IMAGE EVALUATION TLiT TARGET (MT-3) ^4^ 1.0 1.1 11.25 hi 12.8 ^ m 2.5 2.2 I 6" U IIIIII.6 /: o^ ^ y op^^ M Photographic Sciences Corporation 23 WIST MAIN STRUT WMSTER,N.Y. 14SS0 (716)873-4503 '^^ ^ CIHM/ICMH Microfiche Series. CIHIVI/ICMH Collection de microfiches. Canadian Institute for Historical IVIicroreproductions / Institut Canadian de microreproductions historiques Technical and Bibliographic Notes/Notes techniques et bibliographiques The Institute has attempted to obtain the best original copy available for filming. Features of this copy which may be bibliographicaily unique, which may alter any of the imeges in the reproduction, or which may significantly change the usual method of filming, are checicdd below. D D D □ D Coloured covers/ Couvarture de couleur I I Covers damaged/ Couverture endommagie Covers restored and/or laminated/ Couver'iure restaurde et/ou pelliculte □ Cover title missing/ Le titre de couverture manque I I Coloured maps/ Cartes giographiques en couleur Coloured ink (i.e. other than blue or blacic)/ Encre de couleur (i.e. autre que bieue ou noire) I I Coloured plates and/or illustrations/ D Planches et/ou illustrations en couleur Bound with other material/ RelM avec d'autres documents Tight binding may cause shadows or distortion along interior margin/ La re liure serrie peut causer de I'ombre ou de la distortion le long de la marge intArieure Bianlc leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajouttes lors d'une restauration apparaissent dans le texte, mais, iorsque cela Atait possible, ces pages n'ont pas it* filmtes. Additional comments:/ Commentaires supplAmentaires: Thee tothi L'Institut a microfilm^ le meiileur exemplaire qu'il iui a 6t6 possible de se procurer. Les details de cet exemplaire qui sont peut-Atre uniques du point de vue bibliographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la methods normale de fllmage sont indiquAs ci-dessous. n D D D D D D Coloured pages/ Pages de couleur Pages damaged/ Pages endommagtes Pages restored and/or laminated/ Pages restaurAes et/ou peliiculies Pages discoloured, stained or foxed/ Pages dteolortes, tachetAes ou piqu6es Pages detached/ Pages ditachtes Showthrough/ Transparence Quality of print varies/ Quality in^gaie de I'lmpression includes supplementary material/ Comprend du materiel suppKmentaire Only edition available/ Seule Mition disponible Pages wholly or partially obscured by errata slips, tissues, etc., have been refilmed to ensure the best possible image/ Les pages totalement ou partieiiement obscurcles par un feuillet d'errata, une pelure, etc., ont it* film^es A nouveau de fapon d obtenir la meiileure image possible. The! possi of th( filmir Origli begir the ii sion, othei first sion, or lilt The I shall TINU whic Mapi diffei entin begir right requi meth This item is filmed at the reduction ratio checked below/ Ce document est film* au taux de rMuction indiqud ci-dessous. 10X 14X 18X 22X 26X 30X X 12X 16X aox 24X 28X 32X The copy filmed here hes been reproduced thenks to the generosity of: Library, Geological Survey of Canada L'exempisire fiim6 fut reproduit grAce d la ginArositA de: Bibliothique, Commission Gtologique du Canada The Images appearing here are the best quality possible considering the condition and legibility of the original copy and In Iteeping with the filming contract specifications. Les images suivuntes ont 6t6 reproduites avec le plus grand soin, compte tenu de la condition et de la netteti de i'exemplaire fiim6, et en conformity avec les conditions du contrat de fiimage. Original copies in printed paper covers are filmed beginning with the front cover and ending on the last page with a printed or illustrated impres- sion, or the bacit cover when appropriate. All other original copies are filmed beginning on the first page with a printed or illustrated impres- sion, and ending on the last page with a printed or illustrated impression. Les exempiaires originaux dont ia couverture en papier est imprim6e sont filmto en commenpant par le premier plat et en terminant soit par la dernlAre page qui comporte une empreinte d'impression ou d'lliustration, soit par le second plat, salon le cas. Tous les autres exempiaires originaux sont filmte en commenpant par ia premiere page qui comporte une empreinte d'impression ou d'lliustration et en terminant par la dernid:e page qui comporte une telle empreinte. The last recorded frame on each microfiche shall contain the symbol ^»> (meaning "CON- TINUED"), or the symbol y (meaning "END"), whichever applies. Un des symboles suivants apparattra sur la dernidre image de cheque microfiche, selon le cas: le symbols — ► signifie "A SUiVRE", le symbols V signifie "FIN". Maps, plates, charts, etc.. may be filmed at different reduction ratios. Those too large to be entirely included in one exposure ere filmed beginning in the upper left hand cornea. Heft to right and top to bottom, as many frames as required. The following diegrems illustrate the method: Les cartes, planches, tableaux, etc.. peuvent dtre fiimte A des taux de reduction diffirents. Lorsque le document est trop grand pour dtre reproduit en un seul diet 6, 11 est filmd A partir de I'angle sup6rieur gauche, de gauche d droite. et de haut en bas, en prenant le nombre d'images n6cessaire. Les diagrammes suivants illustrent ia mdthode. 1 2 3 1 2 3 4 5 6 BY THE SAME AUTHOR. Chemical and Geological Essays. Second llevised Edition, witli a New Preface; pp. xlvi. and 489; 8vo. «» XN" PRE1F-A.RA.TION" : MINERALOGY ACCORDIKO TO A NATURAL SYSTEM. Thb outlines of this system — the result of more than thirty years of study — are given in the author's Mineral Physiology and Physiography, pp. 279-401, 687, 688, where a tabular view of Classes and Orders appears on page 382. The principal points in the new treatise are: (1) Arrangement of all native and artificial species of the mineral kingdom in classes and orders, upon a Chemical basis; (2) Their complex chemical structure and high combining numbers, or so-called molecular weights; (3) Application to them of the principles of polyraerlsm and of homologous or progressive series; (4) Direct relations of the constitution of solids, not only to their specific gravity, but to their hardness and their greater or less chemical indiflference, which latter are shown to be connected with the variations in the volume of the chemical unit, or so-called atomic volume ; (5) Recognition of the wide distinction between crystalloid and colloid or amorphous species ; (6) Division, on the above grounds, of the orders and sub-orders into tribes, which generally correspond with the orders of the Natural-History system; (7) Sub-division of tribes into genera and species, and designation of these by a binomial Latin nomenclature ; (8) Systematic descriptions of native, and of some related artificial species ; (9) Genetic history of native species, and their artificial production; (10) Their relations to the great groups of rocks in the earth's crust. MINERAL Physiology and Physiography. A SECOND SEBIES OF CHEMICAL AND GEOLOGICAL ESSAYS WITH A GENERAL INTRODUCTION. 'BY THOMAS STERUY HUNT, M.A., LL.D. (Cantah.) FtUow of the Royal Society of London ; Member of the National Academy of Sclencea of the United States, the Imperial Leopoldo-Carolinian Academy, the American Philosophical Society, the American Academy of Sc!onces, the Royal Society of Canada, the Geological Societies of France, Belgium, and Ireland; OfBcer of the Orders of the Legion of Honor, 88. Mauritius and Lazarus, etc., etc., etc. , BOSTON: SAMUEL E. CASSINO. 