geol., vol. 10, 1915, pp. 713-769. 484 MINERAL DEPOSITS At a distance from the larger veins a propylitic type of altera- tion appears, in which calcite and chlorite, together with pyrite and siderite, are the important minerals. The feldspars are altered to calcite with a little quartz; epidote is not abundant. There are transitions between the propylitic and the sericitic alteration, and according to Spurr they were produced by the same waters. Near the veins these waters introduced silica, potassium, and metallic sulphides; as they penetrated farther ANALYSES OF FRESH AND ALTERED ROCKS, TONOPAH, NEVADA - - 1 2 3 4 5 SiO 2 67 69 55 60 72 98 73 50 91 40 TiO 2 .. 72 44 47 07 A1 2 3 Fe 2 O 3 17.67 2 43 16.70 2 23 " 14.66 1 01 14.13 1 51 4.31 0.77 FeO 80 3 51 16 26 0.11 MgO CaO 0.88 45 2.60 4 27 0.33 18 0.21 12 0.18 none BaO 12 19 02 Na 2 O 2 54 4 08 none 24 06 K 2 O H 2 O-. . 5.11 3.17 88 6.03 97 5.11 1 07 1.68 46 H,0 + C0 2 P2O B 3.06 2.76 28 2.95 none 16 2.81 none 09 0.98 none 04 97.57 99.98 99.87 '99.71 2 100.08 1. Partial analysis of relatively fresh "Mizpah" trachyte. Booth, Garrett and Blair, analyst. 2. Altered andesite, Siebert shaft. Propylitic alteration to quartz, calcite, chlorite, and sericite. George Steiger, analyst. 3. Altered trachyte, Mizpah mine. No original minerals remaining. Sericitic alteration. George Steiger, analyst. 4. Altered trachyte, Mizpah Hill. Typical alteration to adularia with a little sericite. George Steiger, analyst. 5. Ore material of Mizpah vein. Dense quartzose rock mixed with kaolin-like material. Silicified trachyte. George Steiger, analyst. 1 Also 0. 17 SO 3 and 0.03 S. 2 Also 0.12 ZrO 2 and 0.06 MnO. DEPOSITS FORMED NEAR THE SURFACE 485 from these channels their metal contents were exhausted, while silica and potassium were still introduced; finally only carbon dioxide and hydrogen sulphide were left in the cooling waters, which, therefore, had little to precipitate and small power of abstracting. The wall rock acted as a screen for the traversing solutions. As noted above, these views are not entirely accepted by Finlayson. The most prominent features of the alteration, as shown by analyses, are the almost complete removal, adjacent to the veins, of ferrous iron, calcium, magnesium, and sodium and the partial removal of ferric oxide. Even the resistant apatite and rutile seem to have been dissolved to some degree, as shown by the relations of phosphorus and titanium. On the other hand, there is a decided increase of silica, and the potassium has in- creased. There is a moderate hydration, but no carbonates appear. A later mineralization, which affected the later andesite at Tonopah, is materially different; the waters by which it was effected appear to have contained practically no gold and silver. The course of the alteration 1 involves no silicification and practic- ally no change in calcium. The sodium is, as before, almost wholly removed, and likewise a large part of the potassium. Carbonates are present in abundance, with pyrite, and some zeolite is prob- ably present, possibly also some talc and hydrargillite. The Development of Kaolin. It has been stated above that kaolin in the altered rocks of mineral deposits results mainly from the leaching by surface waters containing free sulphuric acid, and that this mode of alteration is frequently superimposed upon the products of chloritic and sericitic alteration by ascend- ing waters. The sulphuric acid attacks and removes all calcium, magnesium, sodium, and potassium; and the final result is a mix- ture of kaolin and quartz. Below the influence of free oxygen sul- phides may be deposited with the kaolin; pyrite, more frequently marcasite in arborescent forms, chalcocite, covellite and rich silver ores like argentite and stephanite may also develop (see discussion of sulphide enrichment, Chapter 31). In places for instance, at De Lamar, Idaho this kaolin may contain much gold in extremely finely divided state, undoubtedly concentrated by secondary reactions. 1 Prof. Paper 42, U. S. Geol. Survey, 1905, p. 241. 486 MINERAL DEPOSITS Metasomatic Processes at Silverton, Colorado. The process of kaolinization is well described in Ransome's report on the Silverton district, Colorado. 1 Crystallized kaolinite a rare occurrence was found in the National Belle mine, but is here, too, later than the ore. The normal alteration at Silverton is of propylitic aspect, changing near the veins to sericitic facies. In a series of rocks occurring in the Silver Lake basin the andesite breccia 150 feet from the vein is only slightly altered by the destruction of the dark silicates and by the beginning of replacement of feldspars by sericite (perhaps with some kaolin) and calcite. At 100 feet from the vein the quantity of chlorite and calcite increases. Fifty feet away from the vein the breccia structure is still visible, but the rock is wholly re-crystallized to quartz, chlorite, sericite, calcite, and rutile, with residual apatite. Two feet from the vein there is less chlorite, and the rock consists mainly of sericite and quartz, with some grains of galena. Close to the wall there is but little chlorite, and considerable pyrite has been introduced. This general process corresponds fairly closely to that at Tonopah. In the same region, at Red Mountain, the alteration of the rocks is carried to its ultimate conclusion. The mine waters show the presence of free sulphuric acid and alumina and the white kaolinized rock at the surface shows the following com- position calculated from the analysis. COMPOSITION OF ALTERED ROCK AT RED MOUNTAIN, NEAR NATIONAL BELLE MINE Quartz GO . 9 Kaolinite 26 . 3 Pyrite 5.6 Diaspore 3.8 Sericite 0.6 Apatite 0.6 Rutile 0.6 98.4 A still more advanced silicification is shown by the following analysis of an altered rock at the White Cloud Mine in the same region, from which the mineral composition of the rock may be calculated as follows: 1 F. L. Ransome, Economic geology of the Silverton quadrangle, Bull. 182, U. S. Geol. Survey,^1901. DEPOSITS FORMED NEAR THE SURFACE 487 Quartz , 78.5 Kaolin minerals 16.0 Pyrite... 3.4 Rutile 0.6 Sulphates (Fe, Ca, Ba) 1.3 99 8 This extreme mode of alteration by sulphuric acid solutions results in the almost total elimination of calcium, magnesium, alkali metals, and carbon dioxide; its operation is similar to that of weathering by oxygenated waters, although in that process, of course, pyrite cannot form. It differs radically from the serici- tization and carbonatization described above. Summary. Summing up we may say that sericite and car- bonates mark the alteration by ascending metalliferous solutions at intermediate and moderate depths; that nearer the surface adularia appears as an important metasomatic mineral in addi- tion to sericite; and that still nearer the surface or under the influ- ence of descending sulphuric acid solutions we find kaolin, or, wher.e sulphuric acid is present in abundance, alunite. QUICKSILVER DEPOSITS The Ores and Their General Occurrence. The principal quicksilver ore is cinnabar (HgS), which contains 86.2 per cent, mercury. A black modification of this mineral, called meta- cinnabar, rarely occurs in large amounts and is probably, a sec- ondary sulphide deposited by descending waters. 1 Native quicksilver, silver and gold amalgam, calomel (HgCl), quicksilver oxide (montroydite), and several oxychlorides are evidently secondary minerals, developing from the sulphide (p. 892). Primary but rare minerals are the black telluride of quicksilver, coloradoite; the selenide, tiemannite; the sulphoselenide, onofrite; and other still rarer combinations of the selenides of copper, lead, and quicksilver. Mercurial tetrahedrite is not uncommon, and some varieties contain as much as 17 per cent, quicksilver, although the percentage is usually much smaller. In smaller quantities this metal is also sometimes present in other minerals, for instance, in the native silver of Kongsberg, Norway, and in the dyscrasite of the silver-bearing veins of Cobalt, Ontario (p. 626). The occurrence of quicksilver minerals is by no means confined 1 E. T. Allen and J. L. Crenshaw, The sulphides of zinc, cadmium and mercury, Am. Jour. Sci., 4th ser., vol. 34, 1912, pp. 367-383. 488 MINERAL DEPOSITS to any certain kind of deposits or to any given age or epoch of metallization. However, such minerals are not known to occur in deposits of distinctly igneous origin, nor in pegmatite dikes, nor in veins of the deepest zone. High temperature is evidently unfavorable for their development. The most noteworthy occurrence is that of coloradoite in the gold telluride veins of western Australia, which contain, among other minerals, mag- netite and tourmaline, indicating deposition at fairly high tem- perature. In gold-bearing quartz veins of the ordinary type, believed to have been formed at a considerable depth, but at con- siderably lower temperature and pressure than pegmatite dikes, cinnabar is not an uncommon mineral. It occurs in several of these veins in California, 1 as well as in the similar veins of central Idaho, and is frequently found in the placers derived from the erosion of these veins, as at Stanley Basin and Warren, Idaho. In northeastern Oregon the gold-quartz veins contain mercurial tetrahedrite, as well as secondary cinnabar formed from that mineral. 