1886. Copyright, 1886, By Samuel E. Cassino. Electrottpkd by G. J. Peters & Son, Boston. TO €ffz Mmot^ or BENJAMIN SILLIMAN, WHO DIED 1,35. \-2^ MINERAL PHYSIOLOGY AND PHYSIOGRAPHY. A SECOND SERIES OP CHEMICAL AND GEOLOGICAL ESSAYS. PREFACE. The second title assigned to this volume, — namely, Chemical and Geological Essays, — fails to indicate its character and scope, by reason of the indefiniteness of the word Gcv 'ogy, which is now commonly used to desig- nate both the Natural Philosophy and the Natural History of our earth, except so far as modern geography and meteorology, and the existing flora and fauna, are con- cerned ; descriptive mineralogy and lithology being insep- arable from the study of the earth's crust. In this popular sense, geology is made to include the whole history of organic life in past ages, — a field which rightfully belongs to botany and zoology. The fossil remains of extinct organic forms, valuable as they may be in the diagnosis of stratified sedimentary strata, have, however, no geognostic significance save in their chemical and lithological rela- tions ; and paleontology should, therefore, be distinguished alike from geogeny and geognosy. The proper application of these two terras is defined farther on, in an essay on The Order of the Natural Sci- ences. Therein will be seen the subordination of geogeny to dynamics and chemistry, and of geognosy to descriptive and systematic mineralogy, which are included under the respective heads of Mineral Physiology and Mineral Phy- siography, suggesting, as the more definite title of the volume. Mineral Physiology and Physiography. The essays of which it is made up have been written in accordance with a predetermined plan, which is now TRErACE. accomplished. The first and second are intended to serve as a General Introduction, and to show the relations of the natural sciences to each other and to that complex study which we call geology. In writing the six succeed- ing essays it w.is the author's design to bring together, in a concise form, the facts and the reasonings from which are deduced what he regards as the Principia of geugeny, geognosy, and mineralogy. The chemistry of the atmosphere, and the relations of the earth's aerial envelope alike to outer space and to the gases condensed and the waters precipitated on the sur- face of the globe, as set forth in the third and fourth essays, constitute a necessary preliminary to the study of rock-masses. These, in Essays V., VI., VII., are consid- ered from three different points of view; the genesis and the geognostic relations of the various crystalline rockss, and finally the decay of these, which has deter- mined their present surface-outlines, and has given rise to the materials of the uncrystalline sedimentary strata. In the fifth essay an attempt has been made to show the defects of each of the many contradictory hypotheses hitherto proposed to explain the origin of the crystalline rocks, and to set forth a new one, according to which they have been derived — for the most part indirectly and by aqueous solution — from a single primary plutonic mass, which itself, however, modified both by the action of water, and by partial separations through crystalliza- tion and eliquation, has been the direct source of many exotic rocks. All of these points are more fully discussed in Essay VI. The new hypothesis, as set forth in Essays V. and VI., is the result of nearly thirty years of studies having for their object to reconstruct the theory of the earth on the basis of a solid nucleus, to reconcile the existence of a solid interior with the flexibility of the crust, to find an ' it ' ' PREFACE. Vll adequate explanation of the universally inclined and pli- cated condition of the older crystalline strata, and at the same time to discover the laws which have governed the formation and the changing chemical composition of the crystalline rocks through successive geologic ages. The mineral species which make up the earth's crust next demand attention. A system of classification which should consider their physical ciuiracters, in connection with the chemical composition and tiie mode of formation of mineral sjjecies, has hitherto been wanting. The possibility of such a system, and the principles upon which it might be founded, were pointed out by the author in a series of papers more than thirty years since. He has now, in the eighth essay of the present volume, at- tempted to apply these principles to the study of the natural silicates, which are the most important elements of the crystalline rocks, and to give for these species what he believes to be a natural classification, — followed by an outline of the system as applied to all other native mineral species. The origin of mineral species, their succession, their associations, and the modes of their occurrence alike in massive and in stratified rocks, in veinstones, and in the chemist's laboratory, — in other words, the physiological history of mineral species and their various aggregates, considered both dynamically and chemically, as set forth in Essays V. to VIII., must form the basis of a rational mineralogy and lithology. In this connection are dis- cussed some fundamental principles long maintained by the author, and believed by him to form the basis of " a correct mineralogical system," and, moreover, to " enlarge and simplify the plan of chemical science." That, contrary to the teachings of the Huttonian or metamorphic school in geology, there is an order in the succession of the rocks from the ante-gneissic granite, and VIU PREFACE. that mineralogical constitution and litliological characters, when rightly interpreted, are a sure guide to the relative ages of the various groups of stratified crystalline rocks, was a conclusion early forced upon the author by his studies of these alike in North America and in Europe ; and has led him to propose stratigraphical divisions and a nonienclature which are to-day more or less generally recognized on both sides of the Atlantic. These studies, from 1847 to 1878, were presented in a volume on Azoic Rocks,* published in the latter year, and, with additions up to 1885, are now briefly resumed in the ninth essay. intimately connected with this subject, and at the same time bearing directly upon the different hypotheses touch- ing the genesis of crystalline rocks, is the history of the serpentines, which have been alternately regarded as igneous and as aqueous, as exotic and as indigenous masses, and in either case were supposed to hp.ve been the subject of various metasomatic changes. In the tenth essay, the origin and the geogiiostic relations of these rocks, as found alike among eozoic and paleozoic strata, are considered, and in this connection the history of many crystalline eozoic groups on both continents, but espe- cially in central and southern Europe, has been reviewed, — thus continuing the subject begun in the preceding essay. In concluding, in the eleventh and final essay, the review of the geognostical uistory of the crystalline rocks, continued from Essays IX. and X., the question of the BO- ailed Taconic rocks has been discussed at some length, anii for two reasons. First, because the Lower Taconic series, which has been designated Taconian, appears, as is • Azoic Rocks, etc., by T. Sterry Hunt : Part I. Historical Introduc- tion, 1878, 8vo, pp. xxi. and 253, being Report E of the Second Geologi- cal Survey of Pennsylvania, Harrisburg, Penn. ; Part II., which would have been a special study of these rocks in Pennsylvania, has never ap- peared, but many details thereon are given in Essay XI. of this volume. PREFACE. here shown, to be widely spread over both continents, and to mark the latest known period in the genesis of crystal- line stratified rocks ; and, secondly, because this Taconian series has been by some geologists supposed to represent one of the stages in an imagined process of regional metamorphism by which one and the same group of un- crystalline paleozoic sediments has been made to assume Buccessi\'ely, in contiguous areas, the characters of the various crystalline series from the Taconian down to the Laurentian, both included. To expose the fallacies of this ancient error, and to clear up many of the obscurities which it has thrown alike over the history of theso groups of crystalline rocks and the succeeding Cambrian and Ordovician strata, it was found necessary to examine in some detail the record of stratigraphical research in the pre-Silurian areas of North America, and in so doing to render justice to the work of Amos Eaton, who, more than fifty years since, laid on a sound basis the foundations of American geology. In a volume of selected papers, published by the author in 1874, with the title of Chemical and Geological Essays, in which were discussed the geognostic relations of the Appalachians, of the Alps, and of the Cambrian and Silurian rocks of North America and Europe as known up to that time, the outlines of the present stratigraphi- cal scheme for the eozoic and the lower paleozoic rocks were already, for the greater part, defined ; but the true relations of the Taconian were not then understood. In a Preface to a second edition of that volume, in 1878,* the Taconic question was, however, reconsidered (pp. xix-xxvi), and the a ithor's present conclusions are there briefly set forth. The volume just named contains, moreover, essays on * Chemical and Geological Essays, by Thomas Sterry Hunt, 2d ed., 1&7S 8 vo., pp. xlvi, aud 489. S. £. Cassiuo, Salem [now of Boston], Mass. PREFACE. the origin of limestones, dolomites, and gypsums ; on the chemistry of natural waters, and on petroleum, asphalt, and coal ; on granitic and other veinstones ; and on the theory of ore-deposits. Elsewhere therein, and especially in the first eighty pages, will be found the beginnings of the theoretical views maintained in the present volume, in- cluding disquisitions on the nature and the seat of volcanic action, and on various other points of dynamical geology, considered in connection with the solidity of the earth's interior. Farther on, in pages 426-448 of that volume^ are defined the principles of the mineralogical system which is here developed in the essay on A Natural Sys- tem in Mineralogy. The essays in the present volume, with the exception of the sixth, have already appeared in the Transactions of the Royal Society of Canada, in the Proceedings of the Philosophical Society of Cambridge, England, the Ameri- can Journal of Science, or the London, Edinburgh, and Dublin Philosophical Magazine, A brief notice here pre- fixed to each essay gives "he diite and the conditions of its first appearance. Change;^ have occasionally been made in revision, but wherever there is an addition of significance it is placed within brackets. The same ha-s been done in the case of additional notes. The plan of the present volume was discussed not many ruonths since with t' e author's honored master and his friend of forty years, Benjamin Silliman, tu whom the volume would have been inscribed. It is now dedicated to his memory. Boston, Massachusetts, August, 1886. CONTENTS. I.— NATURE IN THOUGHT AND LANGUAGE (pages 1-26). 1. H18TORICA.L. — Meaning of physics and physiology from Aristotle to Newton, 1. Hippocrates; the school of Alexandria, 8. Physician and mediciner, 10. 2. Philosophical. — Physics, physiology, and physician in modern science, 11. Dynamics, chemics, and biotics, 13. Belations of colloids, 18. Life in matter, 20. Physiophilosophy of Oken, 23. Mineral physiology, 25. IL— THE ORDER OF THE NATI'RAL SCIENCES (pages 27-29;. Natural History and Natural Philosophy, or general physiogra- phy and physiology, 27. Dynamic, chemic, and biotic relations, 28. Tabular view of the natural sciences, 29. III. — CHEMICAL AND GEOLOGICAL RELATIONS OF THE ATMOSPHERE (pages 30-50). Action of the atmosphere on the earth's crust, 30. Fixation of carbonic dioxyd, 34. Sources of this gas, 38. Hypothesis of an interstellary gaseous medium, 40. Dynamic relations of such a medium, 41. Its geological relations, 43. Its astronomical rela- tions, 46. Solar physics and cosmic evolution, 48. IV. — CELESTIAL CHEMISTRY FROM THE TIME OF NEWTON (pages 51-67). Newton's Hypothesis touching Light and Color, 51. Stellar chemistry, and dissociation, 52. Stolchiogeny, 55. Newton on inter- stellary ether, 57. His Principia and Optics examined, 58. Later writers on interstellar space, 62. Terrestrial relations of an inter- stellar gaseous medium, 66. v.— THE ORIGIN OF CRYSTALLINE ROCKS (pages 68-189). 1. Historical and Critical. — Early history; Werner and Ilutton, 68. Neptunist r^nd vulcanist schools, 76. Modified neptunism of Ij^s la xu CONTENTS. Beche, T7. Sorope's Theory of the Earth, 81. Huttonian meta* morphism; pseudomorphism and metasomatism, 82. £ndopIu< tonic hypothesis ; two igneous magmas imagined, 85. Exoplutonic or volcanic hypothesis, 88. Hydroplutonic views of Scrope and others, 96. Metasomatic hypothesis, 98. The various hypotheses reviewed, 104. 2. Development of a New Hypothesis. — Chemistry of the prime- val earth; a solid nucleus, 114. Aqueous origin of natural silicates, 119. The crenitic hypothesis formulated, 131. Crenitic rocks and exoplutonic rocks, 132. 3. Illustrations of the Crenitic Hypothesis. — History of zeo- litic and pectolitic minerals, 135. Secretions of basic rocks, 138. Table of zeolites and related species, 141. Table of protoxyd-sili- cates, 145. Action of heated waters on glass, 147. Recent forma- tion of zeolites, 150. Artificial production of zeolites, feldspars, quartz, etc., 155. Magnesian silicates, 158. Micas and tourma- » lines, 160. Finite and related species, 163. On a genetic classifica- tion in mineralogy, 166. Studies of carbonates, 168. Origin of dolomites, 171. Diageneois; studies of crystallization, 173. 4. Conclusions. — The crenitic hypothesis reviewed ; magnesian salts, 176. Exoplutonic action; folding of strata, 178. Bock-decay, its chemical relations, 180. Genesis of silicates and oxyds, 181. History of the successive groups of crystalline schists, 183. History of exoplutonic rocks, 186. Relation of aqueous to igneous agen- cies, 188. ■ VI. — THE GENETIC HISTORY OP CRYSTALLINE ROCKS. (pages 190-245). Crystalline Rocks Defined, 191. Crystalline silicates in paleozoic rocks, 193. Studies of glauconite, 196. The crenitic hypothesis restated, 199. Origin of Stratiform Structure, 200. The exoplutonists on the eruptive origin of crystalline schists, 201. Non-plutonic intrusion of rock-masses, 204. The endoplutonic hypothesis of the origin of crys- talline schists, 205. Hypothesis of a Liquid Interior of tiie Earth, 207. Eliqua- tion in crystallizing magmas, 208. Illustrated by stratiform dolerites, 210. Studies of chrysolitic plutonic rocks, 211. Secular variation in the composition of plutonic rocks, 214. Its relations to the crenitic]hypothesis, 216. Crystallization from artificial fused mag- mas, 219. Intervention of water in eruptive rocks; igneo-aqueous fusion, 220 (and : 15). Concretionary Banded Granites, 222. Stratiform, calcareous apatite-bearing veins, 224. Their included pyroxenic and feldspathic masses, 226. Plutonic hypothesis of the origin of such veins, 228 (and 236). Their farther geognostic history, and their mineralogy, 229. Studies of such veins in Canada, 232. Their orighoi not CONTENTS. xiii }nian meta- £ndoplu- Exoplutonic Scrope and ; hypotheses f the prime- iral silicates, ic rocks and tory of reo- ! rocks, 138. )rotoxyd-sili- ecent f orma- is, feldspars, Eind tourma- tic classifica- ;. Origin of 173. [nesian salts, )ck-decay, its oxyds, 181. 183. History jneous ageu- ROCKS. in paleozoic c hypothesis i?j nists on the ! intrusion of rigin of crys- ■M !07. Ellqua- *^ j stratiform 11. Secular i relations to V I fused mag- neo-aqueous , calcareous s i feldspathic h veins, 228 mineralogy. origin not plutonic, but aqueous, 236. Calcareous veins and beds on Lake Champlain, 237. Crenitic origin of such veinstones, 238. General corrugation of older crystalline roclcs explained, 241. Conclu- sions, 244. VII. -THE DECAY OP CRYSTALLINE ROCKS (pages 240-278). Historical Sketch, 246. Views of the writer on rock-decay and boulders, 251. Tlie question chemically considered, 253. Kouiuled masses in ancient rocks, 254. Rock-decay in Appalachian crystalline schists, 256. Its relation to pyritous deposits, 257. Limonitcs from pyrite and from siderite, 261. The chemistry of iron-carbonate, :i()6. The decay of serpentines, 208. Antiquity of rock-decay, 209, Foriaa- tion of boulders in situ, 272. Decay in tertiary gravels of California, 272. Rock-decay as related to glacial and erosive action, 274. Studies in Asia, Scandinavia, and Corsica, 275. Conclusions, 277. VIIL — A NATURAL SYSTEM IN MINERALOGY (pages 279-401). 