2 In the placers below the veins of Susan- ville, in the same region, pebbles showing masses of cinnabar in- closed in massive white vein quartz have been found. One often finds apparently reliable statements that during the process of amalgamation and refining of the gold from such deposits more quicksilver was recovered than was added for metallurgical purposes. In small quantities cinnabar occurs in the lead and zinc deposits of Monteponi, in Italy, and at Santander, in Spain. Many occurrences of mercurial tetrahedrite in Europe and South America have been described. In few of these deposits are the mercurial minerals abundant enough to constitute an ore, and in the majority of the deposits formed at a considerable depth the metal is apparently entirely absent. The commercial quicksilver ores are practically con- fined to a small and well-defined group of deposits, which will be described in the following pages and which are of particular interest because their genesis can be fairly accurately ascertained. A scant association of ore minerals characterizes these deposits. Besides cinnabar and metacinnabar, as well as a few minerals derived from the decomposition of the sulphide, they contain 1 H. W. Turner, Am. Jour. Sci., 3d ser., vol. 47, 1894, p. 467. H. D. McCaskey, Min. Res., U. S. Geol. Survey, pt. 1, 1910, p. 905. 2 W. Lindgren, The gold belt of the Blue Mountains of Oregon, Twenty- second Ann. Rept,, U. S. Geol. Survey, pt. 2, 1901, pp. 604, 708. DEPOSITS FORMED NEAR THE SURFACE 489 almost invariably pyrite or marcasite and frequently stibnite, but rarely any of the sulphides of the base metals so common in ore deposits. Among gangue minerals we have predominatingly opal, chalcedony, and quartz, also calcite and dolomite, more rarely ba- rite, and very seldom fluorite ; zeolites are of exceptional occurrence. As to form and structure the ores occur in irregular and "cham- bered" veins and brecciated zones, often also as "stockworks" of minute seams, or as disseminations in more or less porous rocks. The irregularity and brecciated character of the deposits suggest their development near the surface, a conclusion that is often justified by other geological evidence. As to association the deposits occur in rocks of any kind and any age, but almost always in close connection with effusive rocks or in regions of volcanic activity. Hot springs are frequent in many quicksilver districts, and this conspicuous association has led to the view that the metal is a last product of differentiation in many magmas nearing the surface and that the hot springs take it into solution to deposit it near their points of issue. This theory is strongly fortified by the discovery that many springs in volcanic regions are depositing cinnabar at the present time. Although most of the quicksilver deposits have been formed at a relatively late time and in connection with the eruption of Tertiary and recent lavas, it does not necessarily follow that their development has been confined to late geological time. Older surface eruptions were undoubtedly accompanied in places by the formation of quicksilver deposits, but as these were near the surface they were easily eroded. Attention has already been called to this elsewhere, 1 and it has been suggested that the de- posits at Almaden, in Spain, and those in the German Palatinate may belong to such more remote volcanic epochs. Distribution, Production and Use. Quicksilver deposits are widely distributed, although the main production of the metal comes from a few occurrences. G. F. Becker, who first studied the occurrence of these ores concluded that their distribution follows the main structural lines of the continents, especially in the Pacific region and in the Alpine-Himalayan chain. While this is apparently true, it is more correct to say that the ores fol- low the belts of Tertiary and Quaternary eruptions, especially along the important fracture lines of the globe. The Coast Range of California, in which the erogenic move- 1 Beyschlag, Krusch, and Vogt, Die Lagerstatten, etc., vol. 1, 1910, p. 454. 490 MINERAL DEPOSITS ments are largely post-Miocene, contain a belt of quicksilver deposits several hundred miles in length, from which at one time a large production was obtained. 1 A second belt, perhaps less well defined and certainly less productive, extends from north to south over a similar length in western Nevada. The Mexican area, which in spite of comprising many deposits, yields only a slight production, begins in western Texas in the Terlingua district 2 and may be considered to end in the State of San Luis Potosi, Mexico. 3 Farther south, in Peru, 4 quicksilver deposits appear again. The Yauli and Huancavelica districts are best known; the latter at one time was an important producer. In Europe an extensive region in Italy, Austria, and adjacent countries contains quicksilver deposits; this area includes the deposits of Tuscany, 5 Vallalta-Sagron, 6 Idria and vicinity, 7 1 G. F. Becker, The quicksilver deposits of the Pacific slope, U. S. Geol. Survey, Mon. 13, 1888. W. Forstner, The quicksilver resources of California, Bull. 27, California State Min. Bur., 1903. H. D. McCaskey, Mineral Resources, U. S. Geol. Survey, pt. 1, 1910- 1916, particularly in issue of 1911, with literature. 2 W. B. Phillips, The quicksilver deposits of Brewster County, Texas, Econ. Geol., vol. 1, 1906, pp. 155-162. H. W. Turner, The Terlingua quicksilver deposits, Econ. Geol., vol. 1, 1906, pp. 265-281. 3 J. .D. Villarello, Genesis de los yacimientos mercuriales de Palomas y Huitzuco, Mem. Soc. Ant. Alzate, vol. 20, 1903, pp. 95-136. J. D. Villarello, Descripci6n de los criaderos de mercurio de Chiquilistan, Jalisco, Idem, vol. 20, 1904, pp. 389-397. P. A. Babb, Dulces Nombres quicksilver deposit, Mexico, Eng. and Min. Jour., Oct. 2, 1909. F. J. H. Merrill, The mercury deposits of Mexico, Mining World, vol. 24, 1906, p. 244. 4 A. F. Umlauff, El cinabrio de Huancavelica, Boi. 17, Cuerpo de In- genieros de Minas, Lima, 1904, p. 61. 6 V. Spirek, Die Zinnobererzvorkommen am Monte Amiata, Zeitschr. prakt. GeoL, 1897, pp. 369-374; idem, 1902, pp. 297-299. B. Lotti, II campo cinabrifero dell' Abbadia, etc., Rass. Min., vol. 7, 1898, No. 11; Zeitschr. prakt. Geol, 1898, p. 258. See also Rass. Min., vol. 17, 1902, No. 10. 6 A. Rzehak, Die Zinnoberlagerstatte von Vallalta-Sagron, Zeitschr. prakt. GeoL, 1905, pp. 325-330. 7 F. Kossmat, Ueber die geologischen Verhaltnisse des Bergbaugebietes von Idria, Jahrb. K. k. geol. Reichsanstalt, vol. 49, 1899, pp. 259-286. Geologisch-bergmannische Karten, etc., von Idria. Text by Plaminek. Published by the Agricultural Department, Vienna, 1893. (Literature.) DEPOSITS FORMED NEAR THE SURFACE 491 Avala, 1 in Servia, and less important occurrences in Bosnia. Isolated yet highly productive deposits occur in Almaden, in Spain. 2 Some deposits have been found on the western side of the Pacific, mainly in Japan, China, Borneo, Australia, and New Zealand. Before the war, in 1913, the world's production of quicksilver was 124,654 flasks at 75 pounds. Of this Spain produced 43,799, Italy 29,513, Austria 26,720 and United States 20,213. In 1917, the domestic production was 36,315 flasks. California yielded 24,251 flasks. The average price in 1913 was $40 per flask. In 1917 the average for the year was $106. The principal use of quicksilver is for gold amalgamation, drugs, paints and mercurial fulminate (Hg (ONC) 2 ) an explosive used for priming shells. Geological Features. The comparative youth of the deposits is attested by the fact that many of them are found in sedimen- tary or volcanic rocks of Tertiary or Quaternary age. They are not confined to these rocks, however, and may, in fact, occur in rocks of any composition or age. Sandstones, shales, limestone, serpentine, granite, andesite, rhyolite, or basalt may harbor the ores, and the character of the surrounding rocks seems to have little influence on the value of the deposits. The California belt contains ores in Jurassic, Cretaceous, and Tertiary sandstones and shale, in serpentine, and in late Tertiary or Quaternary basalt and andesite. In the Nevada belt the ores occur in Triassic strata in Paleozoic limestone and dolomite or more commonly in rhyolite, probably of middle Tertiary age. In Texas and Mexico the ore-bodies are in Cretaceous strata or in the Tertiary andesite, basalt, and rhyolite which break through them. The Peruvian deposits are in Jurassic beds or in Tertiary volcanic rocks. In the Adriatic region of Europe the ores occur in rocks of many kinds : In Tuscany, Mesozoic and Tertiary limestones and sandstones with trachyte are the enclosing rocks; at Idria, the 1 H. Fischer, Die Quecksilberlagerstatten am Avala-Berge in Serbien, Zeitschr. prakt. Geol, 1906, p. 245. 2 There is no modern and detailed description of this important deposit. The best account is found in Beck's "Lehre von den Erzlagerstatten," vol. 1, 1909, pp. 519-522. 