1. Historical Introduction. — Werner, Mohs, and Jameson, 279. The natural-history method, 280. The chemical method of Berze- lius and his school, 282. Defects of both methods, 283. 2.. Attempt at a Natural System. — Its first suggestion : The Object and Method of Mineralogy, 285. Equivalent volumes, 288. The law of progressive series, 289. Polymerism; higli molecular weights and homologies, 290. An atomic notation , 292. The con- . ception of crystalline mixtures, 294. The doctrine of polysilicates and polycarbonates restated, 298. Isomerism in silicates; physical and chemical differences, 299. Atomic or quantivalent formulas, 302. Unit-weights and unit-volumes, 303. Principles of a natural system resumed, 304. 3. A Classification of Silicates. — Three sub-orders of silicates, 305. Chemical relations of alumina and silica, 306. Genetic history of Protosilicates, Protopersilicates, and Persilicates, 307. Question of quantivalent ratios, 311. External characters of spe- cies, 313. Natural-history orders, Mica, Gem, Spar, Zeolite; col- loids, 314. Division of sub-orders into tribes, 315. Significance of physical and of chemical characters, 318. Atomic symbols and weights, 320. General view of tribes and species, 321. Pectolitoids, with table. 322. Magnesian species, 324. Protospathoids, with table, 327. Protadamantoids, with table, 328. Protophylloids, with, table, 331. Ophitoids, with ta'le, 332. Zeolitolds, with table, • 334. Protospathoids, with table ^ 336. History of feldspars, 338. . Scapolites, 340. Sodalite, cancriniio, etc., 343. Protoperadaman- toids, with table, 345. History of tourmalines, with table, 3.50. • Protoperphyllolds, with table, 353. Micas and chlorltes; venerite, , 355. Pinitoids, with table, 360. Production of hydrous colloids by epigenesis; pseudomorphism, 362. Perzeolitoids and perspathoids; XIV CONTENTS. bismtithlc species, 365. Peradamantoids, with table, 365. Zirconic species, 366. Perpliylloids, with table, 367. Argilloids, with table, 369. Crystalline and colloid tribes, 373. 4. Classification of Othek Species. — The order of Oxydates, and its tribes; unit-volumes, 376. The order of Metallates, its sub- orders and tribes; unit-volumes, 378. The orders of Haloidates and Pyricaustates, their sub-orders and tribes, 380. Notes on the sys- tems of Weisbach and Breithaupt, 381. Tabular view of classes and orders in mineralogy, 383. The question of molecular weights and volumes, 383. Polycarbonates, polysilicates, and polytungstates, etc., 385. Complex inorganic acids of Gibbs, 387. An advance in mineral chemistry, 389. The conception of chemical units and unit-volumes, 390. Significance of specific gravities, 394. Relations of chemistry to physics, 396. Scope of mineralogy, 398. Synoptical tables of the three sub-orders of silicates, 399. IX. — HISTORY OF PRE-CAMBRIAN ROCKS (pages 402-425). 1. Pre-Cambrian Rocks in North America. —The names Eozoic and Archaean, 402. Early studies in New York and Canada; Laurentlan and Huronian distinguished, 404. Various views as to the crystal- line schists of the Atlantic belt, 405. The petrosilex group, 408. Relations of the Huronian or pietre-verdi series, 411. The Montal- ban or younger gneissic series, 411. Ottawa gneiss and Grenville series; Lower Laurentlan, 412. Upper Laurentlan, liabradorian or Norian, 413. Lower Taconic or Taconian, in part confounded with Huronian, 411. The Keweenian series, 415. 2. Pre-Cambrian Rocks in Europe. — Crystalline schists in Wales, 416. Pebidian and Diraetian groups, 417. Arvonian or petrosilex group, 418. Petrosilex-conglomerates of Wales and Massachusetts, 420. Various crystalline series in Great Britain, Ireland, and the Ardennes, 421. Pebidian referred to Huronian, 423. Upper Pe- bidian, Grampian or Montalban, 423. Notice of a recently pro- posed scheme of classification, 424. X.— THE GEOLOGICAL HISTORY OP SERPENTINES, WITH STUDIES OF PRE-CAMBRIAN ROCKS (pages 426-516). 1. Historical Introduction. — Early opinions of geologists, 427. Ser- pentines as of aqueous or of igneous origin, 429. Chemistry of silicates and carbonates of lime and magnesia, 432. 2. Serpentines in North America. — Laurentlan serpentines; Eo- zoon, 435. Huronian serpentines, 436. Montalban serpentines, 438. Serpentines of Pennsylvania and Manhattan Island, 439. Serpen- tines of Staten Island, 440. Taconian serpentines, 442. Silurian serpentines of Syracuse, 443. Sepiolite and talc, 448. 3. Serpentines in Europe. — Serpentines and. greenstones of Corn- wall, 449. Serpentines and ophiolites of Italy; the name of gabbro, 451. Studies by Italian geologists; their different views, 454. 7 CONTENTS. XV 4. Geology or the Alps and Apennines, — The work of Gastaldl, 458. His divisions of pre-Cainbrian roclcs, 460. Studies in tlie Biellese, 462. Tlie pietre-verdi zone, 464. Classification of Von Hauer; older gneiss; pielre-terdi ; younger gneiss, 465. Tlie lustrous schists, 467. The St. Gothard section, 470. Four groups of Alpine pre-Cambrlan rocks, 472. The succeeding paleozoic strata in the Alps and Apennines, 473. Studies in Corsica, Elba, and Sar- dinia, 474. The marbles of Carrara, 477. Granulites and gneisses of Saxony; conglomerates, 478. The Montalban series defined in 1870, 480. Concordant views of Von Hauer, 481. Pre-Cambrian rocks of Bavaria, 481. 5. Sekpentines of Italy. — Gastaldi on their age and stratigraphical relations, 483. Serpentines in Liguria and Tuscany, 485. Hydro- thermal hypothesis of their origin, 487. The serpentine of Monte- ferrato, 490. Its stratigraphical relations, 404. Studies in Liguria and in Lombardy, 406. 6. Genesis of Sekpentines. — Two metasomatic hypotheses, 497. The metamorphic and the hydroplutonic hypothesis, 490. Conclusions of Dieulefait regarding serpentines, 501. Other magnesian silicates, 503. Chrysolite both of igneous and of aqueous origin, 506. Studies of chrysolitic rocks, 507. 7. Geognostic Relations of Serpentines. — Controversies as to aqueous or igneous origin of British serpentines, 510. Stapff on the serpentines of St. Cothard, 511. The non-plutonic intrusion of rocks, 512. Conclusions, 514. XI.— THE TACONIC QUESTION IN GEOLOGY (pages 517-616). 1. Introduction. —Amos Eaton; hi., threefold division of strata, 518. Primitive Quartz-rock and Lime-rock; Transition Argillite, First Graywacke and Sparry Lime-rock, 519. Second Graywacke, 520. 2. The Geological Survey of New York. — Ebenezer Emmons; The Champlain division, 522. The Taconic system, 523. Mather; he confounds the First and Second Graywackes, 523. Vanuxem; the Hudson-River group, 524. Eaton's farther conclusions, 526. The term Ordovician, 528, Tabular view of Eaton's classification, 529. Relations of various pre-Cambrian rocks, 530. 3. Geological Studies in Pennsylvania. — The central valleys, 631. Classification of H. D. Rogers, 5S4. His Primal and Auroral are Lower Taconic; their mineralogy, 535. Their great thickness, 537. The First Graywacke in Pennsylvania, 539. The Lower Primal or Azoic and the Hypozoic series of Roger"", 544. Petrosilex or Arvonian In Pennsylvania and elsewhere, o46. Lower Taconic in Chester, Lancaster, and York Counties, 549. The South Mountain, and a second Laurentian axis, 649. Iron ores of the Primal and Auroral; the mines at Cornwall, Dillsburg, etc., 550. 4. LovrER Taconic in Various Regions, —The Taconic hills; the Stockbridge limestone, 554. Argillites, G55. Lower Taconic in XVI CONTENTS. Virginia, 556. Its extension from the Schuylkill Into North Caro- lina described by Maclure, 557. Five areas of it in North Carolina, 558. Flexible sandstone or itacolumite, 501. Lower Taconic in South Carolina, Georgia, and Alabama; pyrophyllite, rutile, and diamonds, 503. Lleber on the Itacolumite series of King's Moun- tain, South Carolina, 504. He compares it with the diamond- bearing rocks of Brazil and Ilindostan, 564. Mineralogy of the Lower Taconic, 5(^. Henry Wurtz on these rocks in North Caro- lina, 560. Lower Taconic east of the Appalachian valley, in New Jersey, Rhode Island, Maine, New Brunswick, and Nova Scotia, 570. In Ontario; the Hastings series, 674. Lower Taconic on Lake Superior; the Animlkie series, 578. Distinguished from Huronian; studies by Kominger, 580. The name Taconian proposed, 582. Relations of Taconian to eozoic and paleozoic times, 583. 5. Uppeu Taconic or First Graywacke. — Traced from the Hudson to the lower St. Lawrence, 584. The Upper Taconic or Taconic- slate group, defined by Emmons in 1842, and then included in the Silurian system. 580. Its relations to the Champlain division; Lower and Upper Taconic farther distinguished, 588. Upper Taconic in Pennsylvania, 589. The Green-Pond Mountain belt, 590. Upper Taconic in eastern New York; its apparent inversion by parallel faults, 592. Red Sand-rock of Vermont, 593. 6. Upper Taconic in Canada. — History of the so-called Quebec group, 594. An inverted series at Quebec, 690. Stratigraphical breaks, 598. Studies in the Ottawa basin, 599. Relations of the Ordovician limestones, 600. James Hall on the Hudson-River group. 601. Logan on the Cambrian or First Graywacke in New York, 603. Distribution of Ordovician and Silurian along the Cambrian belt, 604. The First Graywacke or Uppei Taconic with the Sparry Lime-rock, is the Hudson-River group, 607. The copper-bearing sandstones and amygdaloids of Lake Superior, called Quebec group by Logan, 610. Their history, 611. They are named Keweenian, 614. A similar series in Arizona and Texas, 616. 7. North American Paleozoic History. — The Eozoic lands and the Cambrian sea, 616. First or Cambrian Graywacke, 617. The Ordovician sea, 618. The Green Mountains and White Mountains, 620. Keweenian and Cambrian series; Movements of strata, 621. 8. The Taconic History Revieaved. — American Cambrian in differ- ent areas compared, 622. Rocks of Grand Caflon group and of Newfoundland, 624. The Taconic system named, 627. First Graywacke or Taconic slate-group and Sparry Lime-rock, 627. Mather and his liypothesis, 628. Speculations of various observers, 630. First Graywacke or Upper Taconic called Hudson-River group, and, later, Quebec group, 633. Its Cambrian age, 635. Farther studies of the First Graywacke series, 638. J. D. Dana on the Taconic question, 642. The Taconic succession defined by Emmons, 643. The Sparry Lime-rock is Upper Taconic, 645. Five unlike views as to Lower Taconic or Taconian, 648. Distribution CONTENTS. XVU of Taconlan In North Amorica, fl.'0. Confonmllnjr the First and Second Graywackes; a great error in American stratigraphy, 653. 9. The Metamoki'IIIc IIyi'otiie8I9. — Tlie teachings of Mather, 655. Views of II. I), and W. B. Rogers, 657. J. D. Dana and others oh Soutlieastern New York, 66;}. Rise and fall of llie doctrine of nieta- ' morijliism, 668. rre-Cambrian age of the Scottish Highlands, 669. 10. Conclusions. — The Lower Taconic, Taconian, or Itacolumitic series; its mineralogy, 674. Its distribution in North America, 675. The Upper T ;oidc. First or Cambrian Graywacke, IIudscn-River group, or Quebec group, 676. Relations of various crystalline series, 678. Taconian in the West Indies, South America, and elsewhere, 680. APPENDIX. MiNEBALOGicAL CLASSIFICATION ; action of fluorhydric acid, 687. Binomial nomenclature ; mineralogical evolution, 688. POSTSCEIPTUM. In discussing the " Question of Molecular Weights," on pages 383-395 of this book, it is said that while solid species must be regarded as poly- merids, and their molecular weight as some multiple of the unit-weight deduced from chemical analysis, " the molecular weights of these are as yet unknown"; and, moreover, that "the relations alike of this unit- weight and unit-volume to those of the molecule to which it belongs are unknown." From the principles stated on those pages, and farther on pages 284-304, we may, however, readily fix the weights of these poly- merids if we consider that the volume, Instead of being an arbitrary quan- tity, is the unit adopted in the chemistry of gases and vapors; and, more- over, that the law of volumes is not limited to tliese, but is universal, and applies equally to their condensation into liquids and solids, which are different polymerids of their corresponding vapors, — the conversion of gases into liquids and solids, and, convcrP"'y» the vaporization of these, being a chemical process. If we take as the unit the voiume of water- vapor (H^O = 18) at 100° C, we find that 1487 volumes of this are con- densed into one volume of ice at 0° C, with a specific gravity of 0.9167, 80 that the molecular weight — or, strictly speaking, the equivalent weight— of ice is 1487 x 18 = 26,766; wliile water is 1628(H20) = 29,304. This quantity being the equivalent weight of water (which is the species adopted as the unit of specific gravity for liquids and solids), shows, when divided by two, the number of times that the weight of the hydrogen vol- ume, Hj, is contained in one volume of that unit. From it we calculate the equivalent weight and the true chemical formula of any liquid or solid species, when its specific gravity (water = 1.000) and its empirical formula are known. The so-called volume of the chemical unit, atom or molecule is the reciprocal of its coetlicient of condensation. / NATURE m THOUGHT AND LANG o AGE. This Essay was presented and read in abstract to the National Academy of Sciences at WusUlngton, April 18, 1881. Privately printed in June, it was publislied in the London, Edinburgh and Dublin Philosophical Magazine for October, 1881 ([V] xli., 233-263) under the title of The Domain of Physiology, or Nature in Thought and Language, and again in a second edition, separately, by S, K. Cassiuo, Boston, in 1882. I. — HISTORICAL. § 1. The importance of a correct and well-defined ter- minology in science cannot be overestimated, since a want of precision in language leads to vagueness in thought, and often to errors in philosophy. There are. few more striking examples of indefiniteness in language than can be found in the use of the words physic, physiology, and their derivatives. The material universe is designated with etymological correctness as physical, that is to say, natural — a term which belongs alike to the organic and the mineral kingdoms ; but in the use of this and of other words having a similar etymology (Gr. (pi'>atg, Lat. natura) we find in modern language many restrictions, limitations, and ambiguities. It will aid us in our present inquiry if we bear in mind that both the Greek physis and the Latin natura involve the notion of a generation or growth, and that the adjectives physical and natural, in their origin, imply the results of a formative process or evolution. The term physia (wjiich we translate by nature), as employed by Aristotle, denotes that which ia at once self-producing, self-determined, and uniform in its mode of action. § 2. The substantive physic (fuatx^, physica, physique), has been employed by philosophers since the time of Aris- totle to signify the knowledge of all material nature. 2 NATUUE IN THOU(aiT AND LANOUAOE. [L " Physical science," as well defined by Clerk Maxwell at the beginning of his little treatise on Matter and Motion, "is tliat department of knowledge whicii relates to the order of nature, or in other words, to the reguhir succes- sion of events. The name of piiysical science, however, is often applied, in a more or less restricted manner, to tliose branches of science in which the phenomena aie of the sim- plest and most abstract kind, excluding the consideration of the more c()m])lex phenomena such as are observed in livint; beings." § 3. To the student of natural phenomena, Aris- totle gave the names of tpvaix6i and qivalokoyoi. These words were adopted in the same sens 3 by the Romans, who made use of the substantives phyaicus and physioln. gia to designate natural philosophers and natural science. Cicero writes of the physicus or physician Anaxagoras, g,nd employs the word physiology to denote " the science of natural things," ill accordance, as he tells us, with Greek usage.