492 MINERAL DEPOSITS disturbed beds of the Alpine Triassic; at Avala, the serpentine and probably Cretaceous limestone cut by trachytic dikes. De Launay 1 has shown that these Adriatic ores coincide in their extension with Tertiary eruptives, and that in all probability, even where these eruptives are locally absent, as at New Idria, the deposits owe their origin to the after-effects of this igneous activity in the form of ascending springs. In the Donetz basin in southern Russia 2 the cinnabar ores lie in Carboniferous strata and have no apparent connection with igneous rocks. Mineralogy of Quicksilver Ores. Cinnabar while usually massive sometimes forms well-defined but small crystals; these are especially common in porous rocks like sandstone and tuff. Aside from pyrite and the more common marcasite the ore mineral most generally accompanying cinnabar is stibnite;many stibnite veins, it should be added, also contain some cinnabar. Quartz is usually present, but far more commonly the silica appears as chalcedony or opal. In the California occurrences opal is particularly abundant and here, as at Avala, much of it is a product of the replacement of serpentine. In California the cinnabar is not often found in the opal itself, but rather in the veinlets of quartz or chalcedony traversing it. Calcite is not an uncommon gangue material, and in the Coast Ranges of Cali- fornia many of the deposits are accompanied by calcium-mag- nesium carbonates derived by replacement from serpentine or allied rocks. Among the sulphates barite is fairly abundant, and at most places there is also more or less gypsum, which may often be a product of primary deposition, although it would naturally also be generated by the effect of decomposition of pyrite in a cal- careous gangue. Fluorite is rare, but is recorded from Guad- alcazar and Idria. In many deposits, particularly those of Idria and the California belt, hydrocarbons are characteristic; they are probably derived from the adjacent sedimentary beds, but are believed to have exerted some influence in the precipitation of cinnabar from its solutions. Inflammable gases, mainly hydro- 1 L. De Launay, La metallogenie de 1'Italie, International Geological Congress, Mexico, 1906. 2 Zeitschr. prakt, Geol, 1894, p. 427, after Kulibin. Tschernyschew and Loutouguin, Guide des excursions du VII Congress geol. internat., No. 16. 1897, pp. 36-45. DEPOSITS FORMED NEAR THE SURFACE 493 carbons, are reported from several localities in California, notably New Idria. Zeolites are not unknown in quicksilver deposits; chabazite colored by cinnabar is mentioned from Almaden, in Spain, and apophyllite from New Almaden, in California. Alunite has been found associated with opal and cinnabar in rhyolites of Nye County, Nevada, near Bullfrog. 1 Millerite, or sulphide of nickel, is not uncommon in the Cali- fornia occurrences; this metal is probably derived from adjacent bodies of serpentine or peridotite. The rare tiemannite and onofrite were at one time obtained near Marysvale, in southern Utah, and some quicksilver was recovered from such ore. The minerals occurred as fissure filling in limestone, but in a district of volcanism with rhyolitic and andesitic flows. Selenides are also reported from Guadalca- zar, in San Luis Potosi, Mexico. It is probable that careful examination would disclose the presence of selenides at many other places. The mineral livingstonite (HgSb4S7) occurred in considerable quantities at Huitzucq, Guerrero, Mexico. Regard- ing oxidation of quicksilver ores see p. 892. Quicksilver ores should contain at least 0.5 per cent, of the metal. The richest ores are those of Almaden, Spain, which are said to average 8 per cent., while at New Idria ores are mined which contain less than 0.5 per cent. Structure. In their structural relations the majority of quick- silver deposits clearly indicate their origin near the surface. Sharply defined continuous fissure veins occur only exceptionally; far more common are irregular and "chambered" veins that is, fissures accompanied by brecciated masses in which the ore minerals have lodged (Fig. 150). Another common mode of occurrence is as disseminations in porous rocks, like sandstone and tuffs, or again as "stockworks," the ore minerals filling little crevices and fissures in limestone, serpentine, or other rocks. Cinnabar is often deposited in open cavities or in pores or in the soft mud of altered rocks. It does not replace limestone on the scale of the lead deposits, but that replacement has occasion- ally occurred seems to be beyond doubt. A large number of deposits have been found to cease or become impoverished at a depth of a few hundred feet. In contrast to 1 Adolph Knopf, Some cinnabar deposits of western Nevada, Butt. 620, U. S. Geol. Survey, 1915, pp. 59-68. 494 MINERAL DEPOSITS this, the celebrated mines of Almaden, Spain, are said to have found richer ore in depth, and the workings have now attained a depth of 1,300 feet. The deposits at Almaden occur in three beds of steeply dipping Silurian quartzite separated by bituminous slates. In part the cinnabar may occur as filling of the pores of the rock, as G. F. Becker 1 suggests, but Beck 2 has shown convincingly that there has also occurred an actual replacement of the sandstone grains FIG. 150. Diagrammatic vertical cross section of the Redington cinnabar mine, California, showing brecciated ore chambers near surface, changing in depth to more regular fissures filled with cinnabar. Total depth about 600 feet, m, Metamorphic rock; n, Cretaceous sandstone. After G. F. Becker, U. S. Geol. Survey. by the ore mineral (Fig. 151). The ore-bodies are as much as 45 feet in thickness, and the average tenor of the ores is unusually high; they are said to contain 8 per cent, quicksilver. Granite and diabase break through the sedimentary series, but the geological history of the deposit is too imperfectly known to draw safe conclusions as to its age or mode of origin. Almaden is the richest and most productive quicksilver region in the world. The value of its metallic product, according to the handbook of Beyschlag, Krusch, and Vogt, during the period 1564 to 1907, is estimated at 212 million dollars. 1 G. F. Becker, Mon. 13, U. S. Geol. Survey, 1888, p. 399. 2 R. Beck, Lehre von den Erzlagerstatten, vol. 1, 1909, p. 521. DEPOSITS FORMED NEAR THE SURFACE 495 At Idria, in southern Austria, is located another of the great quicksilver mines of the world. The ores are contained in Triassic beds of shale, marl, and dolomite; they are apparently connected with and in part occur in great overthrusts and faults. The ore-bodies, which apparently do not extend below a depth of 1,000 feet, in places follow the stratification and were formerly believed to be of syngenetic origin. The ores are usually desig- nated as "impregnations" in shale or marls, but small veins and stockworks are also found, especially in the dolomite. The age of deposition is certainly post-Cretaceous, probably Tertiary. FIG. 151. Rich ore, Almaden, Spain, showing cinnabar between grains of quartzite and formed by replacement hi quartz, c, Cinnabar; s, sericite; p, pyrite; z, zircon. Magnified 70 diameters. After R. Beck. Schrauf, who has given long study to Idria, believes that the ore occurring in the dolomite is a later migration from the some- what older deposit in the shales. The California region offers types of almost all the various structural developments. In the region north of San Francisco, near Clear Lake, serpentine, radiolarian cherts, altered Franciscan sandstone (Jurassic?), and Cretaceous sandstones prevail; the rocks are greatly shattered and late Tertiary to Quaternary andesites and basalts break through them. The occurrences of 496 MINERAL DEPOSITS cinnabar are numerous, and some of them have yielded a large production, but the ore-bodies generally become impoverished at a depth of a few hundred feet. The deposits form fissure veins largely filled with attrition material, and this is impregnated with cinnabar, pyrite, opaline silica, and calcite. Masses of ore often extend into the country rock from these fissures (Fig. 152). Or again, as in the Great Eastern mine, the ore forms tabular masses between serpentine and sandstone, or pipes in opaline or chalcedonic rocks between the same formations (Fig. 153), or finally it may be developed on the contact of basalt and sand- stone. The Redington mine (Fig. 150) was operated on a large FIG. 152. Vertical cross section through workings of Napa Consolidated mine, California. Irregular veins in horizontal Cretaceous sandstone, widening to chambers along bedding planes. After G. F. Becker, U. S. Geol. Survey. chambered deposit at the surface which was found to be continued below by more regular and narrow veins. Throughout this region hot springs are found in and around the ore deposits. The great mine of New Almaden, in Santa Clara County, south of San Francisco, is opened in shattered greenstone, serpentine, radiolarian chert, and sandstone of the Franciscan series. Con- sidered in detail the ore-bodies are stockworks, but they are arranged along definite fissures and have on the whole a vein- like character. There are two main fissures of varying dip along and from which the ore-bodies extend. The hanging wall is DEPOSITS FORMED NEAR THE SURFACE 497 usually an impermeable, slickensided clay. There are no hot springs and no eruptives in the vicinity. The mine has been opened to a depth of 2,100 feet and a continuous ore-body extended down to the 1,600-foot level. During the last few years little work has been done in the lower levels. At New Idria, at the south end of the Mt. Diablo Range, impor- tant deposits have been worked and the larger part of the produc- tion*of California is now derived from this mine. The rocks are disturbed greenstones and sandstones of the Franciscan series, imconformably covered by tilted Chico (Cretaceous) and Tejon FIG. 153. Vertical cross section of the Great Eastern mine, California, showing pipe of cinnabar contained in opaline replacement gangue. After G. F. Becker, U. S. Geol. Survey. (Eocene) sandstone. The ores appear in three forms as normal veins, as irregular stockworks, and as impregnations in sand- stone. The mine is opened by tunnels, the lowest level being at a vertical depth of 1,060 feet. There are no volcanic rocks in the immediate vicinity. In the Terlingua district, Texas, near the Mexican boundary, the ores are found in the Upper Cretaceous shales and the Lower Cretaceous limestone. Volcanic rocks are represented by sheets, dikes, and flows of andesite, rhyolite, and basalt. In the lower 498 MINERAL DEPOSITS limestones the ores are mainly in nearly vertical calcite veins, or in lodes of friction breccia (Fig. 154). The other associated minerals are chalcedony, gypsum, aragonite and pyrite. Genesis. The uniform character of the quicksilver deposits points to a common genesis for all of them. The earlier belief that the ores were products of sublimation is generally abandoned, for the usual mode of occurrence, with minerals of aqueous origin, such as calcite, opal, chalcedony, and often barite, is decidedly opposed to such a view. Becker has pointed out that, as the character of the enclosing rocks has little influence on the deposits they are most probably derived from a common, deep-seated source. Their structure indicates deposition near the surface, Scale 100 200 300 Feet FIG. 154. Vertical cross section of California Hill, Terlingua, Texas, showing cinnabar veins with large ore-bodies below impervious shale. After H. W. Turner. as does also the physiographic evidence at many places for instance, where the ore appears in the crevices of Quaternary and little-eroded lava flows. When it is noted that hot springs and volcanic surface flows are present in almost all regions of importance (except Almaden in Spain, Idria in Austria, and Nikitowka in Russia), and that cinnabar in considerable quantities is associated with hot spring deposits, or is actually found deposited by hot springs, the argu- ment becomes very strong indeed that such solutions have formed the majority of the deposits. For the few deposits that have no such clear connection with volcanic rocks the characteristic DEPOSITS FORMED NEAR THE SURFACE 499 mineral association still holds good, and \ve are forced to the hypothesis that volcanism and hot-spring action are the causes of these also, thoughthe products of the igneous activity may have failed to reach the surface and the hot springs may have subsided. The evidence relating to cinnabar deposited by hot springs is summarized in the following paragraphs. At Steamboat Springs, in Nevada, near the California bound- ary, cinnabar is contained in the hot ascending sodium chloride waters, together with antimony, arsenic, and sulphur, and is actually being deposited in the sinter. 1 Close by, but at a higher level, is a low-grade quicksilver deposit in decomposed granite, and this in all probability was also formed by the same springs when issuing at a higher level. Underneath the sinters of the present springs the gravels contain crystallized stibnite and pyrite. At Sulphur Bank, 2 in the California quicksilver belt, Le Conte, Christy, Rising, Becker, and Posepny have studied the deposition of cinnabar and sulphur by ascending hot sodium carbonate and borate waters and have all arrived at the conclusion that such deposition, together with that of pyrite and opal, is actually taking place. The Cretaceous sandstones and associated Franciscan metamorphic rocks are here overlain by flows of both normal and glassy basalt and by cinder cones, pointing to very recent eruption. The hot springs have altered and bleached the basalt. Sulphur is deposited at the surface by the oxidation of H 2 S, or by reaction between S0 2 and H 2 S. Below the super- ficial deposit of sulphur cinnabar is found in the basalt, as well as in the underlying shales and sandstones; it occurs mostly in veinlets and joints together with the pyrite and opal above mentioned. (Cfr. p. 113.) The Rabbit Hole sulphur deposit, in Humboldt County, Nevada, described by G. I. Adams, 3 is evidently a product of hot springs, and near it are considerable areas of rhyolite. The rocks are silicified, and opal, alunite, gypsum, and some cinnabar are present as associated minerals. 1 G. F. Becker, Man. 13, U. S. Geol. Survey, 1888, Chapter XL s J. LeConte and W. B. Rising, The phenomena of metalliferous vein formation now in progress at Sulphur Bank, Cal., Am, Jour. Sci., 3d ser., vol. 24, 1882, pp. 23-33. G. F. Becker, op. tit.. Chapter VII. F. Posepny, The genesis of ore deposits, 2d ed., 1902, pp. 32-36. 3 Butt. 225, U. S. Geol. Survey, 1904, pp. 497-502. 500 MINERAL DEPOSITS In the Hauraki peninsula of New Zealand, near OmapereLake, 1 where basalts overlie Mesozoic shales and sandstones, mercury and cinnabar have been found in the deposits of the hot springs at several places. E. Cortese 2 reports the occurrence of cinnabar in connection with sulphur deposits which result from still active hot springs in Chaguarama Valley, Venezuela. The cinnabar occurs in Ter- tiary bleached sandstone, together with pyrite. Borax deposits are also said to occur in the same locality. A careful investigation would doubtless disclose the presence of cinnabar in many other spring deposits. If it is found in more than traces the best way to test such material, as well as ores, is by the miner's pan, in which the bright-red grains of cinnabar show conspicuously. Quicksilver is apparently contained in hot-spring waters carry- ing sodium carbonate, sodium chloride, or sodium borate; some- times all three salts as well as carbon dioxide and some hydrogen sulphide are present. Near the surface these springs may become acid owing to the oxidation of hydrogen sulphide. Regarding the mode in which mercury is carried in solution, Becker's views, 3 based on the laboratory experiments of W. H. Melville, still appear to furnish the best explanation. While the solubility of mercuric sulphide in alkaline compounds con- taining sulphur had long been recognized, the evidence was to some degree conflicting. Becker showed that mercuric sulphide is freely soluble in solutions of sodium sulphide, as well as in a mixture of Na2 and NaOH, and also in warm sodium sulphydrate (NaHS). When neutral sodium carbonate is treated with hydrogen sulphide, sodium sulphydrate and probably also sodium sulphide will form; these dissolve mercurial sulphide, and double salts of the general formula HgS.wNazS doubtless form. Incidentally it was found that the same reagents would dissolve metallic gold, pyrite, sphalerite, and cupric sulphide. The solubility of the 1 1. M. Bell and E. de C. Clarke, Bull. 8, New Zealand Geol. Survey, 1909, p. 87. A. Liversidge, Jour. Roy. Soc. N. S. W., vol. 2, p. 262. J. Park, Trans., N. Z. Inst. Min. Eng., vol. 38, 1904, p. 27. Andre P. Griffiths, The Ohaeawai quicksilver deposits, Trans., N. Z Inst. Min. Eng., vol. 2, 1898, p. 48. 2 Eng. and Min. Jour., Nov. 10, 1904. 3 G. F. Becker, Mon. 13, U. S. Geol. Survey, 1888, Chapter XV. DEPOSITS FORMED NEAR THE SURFACE 501 sulphides of arsenic and antimony under similar conditions is well known. It is therefore easy to perceive that a spring water containing sodium carbonate and hydrogen sulphide would form a suitable solvent for the compounds mentioned. The precipitation would be easily effected by oxidation of the water and the development of free acids, by dilution, by cooling, or by the presence of organic or ammoniacal compounds. That the latter two agents are active in many cases there is little doubt. Relation to Other Ore Deposits. Although the cinnabar de- posits form a well-defined group, they are not to be separated entirely from other classes of ore deposits. Some of them con- tain other metallic minerals, and there are many that show a transition to the stibnite and arsenical veins. The Nevada belt especially furnishes many instances of a close relationship to gold and silver bearing veins on one hand and to stibnite veins on the other hand. It is true, however, that no cinnabar deposit has yet been found to change gradually into ores of different character as depth is attained. No deposits have been worked below a depth of 2,000 feet vertically beneath the croppings. There is little doubt that these ores were formed from the most volatile parts of the magmas, carried in solution by ascending waters until they came close to the surface. But the irregular distribution of the deposits and their entire absence from large areas of volcanism form problems that are yet unsolved. STIBNITE DEPOSITS Mineralogy, Production and Uses. Stibnite (Sb 2 S 3 with 71.4 per cent. Sb) is the principal ore mineral of antimony. Its oxida- tion near the surface results in various oxides (p. 899) of yellowish or white color such as senarmontite (Sb2Og), cervantite (Sb02), and stibiconite (H 2 Sb 2 O 5 ). While stibnite occurs in many depos- its in small quantities, especially in quicksilver ores, it is the characteristic and dominant mineral in the stibnite veins where it is accompanied by quartz gangue and a scant amount of other sulphides, such as arsenopyrite, realgar, pyrite and zincblende, more rarely jamesonite and similar sulphantimonides. Such ores often also carry gold and the association of the stibnite with some gold quartz veins has often been noted. For the purpose of making pure antimony the presence of arsenic and copper is undesirable. 