* § 4. The earlier English writers followed the Greek and Latin usage, and employed the substantive physic (or physike) in the same sense as Aristotle. Thus, in the four- teenth century, Gower defines physic as that part of phi- losophy which teaches the knowledge of material things, the nature and the circumstances of man, animals, plants, stones, and everything that has bodily substance.f Des- * Cicero, Varr. lib. I. R. R. cap. 40. " Si sunt somlna in afire, ut ait physicus Anaxagoras"; also De Nat. Deorum, I, 4. "Rationem naturae quam pliysiologiam Graeci appellant." In" the Totius Latlni- tatis Lexicon of Facciolatus and Forcellinus we find the definition : Physiologia, scientia quae de naturis rerum disserit, eadem ac Physica. t Gower, dividing theoretical philosophy into three parts, Theologia, Physica, and Mathematica, tells us : — "Physike is after the seconde, Thi-oiigU which the philosopbre hath fonde, • ' ;. ^. To teche sondrle kuowlechyngea __ , Upon the bodeliches thinges Of man, of beast, of herb, of stone, • . Of flsh, of fowl, of euerlch one That be of bodily substance, The nature and the circumstance." /| CONFESSio Amantis, book Tii. \ t.l NATURE IN THOUGHT AND LAXQUAOE. cartes, in tlie sovonteenth century, employed the word (in French phi/siqiie^ witli the same signification, and it wast Buhsetiuently used by Locke in a still more comprehensive sense. He writes of "the knowledge of things us they are in tlieir own proper beings, their ct)nstitution8, prof)- ertics, and oi)erationa ; whereby I mean not only matter and body, but spirits also, which havq their proper natures, constitutions, and operations, as well as bodies. This, in a little more enlarged sense of the word, I call (pvamfi or natural philosophy." * § 6. We have seen that in Latin the words physic and physiology were used synonymously. That they were thus understood Ijy English writers is apparent from the Universal English Dictionary of Edward Phillips ((!th edition, 170G), where Physiology is defined as " a discourse on natural things ; physics or riatural jihilosopliy ; being either general, that relates to the affections or properties of matter, or else special and particular, which considers matter as formed or distinguished into such and such sjje- cies." Cotgrave, a lexicographer of the seventeenth cen- tury, in his " French and English Dictionary," also defines Physiologie as "a reasoning, disputing, or searching-out of the nature of things," a definition which is cited by Charles Richardson in his English Dictionary, under Physiology. § 6. It was to those who occupied themselves with ab- stract or general physiology (as defined by Phillips) that the Greeks gave the name of physiologists, first applied to the philosophers of the Ionian school, who sought to de- rive all things from one or more material elements, and thus had a physical basis for their system of the universe, as distinguished from the school of Pythagoras, whose system was based on numbers and forms. Of Empedocles, the author of a didactic poem on Nature in which we first find enunciated the doctrine of the four elements, fire, air, earth, and water, Aristotle, in his Poetics, makes the criti- cism that he was mors of a physiologist than a poet. * Human Understanding, b. vii., c. 21. NATURE IN THOUGHT AND LANGUAGE. VL Humboldt repeatedly employs the word physiology and its derivatives in the same general sense. Thus, he writes of "the natural philosophy of the Ionian physiologists" (physiologien), which "was devoted to the fundamen- tal ground of origin, and the metamorphoses of one sole element " ; of the " physiological fancies of the Ionian school," and of the teachings of Anaxagoras of Clazomenae, " in the latter period of development of the Ionian physi- ology." * Of Anaxagoras it may be observed that his views marked a great advance over those of his predeces- sors, and that he merited the encomium pronounced by Aristotle that he was the first philosopher who had written soberly of nature. § 7. We find the word physiology and its derivatives employed in the same general sense by English writers in the seventeenth century. Thus, Cudworth speaks of " the old physiologers before Aristotle," and writes " they who first theologized did physiologize after this maixner, inas- much as they made the Ocean and Tethys to have been the original of generation," f while Henry Moore says, " It will necessarily follow that the Mosaical philosophy, in the physiological part of it, is the same with the Carte- sian." I Coming down to later writers, we find the word physiologist used in a general sense, as equivalent to our modern term naturalist. Thus, Dugald Stewart calls Cuvier " the most eminent and original physiologist of the present age," and Burke writes, "The national menagerie is collected by the first physiologists of the time." § We may note in this connection the two series of abridg- ments of the Philosophical Transactions of the Royal Soci- ety — the first, from its commencement to ITOO, and the second to 1720 — both published with the imprimatur of Newton as president of the Society. In these collections * Cosmos, Otte's translation, Harper's ed., II., 108, and III., 11. t Intellectual System, pp. 120, 171. t Philosophical Cabbala, Appendix, c. 1. § Stewart, Philosophy of the Human Mind, II., c. 4; and Burke, Letter to a Noble Lord. I \ f -Am IE. jM gy and M ; writes ,^^H ogists ' .1^9 idamen- mM of one '^1 3 Ionian ;^B omenae, |^ n physi- M that his S 3redeces- ^| meed by « vho had ^^H jrivatives |B jrriters in |9 :s of " the M they who H iner, iuas- 9 lave been 9 )ore says, ^ lilosophy, 1 he Carte- C- the word j| nt to our m ^art calls ■ rist of the nenagerie " § of abndg- oyal Soci- , and the imatur of M loUections B L] NATURE IN THOUGHT AND LANGUAGE. In., 11. and Burke, the classification of the papers is as follows : (1) " Matlie- matical," including pure and applied mathematics ; (2) "Physiological," embracing all meteorological phenom- ena, tides, terrestrial magnetism, mineralogy, geolog}-, botany, zoology, and the study of the physical wrrld in general. Subjects relating to the human body, liowever, such as anatomy and medicine, were excluded from part 2, and, with chemistry, made a first division of part 3, in the second and last division of which were included philo- logical and miscellaneous papers. § 8. Of the " special and particular physiology," as distinguished by Phillips, we have an example in Glanvil, who, in the seventeenth century, writes of the physiology of comets.* The citation from Burke, identifying physi- ologists with zoologists, may also perhaps be taken as an example of a special use of the word, wliile in later times we have come to speak of Vegetable Physiology, Animal Physiology, Human Physiology, and even of Jiental Phys- iology, a term employed by Dr. Thomas Brown of Edin- burgh,! who speaks of " physiology corporeal or mental." J * "So that we need not be appalled at blazing stars, and a comet is no more ground for astrological presages than a flaming chimney. The unparalleled Descartes hath unravelled their dark physiology, and to wonder solved their motions." Joseph Glanvil, Scepsis Scientifica, . . . an Essay on the Vanity o* Dogmatizing, 1665, c. xx. t The grounds upon which Brown based this extension of the term physiology may be gathered from the following passages: " There is, in short, a science which may be called mental physiology, as there is a science relating to tlie structure and offices of our corporeal frame, to which tiie term physiology is more commonly applied." He farther speaks of the ^'physiology of the mind, considered as a substance capable of the various modifications or states which, as they succeed each other, constitute the phenomena of thought and feeling," and declares that "the mind is as an object of study ... to be compreliended, with every other existing substance, in a syi^tem of general i^hysics," Brown, The Philosopliy of the Human Mind, lectures I., II., and V. t Since the writing of this essay, Prof. Osborne Reynolds, in Nature for June 9, 1881 (vol. xxiv, page 123), has made a happy use of the word in question in Avriting of the locomotive engine of George Stephenson, of