502 MINERAL DEPOSITS In past years the production of antimony has not been great owing to its low price and limited usefulness. Since the great war began the price has increased to 17 cents per pound. The world's production of antimony metal may be estimated to 50,- 000 metric tons. The supply is mainly obtained from China, from Central France, and from the State of Queretaro, Mexico. In the United States but little pure antimony is produced though under stress of war the production of such ores has risen to several thousand tons. Antimony is used for bearing metals, type alloys, shrapnel bullets and its salts find a varied use in the industries. The sulphide is used in pyrotechnics. Another source of antimony is in the replacement deposits and veins containing mainly galena but associated with tetrahedrite and more rarely with jamesonite, bournonite, boulangerite and other lead sulphantimonides. As a rule these are related to intrusive action and the small amount of antimony contained is recovered as "hard lead" or antimonial lead in the smelting operations. From 2,000 to 3,000 tons of this alloy is produced annually in the United States. No antimonial mineral is known to occur in magmatic deposits; they are certainly rare in the contact-metamorphic and other high temperature deposits though in these jamesonite, tetrahe- drite and stibnite have been occasionally reported (p. 738). Occurrence. The stibnite veins have wide distribution but are rarely rich. They are in part formed near the surface, but many deposits are of more deepseated origin and occur in or near intrusive rocks. To the former type belong the stibnite veins with a gangue of fine-grained and drusy quartz which intersect flows of rhyolite and basalt in western Nevada. The antimony sulphide is as a rule beautifully crystallized in acicular and pris- matic forms; it is often accompanied by a little pyrite, zinc blende, and arsenopyrite, sometimes also by tetrahedrite and cinnabar. Such veins carry a little silver and less gold. The intimate re- lationship of these veins with the gold and silver veins proper is, however, shown by the occurrence in one of them, at National, Nevada, 1 of a shoot of remarkably coarse gold of the electrum variety. Stibnite veins of uncertain affiliations are found in central Arkansas but are of no great importance. The veins follow the steep stratification of Carboniferous shale and sandstone, and 1 W. Lindgren, Bull. 601 ; U. S. Geol. Survey, 1915. DEPOSITS FORMED NEAR THE SURFACE 503 the stibnite fills the spaces between the quartz combs. 1 Stibnite shows a marked tendency to form replacements in limestone and shale. Such deposits in Eocene shale below a thick series of ande- sites have been described from southern Utah. 2 They are un- doubtedly hot spring deposits. Of such nature are also the deposits at Pereta, in Tuscany, where the mineral is associated with realgar and cinnabar and occupies veins, seams, and irregular pockets in Tertiary calcareous and detrital rocks. The country rock is in part silicified, in part altered to gypsum or alunite, and exhalations of hydrogen sulphide testify to the recent age of the deposit. Beck 3 describes important deposits of stibnite at Kostainik, in Serbia, where the mineral occurs in nests and veins in trachyte or in Triassic clay shales, but also as metasomatic bodies replacing the beds along the contact of limestone and shale. The gangue is a drusy fine-grained quartz. The stibnite veins of Japan, renowned for their beautiful crys- tals, are found in Mesozoic and Paleozoic rocks but little is known about their affiliations. The deposit of Djebel Kami mat, 4 in Algeria, containing sen- armontite, and that of Altar, 5 Sonora, from which stibnite is reported as the principal ore mineral, appear both to be replace- ment deposits in limestone. At the Algerian locality the replacement veins spread out in Cretaceous sediments, while at Altar the ore is said to occur in Carboniferous limestone. Both deposits are probably oxidized replacements of stibnite. Stibnite veins affiliated with intrusive rocks differ but little from the deposits described above. They are known from central France, where narrow veins intersect granite and surrounding schist. Similar deposits are not uncommon elsewhere for in- stance, in Kern County, California, where the quartz veins also contain gold, in the Coeur d' Alene district and elsewhere. Stibnite is very common in Alaska and generally is found in gold-quartz veins. A. H. Brooks 6 who enumerates 67 occurrences suggests that the stibnite may have been introduced in older gold-quartz veins during a later and Tertiary mineralization. 1 F. L. Hess, Bull. 340, U. S. Geol. Survey, 1908, pp. 241-256. 2 G. B. Richardson, idem, pp. 253-256. 3 R. Beck (after W. von Fircks), Zeitschr. prakt. Geol, 1900, pp. 33-36. 4 L. De Launay, Gttes mine'raux, 1, 1913, p. 772. 6 E. T. Cox, Am. Jour. Sci., 3d ser., vol. 20, 1880, pp. 421-423. 6 Antimony deposits of Alaska, Bull. 649, U. S. Geol. Survey, 1916. 504 MINERAL DEPOSITS GOLD -QUARTZ VEINS IN ANDESITE Transylvania. 1 In northwestern Hungary and in adjoining parts of Transylvania gold-bearing veins of Tertiary age have been developed after eruptions of andesites and dacites. A mining industry, begun centuries ago, still flourishes in this region. The literature is very extensive, and only a few deposits can be mentioned here as examples. The geological formations in the western part of the gold- mining region of Transylvania consist of Mesozoic melaphyres, Cretaceous shales and sandstones, and Miocene sediments, all penetrated by late Tertiary eruptions of andesites and dacites. The igneous rocks appear as lava flows, tuffs, and volcanic necks. The veins near Brdd, at present the most productive district, fill well-defined steeply dipping, in places branching fissures which generally intersect volcanic rocks or Cretaceous sediments. The simple veins are as much as 1 meter in thickness; the lodes attain a thickness of 10 to 20 meters. The deposits have been worked to a depth of about 270 meters. The fissures are tec- tonic, not contraction joints. They are of Miocene age. The surrounding rocks, particularly the andesite, have suffered extensive propylitization, the femic minerals being extensively decomposed, while feldspars remain fresh. Pyrite is not common except near the veins. Calcite is abundant. Schumacher does not believe that propylitization is caused by "intensive pene- tration by gases from the not yet wholly solidified intrusions," an opinion expressed by Stelzner and Bergeat. He nevertheless considers the process distinctly earlier than the veins and inde- pendent of them. The alteration continues to the greatest depths attained. " Kaolinization " near the veins is a wholly 1 Bela von Inkey, Nagyag und seine Erzlagerstatten, Buda-Pest, 1885. Bela von Inkey, De la relation entre l'6tat propylitique des roches andesitiques et leur filons mineYaux, Internat. Geol. Congress, Mexico, 1906. M. v. Palfy, Das Goldvorkommen im siebenburgischen Erzgebirge, etc. Zeitschr. prakt. Geol., 1907, pp. 144-148. C. Semper, Beitrage zur Kenntniss der Golderzlagerstatten des sieben- burgischen Erzgebirges, Abh. K. preuss. geol. Landesanstalt, Neue Folge, Fasc. 33, 1900. F. Schumacher, Die Golderzlagerstatten der Rudaer Zwolfapostel- Gewerbschaft zu Brdd im Siebenburgen, Zeitschr. prakt. Geol, 1912, pp. 1-85. DEPOSITS FORMED NEAR THE SURFACE 505 different process according to Schumacher and is superimposed upon propylitization. Kaolin and sericite are both present in the altered rock, but the possible influence of descending waters on kaolinization is inadequately treated. The alteration of the wall rock in a vein 0.5 meter thick extends only about 10 centi- meters from the vein, but many smaller veins have proportion- ately wider zones of alteration. An older set of veins is composed of clastic material of shale and igneous rocks ("Glauch veins")- They are interpreted as having been filled by ascending liquid muds. Similar veins in the Silverton district, Colorado, have been described by Ransome. a/ FIG. 155. Rich gold-bearing quartz, Brad, Transylvania, a. Granular quartz; b, gold between grains; c, plates of gold accompanied by crushed quartz. Magnified. After F. Schumacher. The gangue of the ore-bearing veins is composed of quartz (rarely chalcedony), calcite, rhodochrosite, and barite, a fre- quently recurring association in veins of this class. The quartz is usually fine-grained, sometimes drusy, cellular, or honey- combed, but not amethysthine. Pyrite in small crystals l is abundant; the concentrates contain 10 grams of gold and 69 grams of silver per ton, while the pyrite in the country rock contains 7 to 15 grams of gold and 10 to 15 grams of silver per ton; both kinds are therefore poor in gold. Marcasite^has^been notedjin only one mine, where it occurs on quartz, associated 506 MINERAL DEPOSITS with free gold. Zinc blende is associated with pyrite and is poor in gold but contains more silver than the pyrite. Chalcopy- rite and galena where present are poor in gold but contain several hundred grams of silver per ton. The principal ore mineral is native gold, which occurs com- monly in crushed quartz or in little fissures (Fig. 155), or as sheets or wires between the quartz combs of veinlets. Some of it is found in coarse quartz and is apparently older than the quartz or of contemporaneous origin. It occurs also in sheets along the cleavage planes of calcite and in lumps or nodules in clay. In part it is therefore of relatively late origin. The gold contains 28 per cent, silver and the ores average 10 grams of gold per ton. Tellurides and rich silver minerals are rare. boale Viktor Adlt ' 10 20 30 40 50 00 70 SO 90 MetetS FIG. 156. Section of stoped area in vein at Brad, Transylvania, showing pockets of rich ore; also rich shoots following intersections with barren veins. After F. Schumacher. The structure of the veins is irregularly massive, though in places crusted, banded, and drusy. Brecciated structures are common. Small prismatic and rectangular pseudomorphs of quartz are considered as replacements of gypsum but strongly resemble the similar casts of celestite at Cripple Creek, Colorado. The ore-shoots are irregular; sometimes they are narrow but extend with steep dip for 100 or 200 feet vertically. Shoots often occur at junctions and intersections (Fig. 156). At a junc- tion of two veins with a narrow pyritic seam was found a pocket from which in one day gold weighing 55 kilograms was extracted. Near the surface the veins were poor. The richest zone extended from about 100 meters below the surface down to a depth of 320 meters. The remarkable dependence of the rich ore on narrow seams of pyrite is evident and recalls analogous conditions DEPOSITS FORMED NEAR THE SURFACE 507 in the Thames district, New Zealand, and the "indicators" of Victorian quartz mines and many other gold deposits (Fig. 157). B. von Inkey held that the gold was concentrated by leaching from the country rock. Schumacher believed that it was depos- ited by ascending hot waters in the vicinity of necks of intrusive rocks. Beyond the intrusive necks the veins persist but contain only gangue minerals. M. Dittrich examined fresh and propy- litic andesite, using the cyanide process, but found gold in neither. While much of the gold is distinctly later than the gangue it is difficult to say whether we have to deal here with the effect of descending waters or with the last phases of vein formation. A similar problem is offered in the rich pockets of veins at Thames, in New Zealand; in neither place is enrichment by descending waters satisfactorily proved. FIQ. 157. Pockets of native gold (a) in quartz vein (g) along intersec- tions with pyritic seams (&). After F. Schumacher. Hauraki Peninsula, New Zealand. 1 The Hauraki region in the northern island of New Zealand is richly mineralized in sev- eral districts. Its rocks consist mainly of andesite and dacite flows of Eocene or Miocene age covered by Pliocene rhyolites. A production of about $30,000,000 is recorded from the Thames district, though but little gold is now obtained there. The 1 James Park, Geology of Hauraki gold field, Trans., N. Z. Inst. Min. Eng., vol. 1, 1897, p. 3. P. G. Morgan, Geology, etc., of Waihi, Trans., Austr. Inst. Min. Eng., vol. 8, 1902, p. 166. J. M. Bell and C. Fraser, The great Waihi gold mine, Bull 15, New Zealand Geol. Survey. A. M. Finlayson, Econ. Geol, vol. 4, 1909, pp. 632-645 (with literature). Arthur Jarman, The geology of the Waihi-Grand Junction mine, Trans., Inst. Min. and Met. (London), vol. 25, 1916, pp. 340, with discussion. 508 MINERAL DEPOSITS veins are contained in broad belts of soft, propylitic andesite (see p. 480) and dip 40 or more. Great masses of low-grade quartz occur, but the gold is derived mainly from rich pockets occurring down to a depth of 400 to 600 feet below the surface. One of these pockets in the Caledonia mine, about 1871, yielded 9 tons of gold in 15 months. The veins have been followed from a height of 1,500 feet above the sea to 640 feet below it, but owing to intervening faults the real vertical extent is only 1,200 feet. Park states that the veins do not continue into the underlying Jurassic shale and that they are thus limited to the thickness of the lava flows in which they occur. The rich shoots occur mainly where the veins are intersected by small stringers or "leaders." Opinions differ widely as to whether this concentration in pockets is due to descending waters or not. In all probability, however, it was one of the latest phases of the primary mineralization. The principal ore mineral is gold alloyed with 30 to 40 per cent, silver, but some pyrite, chalcopyrite, zinc blende, galena, stibnite, and pyrargyrite also occur. Arsenopyrite and native arsenic, the latter secondary, occur at Coromandel. The gangue miner- als, besides quartz, are dolomite and, occasionally, rhodonite. The Karangahake deposits, 40 miles south of the Thames district, are also in propylitized andesite and dacite but differ somewhat from the type described and consist in brief of calcite and quartz with more or less sulphides. The best known de- posits are at Waihi. The Waihi lodes are conspicuous and were discovered in 1878; in part the croppings are covered by rhyolite and the development of the deposit therefore falls between the two eruptions. The ore proved difficult to amalgamate and the mines achieved importance only after the introduction of the cyanide process. To the end of 1917 the total production amounted to about $57,000,000. In 1917 the ore averaged $8 in gold and 1 ounce of silver per ton. The country rock is a green propylitic dacite with some pyrite, calcite, and seams of quartz and adularia. This rock often adjoins the veins without further alteration, but transitions to the quartz filling by silicifi- cation are said to exist. The vein system is complex, and sixteen steeply dipping and interconnecting veins are known. Of most importance is the Martha lode (Fig. 158), a wide fissure vein with brecciated walls; the quartz is formed largely by filling, in part by silicification. On the 500-foot level the lode is in some places 80 feet wide; for DEPOSITS FORMED NEAR THE SURFACE 509 half of this width it is barren, but the other half is said to average $15 to $20 per ton. The proportion of gold to silver by weight is 1:3 or 1:4 and this average was maintained from the surface down. The water level stood within 200 or 300 feet of the sur- face. The lode is said to contain ore for a horizontal distance of 1,700 feet. The developments in the deepest levels are said to be disappointing as to the quantity of ore, but the lode itself maintains its strength. A lively controversy has lately developed in regard to near- surface intrusions. Bell and Frazer consider the dacite intrusive in andesite flows, the inference being that ore may only be ex- pected in that rock. Jarman believes there are no intrusives but only a series of flows. FIG. 158. Cross-section of Waihi mine, New Zealand, showing lode system in andesite and dacite (G), covered by post-mineral rhyolite (B). After C. Fraser. A little pyrite was found in the first level in the Martha lode ; on the second level the sulphide ore on the foot-wall was a few feet wide; on the 500-foot level 20 feet of sulphide ore was exposed on the foot-wall, while the remainder of the vein, at this place 40 feet wide, was thoroughly oxidized, with much black man- ganese oxide. This sulphide ore is of nearly the same value as the oxidized ore, containing perhaps a little more gold and a little less silver. The ore consists of quartz and calcite, with 3 per cent, of pyrite, zinc blende, galena, and argentite. The sulphides are often banded and the gold values are mainly in the pyrite; the 510 MINERAL DEPOSITS bullion contains some selenium. Throughout the oxidized zone the calcite is dissolved, leaving a lamellar quartz ore stained by manganese, but this change is produced mainly by descending waters. In other mines of the district there are indications of a pseudomorphic deposition of silica, similar to that of the De Lamar mine, Idaho (p. 513), by a late phase of ascending solutions. In at least some mines in the Karangahake district the ore be- Later Atidesite FIG. 159. Cross-section of the San Rafael lode, El Oro, Mexico, show- ing branch veins. comes poor when the zone of the calcite, unchanged by descend- ing waters, is reached. The depth of the oxidation in the Waihi mine below water level is noteworthy and probably indicates a dry, intervolcanic epoch. El Oro, Mexico. As there are few important gold deposits in Mexico, the occurrence at El Oro, 70 miles northwest of the fed- DEPOSITS FORMED NEAR THE SURFACE 511 eral capital, is of special interest. The district is situated on the volcanic high plateau at an elevation of about 10,000 feet. The barren and unaltered andesites of this plateau overlie the ore- bearing formation which consists of a thick flat dipping series of well stratified black bituminous shale with some sandstone; in places these Jurassic sediments are covered by earlier andesites, which near the vein contain pyrite and chlorite and in other places they are intruded by sills of similar andesitic rocks. These earlier andesites are held to be of Miocene age. The lodes, about ten in number, only outcrop at one or two places and have been opened by cross cut tunnels and shafts. Almost all of the important work has been done since 1904. There are two principal master lodes, the San Rafael and Dos Estrellas striking