which he says, "the physiology of the machine resembled that of the human system"; while he speaks of its inventor as "he who produced the locomotive physiologically perfect." rz^ NATUKE IN THOUGHT AND LANGUAGE. tl. § 9. There is an example of a special application of the words physiology and physic which requires farther con- sideration. We have already cited Cotgrave's first defini- tion of the word Physiologic, to which he adds, as a sec- ondary meaning, "anatomizing physiv., or that part 6f physic which treats of the composition or structure of man's frame." In more recent times, however, the term has come to mean, not the anatomy, composition, or struc- ture of the human frame, but its functions, to which sig- nification physiology is, in popular language, limited, though now by didactic writers extended to include the functions of the lower animals, of plants, and even of the human mind. The word physic, as we have seen, was used by Gower in the general sense cf a knowledge of all • aaterial things, but his contemporary, Chaucer, employed it, in a special and restricted sense, to designate the science of medicine. Thus, he calls his practitioner of the medical art " a doc- tor of physic," and in his tiescription of this personage adds that " gold in physic is a cordial." * Subsequently, and to our own time, we find the terra applied, in Chau- • "With us there was a doctour of phisik, In all the world ne was there non him lyk To speke of phisik and of surgerye, For he was grounded in astronomye. ■ • • • ■ He knew the cruse of every maladye, Were it of hot or cold or moyste or drye, And where engendered and of what h'umoure; He was a very parfight practisour. I I Well knew he the old Esculapius, And Dioccorides, and eke Rufus, Old Ilippocras, Kali and Gallien, Serapion, Rasis and Avicen, Averrois, Damascene and Constantin, Bernard, and Gatisden and Gilbertin. For gold in phisik is a cordial. Therefore he loved gold in special." CuAucEB, Canterbury Tales, Prologue, I] NATURE IN THOUGHT AND LANGUAGE. cer's sense, alike to the art of healing and to its medica- ments. If we search for the origin of this peculiar use of the word physic, we shall find it employed with the same meaning in medieval Latin.* In French also, ac- cording to Littr^, the term physique was in the thirteenth century applied to the science of medicine, the professors of which were tlien called phi/siciens, | a designation which they kept till the time of Rabelais, and, as we know, still retain in English, though the term physicien is at present applied in French only to students of physical science in the restricted sense mentioned in § 2, including what, in didactic phrase, is now called physique in French and physics in English. § 10. It is a curious inquiry how these terms came to have this restricted use in the middle ages, and how the name of physicus or physician, originally applied to the student of material things — and by pre-eminence to An- axagoras of Clazcmenae, who was called "the ph}sician," (6 eeies." In 1854, in an essay entitled "Thoughts on Solution,"* I, however, declared, with regard to Kant's view, that " the conception is mechanical, and therefore fails to give an adequate idea. The definition of Hegel, that the chemical process is an identifieation of the ditt'erent, and a differentiation of the identical, is, however, completely adequate. Chem- ical union involves an identification not only of the vol- umes (interpenetration, mechanically considered), but of the specific characters of the combining bodies, which are lost in those of the new species. . . . We may say that all chemical union is nothing else than solution ; the unit- ing species are, as it were, dissolved in each other, for solution is mutual." The above considerations will serve to show the essen- tial nature of chemism, a process resulting in the genesis of chemical species, which are mineral or inorganic. § 20. The force involved in the chemical process mani- fests itself as radiant energy and electricity, and there is apparently a tendency among modern dynamicists to confound these activities with chemism itself, and thus to lose sight of the essential significance of the chemical pro- cess as already defined. Thus Clifford wrote of molecular motion "which makes itself known as light, or radiant heat, or chemical action,"! while Faraday was wont "to express his conviction that the forces termed chemical affinity and electricity are one and the same." Helmlioltz, from whom I here quote, adds, " I think the facts leave no * Of the two essays above quoted, the first appeared in 18.53, in the American Jonrnal of Science for March, and also in the L., E. and D. Phi- los. Magazine [4] v., trlQ, and was translated into German in the Chemis- ches Centralblatt for 1853, page 849. The second was published in the American Journal of Science for January, 1854, and also in the Chemical Gazette for 1855, page 90. Both will be found in the author's volume of "Chemical and Geological Essays," in which, for the extracts here given, Bee pages 427, 428, and 450. t W. K. Clifford, Essays II., 17. 16 NATUKU IN THOUOIIT AND LANGUAGE. [I. •I'll! tiiii'" 4. doubt that the very mightiest among the chemical forces are of electrical origin, . . . but I do not Huppose that other molecular forces are excluded, working directly from atom to atom." * The activities which appear in dynamic and in chemio phenomena are one in essence, for force is one. The same is true of the activities manifested in organic growth, and even in thought; but the unity and mutual convertibility of different manifestations of force afford no ground for confounding, as some would do, dynamics with chemics, or with vital or mental processes. All of tiieso phenomena are but the evidences of universal animation, or, in other words, of an energy which is inherent in matter, the mani- festations of which, as matter rises to higher stages of development, become more complex, as organic individ- uals are themselves more complex than mineral forms, f § 21. From the process which generates chemical species we pass to that which gives rise to organized individuals, * Helmholtz, The Faraday Lecture, April 5, 1881; abstract prepared by its autlior; Nature, vol. xx'U., p. 539. t [This view of hylozoisiu v .s well set forth l)y rosinini. According to him, in the words of his intPit-rtacr, Davidson, " tlic ultimate particles of matter are animate, each atom having united with it, and forming its unity or atomicity, a sensitive principle. Wlien atoms chemically com- bine, their sensitive principles become one. '. . . The unit of natural existence is neither force nor matter, but sentience, and through this all the material and dynamical phenomena of nature may be explained." From the unifications of these sensitive principles, or elementary souls, which take place in the combinations of matter, higher and higher mani- festations of sentience appear, constituting the various activities dis- played in crystals, in plants, and in animals. From these elementary souls organic souls are built up, and "when these are resolved into the elementary ones through the dissolution of the organized bodies, the existence of the souls does not cease, but is merely transformed." [See "The Philosophical System of Kosmini." by Thomas Davidson (18S2), pp. 284-301.) This volume was unpublished, and these views of Rosmini were unknown to me, at the time of writing the above pages. The em- inent biophysiologist, the late William B. Carpenter, in an essay on " Life," published in 1847, in Todd's " Cyclopedia of Anatomy and Physi- ology," Vol. III., p. 151, contends that organization and biotical func- tions arise from the natural operation of forces inherent in elementary matter.] I.l NATURE I^ THOUGHT ANU LANOUAUE. 