N.N.W. and dipping steeply S.S.W. The production from the San Rafael alone since 1904 is approximately $40,000,000 from not less than 5,000,000 metric tons of ore. The Dos Estrellas lode for many years yielded about $5,000,000 per annum. The lodes occupy fault fissures, which in their upper parts at least were open and much of the ore has been deposited by filling of open space. In depth and especially in andesite rock much of the quartz is formed by replacement. In 1913, the greatest depth attained was 2,000 feet; 200 feet of which was in barren cap andesite. The filling consists of fine-grained quartz inter- grown with much coarse-grained calcite. There is only a small percentage of pyrite and zinc blende. The gold is never visible and even close panning often fails to bring a color from rich ore. In the upper levels some stopes are from 60 to 100 feet wide but in depth the lode contracts to smaller dimensions of 3 to 15 feet. The "branch veins" are an interesting feature of the large lodes at El Oro. They are steep and persistent stringers caused by the settling of the hanging wall (Fig. 159) and are usually rich containing from $15 to $35 in gold and from 5 to 20 ounces of silver to the ton. The ore from the main lode contains only $5 to $15 in gold with 2 or 3 ounces of silver per ton. The ore shoots in the main lodes are of primary origin and ex- tend horizontally, to a depth of 500 to 700 feet below the capping andesite. The San Rafael has been stoped continuously for one and one-half miles through three properties. These horizontal 512 MINERAL DEPOSITS shoots are probably caused by an upper, impermeable barrier, now eroded, of andesite or clayey rock. The ore of the branch veins is usually well banded by crusti- fication (Fig. 160) and is much richer in sulphides, mainly zinc blende and pyrite, than the main lode. Oxidation of the main lode antedates the capping of younger andesite and is practically complete to a depth of 800 feet. The calcite is dissolved and the quartz remains as a porous, cellular FIG. 160. Photograph of branch vein on 1 ,000-foot level, El Oro Mining and Railway Company, showing pronounced banding by sulphides; vein three feet wide contains 9% ounces gold and 165 ounces silver per ton. Open cavity in middle. mass. The gold being in the quartz there is a considerable ap- parent enrichment of gold in the oxidized zone. Some silver has probably been leached, but no evidence was found of transpor- tation of gold. There is little manganese in the ore. GOLD-QUARTZ VEINS IN RHYOLITE The Tertiary rhyolites in the Cordilleran region often contain gold-bearing veins These veins are poor in ore minerals other than gold but usually contain some argentite, pyrargyrite, and DEPOSITS FORMED NEAR THE SURFACE 513 pyrite. The gold is frequently coarse and accompanied by more or less silver. Among the gangue minerals quartz prevails, but in most cases it is associated with much adularia, probably derived from the surrounding potassic rock. Calcite and fiuorite are also common, but barite is rare. The veins are almost char- acteristically pseudomorphic, with bladed and cellular quartz and adularia, which replace calcite and fluorite. Both veins and sheeted zones occur; in the latter there is little gangue and the gold, as in the Jumbo mine at Hart, California, may be embedded in apparently fresh rhyolite. There is no real propylitic alteration of the country rock but often extensive silicification and much finely disseminated pyrite. The silicification is attended by concentration of potassium as adularia or sericite. The decomposed upper zone of the veins FIG. 161. Vertical section of the vein system at De Lamar, Idaho. contains clay seams that may be extremely rich in gold and secondary silver minerals, as at De Lamar, Idaho, and Rawhide, Nevada. At the De Lamar mine 1 a series of parallel, gently dipping veins of the kind described abut against a fissure filled with tough clay (the "iron dike") near which the best ore is found (Fig. 161). Below a vertical depth of about 800 feet the values are low, although the veins persist. Free gold is rarely seen. The pro- portion of gold to silver by weight is 1 : 20. The veins are ordi- narily 1 to 6 feet thick and distinctly filled, though transitions by silicification were also noted. The filling is wholly quartz, pseudomorphic after calcite, and forms a cellular mass of thin plates coveerd by minute crystals (Figs. 147 and 148). The 1 W. Lindgren, Twentieth Ann. Rept., U. S. Geol. Survey, pt. 3, 1900, p. 122. 514 MINERAL DEPOSITS value of the ore ordinarily ranges from $10 to $20 per ton. The De Lamar mine has yielded gold and silver to the value of about $7,000,000, but is now closed. F. C. Schrader has described similar deposits in the Black Mountains 1 in Mohave County, Arizona, and the Jarbidge dis- trict, 2 Nevada. At both places adularia is extremely abundant and often forms over 50 per cent, of the gangue (Fig. 162). At the Gold Road mine, in the Black Mountain district, the vein is wide and long; the replacement of calcite and fluorite by quartz and adularia is very clearly shown here. The ore averages $10 per ton and the mine has yielded gold to the value of several FIG. 162. Thin section showing association of fine-grained quartz (Q), with admixed adularia, argentite (A), and native gold (G), Jarbidge district, Nevada. Magnified 105 diameters. After F. C. Schrad&r. million dollars. At Jarbidge a great number of veins have been found intersecting an older rhyolite. An interesting feature is the injection of rhyolite into some of the veins, which are dis- tinctly earlier than the late Tertiary rhyolite flows and were thus formed during a short epoch between two eruptions. En- tirely similar veins are found at Rawhide, Gold Circle, 3 Round Mountain, 4 and many other places in Nevada. 5 1 Bull. 397, U. S. Geol. Survey, 1909. 2 Idem, 497, 1912. 3 A. F. Rogers, Econ. Geol., vol. 6, 1911, p. 790. W. H. Emmons, Bull. 408, U. S. Geol. Survey, 1910. 4 F. L. Ransome, Bull. 380, U. S. Geol. Survey, 1909, pp. 44-47. 6 S. H. Ball, Bull. 308, U. S. Geol. Survey, 1907, p. 46. DEPOSITS FORMED NEAR THE SURFACE 515 In the Bullfrog district, 1 Nevada, a thick complex of tilted and faulted Miocene rhyolite flows is cut by gold-bearing veins. These veins show various gradations from sheeted zones, in which parallel banded veinlets of alternating crusts of calcite and quartz are separated by thin slabs of rhyolite, through irregular stringer lodes, to lodes made up largely of angular fragments of rhyolite cemented by quartz and calcite. The calcite is in part replaced by cellular quartz, but the process has not been carried to completion. The extremely rich ore of the National 2 vein, in northwestern Nevada (Fig. 163), has yielded about $3,000,000 from a narrow Fia. 163. Dendritic gold (black) in extremely fine-grained quartz of prob- able colloid deposition. White area is coarsely crystalline comb-quartz. National, Nev. Magnified 15 diameters. shoot followed to a depth of 800 feet on the vein. The veins of that district, except for this occurrence, are of the stibnite type. The native gold contains 50 per cent, silver and is more properly called electrum. 1 F. L. Ransoine, W. H. Emmons, and G. H. Garrey, Bull. 407, U. S. Geol. Survey, 1910. 2 W. Lindgren, Bull. 601, U. S. Geol. Survey, 1915. 516 MINERAL DEPOSITS ARGENTITE-GOLD-QUARTZ VEINS Tonopah, Nevada. 1 The Tonopah district, discovered in 1900, is situated in a group of desert hills in western Nevada about 30 miles north of Goldfield. It is now the most important of the western silver- and gold-producing localities. In 1916 the pro- duction amounted to nearly $2,000,000 in gold and 8,700,000 FIG. 164. Vertical cross-section showing Mizpah vein of first period in Mizpah Trachyte (M. T.), cut off by intrusive sheet of West End rhyolite (W. R.). Later normal faulting has dislocated vein and along the fault fractures mineralization of the second period has taken place prolonging the vein for short distance in the rhyolite. After J. E. Spurr. ounces of silver. The ore, which is treated by concentration and cyaniding, yielded $17 a ton. The total output amounts to $92,400,000. In 1917 the production was somewhat smaller. 1 J. E. Spurr, Geology of the Tonopah mining district, Prof. Paper, 42, U. S. Geol. Survey, 1905, also, Min. and Sci. Press, Apr. 22, 1911. J. A. Burgess, Geology of the producing part of the Tonopah district, Econ. Geol, vol. 4, 1909, pp. 681-712. A. Locke, The geology of the Tonopah mining district, Trans. Am. Inst. Min. Eng., vol. 43, 1913, pp. 157-166. V. C. Heikes, in Min. Res., U. S. Geol. Survey, annual publication. J. E. Spurr, Geology and ore deposition at Tonopah, Nevada, Econ. Geol., vol. 10, 1915, pp. 713-769. DEPOSITS FORMED NEAR THE SURFACE 517 According to Spun the veins intersect a complex volcanic series of flows and near-surface intrusions. The oldest rock is a highly altered trachyte flow (Mizpah trachyte) glassy in its lower part. This is the "earlier andesite" of former reports and contains the valuable veins. Andesite and "West end" rhyolite are intruded into this flow, which was later covered by the " later " or "Midway" andesite flow; at a still later epoch there was a series of flows and intrusions of rhyolite. The most important of the intrusives is the "Tonopah" rhyolite which now outcrops north of the district and is also found in the mine workings. The differentiation of flows and intrusions near the surface is most difficult and have led to conflicting views. Burgess con- sidered the complex to consist of a series of flows. A Miocene age is attributed to the volcanic rocks. The principal quartz veins are later than the enclosing trachyte and earlier than the following intrusives. They are also older than the "Midway" andesite. A second set of veins was formed after the rhyolite intrusion and before the "Midway" andesite. A few of these are productive. A third set of quartz veins was formed after the intrusion of the Tonopah rhyolite and contain a small amount of sulphides of lead, zinc and copper. All these veins formed at shallow depths and the different types are held to represent various stages of temperature; part of the second and all of the third set of veins are believed to have been deposited at higher temperatures corresponding to the rhyolite intrusions. The faulting is complex (Fig. 33). Important movements took place at every stage and were caused by the volcanic dis- turbances. The importance of the volcanic succession is clearly seen. If an apparently underlying rock is really intrusive and later than the vein formation no continuation of the bonanza veins can be expected in it (Fig. 164). The conditions which resulted in the consolidation" of large masses of intrusive magma to more or less glassy rocks seem peculiar and perhaps deserve further investigation. The productive veins of the earliest period show few outcrops at the surface; they have an easterly strike and various northerly dips. The veins are of moderate thickness though some stopes are 30 feet or more in width. Propylitic alteration (p. 479) affects the trachyte and the andesite but next to the veins there is much silicification with sericite and adularia. The principal 518 MINERAL DEPOSITS gangue mineral is white quartz of fine but variable grain, with banded structure and "chalcedonic appearance," containing parallel bands of finely divided sulphides. The veins are in part filled, but in part appear to have been formed by replacement of the country rock. The primary ore contains some black par- ticles of finely divided gold alloyed with much silver. Argentite and polybasite are the principal ore minerals, with small amounts of pyrite, chalcopyrite, galena, and zinc blende. Selenium is present, probably as a silver selenide. Among gangue minerals there are, besides quartz, rhodonite, adularia, and various car- bonates. The secondary ores developed by oxidation and sul- phide enrichment are described in Chapter 31. Hiibnerite (tung- state of manganese), and scheelite (tungstate of calcium), are believed to belong to the second period of vein formation. The relations of the metals in exceptionally rich concentrates are as follows: Ag 25.92 percent. Sb 0.92 per cent. Au 0.82 per cent. Fe 9. 81 per cent. Pb 6. 21 per cent. MgO 1.49 per cent. Zn 5 . 84 per cent. CaO 3 . 70 per cent. Cu 1 . 32 per cent. S not determined. Se 2. 56 per cent. CO 2 6. 34 per cent. As 0.19 per cent. SiO 2 15. 18 per cent. The Comstock Lode. 1 Among other deposits of this type the Comstock lode deserves special mention. Discovered in 1859, it has yielded, to the end of 1911, a total production of $381,400,000 in silver and gold, of which the gold amounted to $153,000,000. The bonanza period fell in the seventies of the last century and although the production since then has declined greatly, yet during the last few years a systematic unwatering of the deep levels has resulted in a noteworthy rise in output. In 1916 the lode yielded $483,000 in gold and 286,000 ounces of silver, the ore having an average value of $10.64 per t6n. The Comstock lode, situated near the summit of the Virginia Range, east of the Sierra Nevada, is a fault fissure of great throw (Fig. 165), trace- able two and one-half miles and in places several hundred feet wide, the vein matter of the lode spreading in the hanging wall. Great bonanzas of crushed, quartz, in part exceedingly rich in silver minerals, were found at intervals along the^lode, especially 1 G. F. Becker, Man. 3, U. S. Geol. Survey, 1882. J. A. Reid, Bull. 4, California Univ. Dept. Geology, 1905, pp. 177-199. DEPOSITS FORMED NEAR THE SURFACE 519 in chambers or vertical fissures probably produced by normal faulting of the hanging wall. The greatest vertical depth at- tained below the outcrop is about 3,000 feet. Mining has been greatly hampered by enormous quantities of hot water containing mainly calcium sulphate. None of the great bonanzas were found below a depth of 2,000 feet. The lode intersects igneous rocks of deep-seated type, showing transitions and variously classified as diorite, diabase, and augite andesite. 1 These are covered by andesite flows of distinctly FIG. 165. Vertical cross-section through the Comstock lode, showing chambered outcrop and bonanza in vertical vein in hanging wall. After G. F. Becker, U. S. Geol. Survey. Tertiary age, which are also mineralized. Both classes of rocks have suffered propylitic alteration, and analyses of the clay gouge near the veins show that sericitic alteration has set in along the principal channels which the solutions followed. The ores consist of quartz and some calcite, in places banded with pyrite, galena, chalcopyrite, zinc blende, and finely distributed rich silver minerals. The valuable minerals are mainly native gold, argentite, stephanite, and polybasite. 1 A. Hague and J. P. Iddings, Bull 17, U. S. Geol. Survey, 1885. 520 MINERAL DEPOSITS There are in places two generations of quartz, as shown in Fig. 166, the older quartz containing principally pyrite. Zeolites are reported in the altered country rock but are apparently not common. According to Reid the descending waters, rich in sulphates, contained notable amounts of gold and silver, and small quantities of these metals were also present in the ascending hot sulphate and carbonate w r aters. Opinions still differ as to the relative importance of sulphide enrichment in this place. FIG. 166. Rich ore, Ophir mine, Comstock lode, showing earlier fractured quartz with fine-grained pyrite and some argentite (a) and later vein with three generations of galena and argentite (6) with some pyrite, chalcopyrite, and quartz (q). Drawn from specimen in collection of Massachusetts Institute of Technology, natural scale. ARGENTITE VEINS The argentite veins have numerous representatives in Mexico, as at Pachuca, Real, del Monte, and Guanajuato. In general they intersect andesitic rocks of supposedly Miocene age but also cut adjacent or underlying Mesozoic sediments. The rich and long worked veins of Pachuca 1 have come into renewed prominence by the successful application of the 1 J. Aguilera and E. Ord6fiez, El mineral de Pachuca, Boletin, Inst. geol. de M&dco, Nos. 7, 8, 9, 1879. DEPOSITS FORMED NEAR THE SURFACE 521 cyanide process to their ores. A complicated vein system inter- sects andesite flows covering Cretaceous sediments. The andesite is extensively propylitized and this altered rock also adjoins the veins, near which, however, a silicification is often superimposed upon the chloritization. In places the andesite is entirely silicified. The veins are filled fissures, crustified or brecciated, with quartz, sometimes amethysthine, as the prin- cipal constituent of the gangue; there are also rhodochrosite, rhodonite, adularia, and calcite, the last named being the most recent. The ore minerals consist of argentite, stephanite, polybasite, pyrite, galena, and zinc blende. The veins have been followed to a depth of 2,000 feet and here contain argentite, pyrite, and zinc blende. The ores average about 18 ounces of silver to the ton. The oxidation is marked by the zones of the "colorados" and the "negros," the first of which contains limonite with silver haloid salts and the second much manganese. The "negros" are said to contain more gold than the deeper ores, which are very low in this metal. There are many other old and famous silver-mining districts in Mexico, the veins of which are similar to those of Pachuca. Among them are Guanajuato, Zacatecas, Sombre rete, Fresnillo, Batopilas, and Parral. In many of these districts, however, the veins are contained in Cretaceous or Jurassic slates and sand- stones, though, without much doubt, the mineralization is genet- ically connected with the surrounding igneous rocks. The latter are in some cases of intrusive origin. GOLD TELLURIDE VEINS Cripple Creek. 1 The veins of Cripple Creek, situated in an otherwise barren part of Colorado, have since 1891 annually yielded a large amount of gold, which in 1900 reached a maxi- mum of $18,000,000. In 1916 the production of gold was valued at $12,120,000, but the silver recovered amounted only to $53,- 000. The total output of the district to the end of 1916 is $285,- 1 W. Cross and R. A. F. Penrose, Sixteenth Ann. Report, U. S. Geol. Survey, pt. 2, 1896, pp. 1-209. W. Lindgren and F. L. Ransome, Prof. Paper 54, U. S. Geol. Survey, 1906. Horace B. Patton, The Cresson Bonanza at Cripple Creek, Min. and Sci. Press, Sept. 15, 1917. 522 MINERAL DEPOSITS 5 kl 1 > I ;< OJ CL 1 1 a> "3 j) 'o j i (> 1 40 ; d> ^ / I '3 " i o ^ ICQ ?' ^ ' c ' , : O g ! I o 11 * I s ^ \ 1 Is ll \ i!~ "o g i '*! > a "^ ', 5; '"5 i3 |> j j. o' 3 ^ 'c CO i I 1' I ^ o- s v| J 3 c^ *d' [ -/ _0 r> S 1 i J l O 5' j ; G9 m 3 a ll \ fi fe ro c X5~ 3 1