17 ill wlrch appear a new class of pluMioinena, distiiiguisliod alike 'nun those of dynainics and llioso of clieiiiisin. These new manifestations, wliich are cuIUmI vital, involve dynam- ical and chemical activities, hnt disi)hiy, in addition to these, still higher ones. Matter, on this more elevated plane, not only bueomes individnalized, but adajjts itself to external conditions, by orgaid/ation, and exhibits in the resulting forms the power of growth by assimilation, and of reproduction. The study of these forms in all their rela- tions is the object of Hiology. Organogeny, or the; process of morphological growth and (levelo[)ment, distinguishes the biological from the nuneralogical individual. The ac- tivities of the crystal are purely dynamic, and its crystal- line individuality must be destroyed before it can become the subject even of chendsm, while the plant and the aiumal exhibit not oidy dynamical and chemical, but organogenic activities, which last are designated as vital phenomena. Tiio study of these constitutes a third division of physics, which may be conveniently designated as lliotics (fro;n ^toTix6i^ pertaining to life), and has to do with organic growth, development, and reproduction, activities which do not api)ear in the mineral kingdom. Mineralogy is the science of inorganic matter, and studies its dynamical 'and chemical relations, while Biol- ogy, which is the science of organic matter, adds to these the study of biotic relations. The dynamic and ehemic activities which in the mineral kingdom give rise to the crystalline individual, are therein in static equilibrium. The organic individual, on the contrary, is j'.inetic, and maintains its equilibrium only by perpetual adjustment with the outer world. § 22. General physic, or the study of nature, presents itself under a twofold aspect, the historical and the philo- sophical ; the former gives rise to physiography, while to the latter the name of physiology more i)roperly belongs. Physiography describes si)0oif.<3 and individual forms, and their external relations, while physiology investigates the 18 NATUEE IN THOUGHT AND LANGUAGE. CI* r II 'II I processes by which these forms are produced, and gives us the logic of nature. The physiology of matter in the abstract is dynamic, that of mineral forms is both dynamic and chemic, while that of organic forms is at once dynamic, cheraic, and biotic. Nature in all its manifestations constitutes a unity, and it is the object of general phsyiology to study the process of ceation in the material world from primal matter up- ward through its various forms until it attains to organi- zation, and at length, in man, to self-consciousness, where the domain of physiology ends and that of psychology begins. § 23. In accordance with th'e views here enunciated, all matter is in a sense living, "all movement is radically vital," * though we, in common language, refuse the desig- nation of vital to those lower forms of material activity which appear in dynamic and chemic phenomena, reserving it for such as are supposed to be peculiar to organized forms, which, to prevent misconception, I have called biotic. When matter, through chemism, attains the con- dition of protoplasm, which may be chemically described as a colloidal albuminoid united with more or less water, it begins to exhibit that form of activity which we term vital, or biotic. " The mobility and 'the spontaneous move- ments of this substance," says Allman,f "result from its proper irritability. From the facts there is but one legiti- mate conclusion, that life is a property of protoplasm." J § 24. Many of the peculiar characters of protoplasmic * Stallo, Philosophy of Nature, p. 66. t Alhnan, Presidential Address before the British Association for the Advancement of Science, in 1879. t The views set forth in this and the three sections preceding may be compared with those concisely expressed by Huxley since the preceding pages were first printed, in his address in August, 1881, before the Inter- national Medical Congress in London. He thereir' concludes that the "contrast between living and inert matter, on which Bicliat lays such stress, does not exist. . . . Living matter differs from other matter in degree, and not in Itind ; the microcosm repeats the macrocosm, and one chain of causation connects the nebulous original of suns and planetary W ! n NATUEE IN THOUGHT AND LANGUAGE. 19 matter appear to be common to chemical species in the colloidal condition. The remarkable properties exhibited by colloids led their discoverer, Graham, twenty years since, to declare, " The colloidal is, in fact, a dynamical [kinetic] state of matter, the crystalloidal being the stati- cal condition. The colloid possesses Energla ; it may be looked upon as the probable primary source of the force appearing in the phenomena of vitality. To the gradual manner in which colloidal changes take" place (for tliey always require time as an element) may the characteristic protraction of chemico-organic changes also be referred." * Following Graham, Herbert Spencer has noted that plia- bility, elasticity, the power of absorbing water with change of bulk, and the phenomenon of osmosis, — the whole of which are well designated by him as showing sensitiveness to external agencies which are mechanical or quasi-me- chanical — are possessed in common by mineral colloids and by organized substances. These phenomena are exam- ples of that " continuous adjustment of internal relations to external relations " which characterizes organic life.f "When the chemist shall have succeeded by his synthesis in producing a colloidal albuminoid having the same chemi- cal constitution as protoplasm, there is, as Barker has well said, reason to expect that it will exhibit all the phenomena of life which appear in the protoplasmic matter common to plants and animals. § 25. Barker has, in this connection, asked the impor- tant question : What are we to understand by organic life, and what is the true meaning of vital, as applied to a function?! If, with him, we answer, following Kiiss, — systems with the protoplasmic foundation of life and organizations." (Nature, Aug. 11, 1881, vol. xxiv., p. 34G. ; * Thomas Graham. Chemical and Physical Researches, p. 554, from Philosophical Transactions for ISOl, p. 183. t Herbert Spencer, Principles of Biology, vol. i., part 1, chapters 1 and 2. t Geo. P. Barker, Address as President of the American Association for the Advancement of Science, Boston, August, 1880. I have in this paragraph closely followed Professor Barker's argiunent. 20 NATUKE IN THOUGHT AND LANGUAGE. [L I • i; iil ! i • "life is all that cannot be explained by dynamics and che»^^- ism," we shall find, restricting our inquiries to the animal economy, that a large part of the phenomGua commonly called vital, — and as such included under the head of ani- mal physiology, — are dynamic or chemic. Tlie law of the conservation of energy applies as rigidly to a living animal as to a thermic engine, and the amount of work done, or of heat evolved, is measured by food consumed in the former as it is by the fuel burned in the latter ; the energy mani- fested in both cases being dependent on the oxydation of carbon and hydrogen. Recent inquiries go far to confirm the view that muscular contraction is electrical, and that electrical manifestation in the muscles is, as in our ordinary batteries, dependent on chemism. The tendency of late investigations is to bring nervous activity into the same category, and the electrical nature of capillarity has been shown by Draper and by Lippmann. The animal circula- tion is a mechanical result of muscular contraction ; the aeration and the coagulation of the blood, and the process of digestion, are chemical, while absorption finds an expla- nation in the phenomena of diffusion and osmosis. When the energy which is in matter is manifested without reference to species, we call it simply dynamics ; when it results in the production of mineral species, we call it chemics, or chemism ; and when it gives rise to organisms, which may be defined as kinetic individuals, we distinguish it as vital, or biotic. In matter, we must recognize with Tyndall " the promise and the potency of all terrestrial life." * * [Address as President of the British Association, Belfast, 1874. Ap- pleton's ed., p. 59. In another version of this address, cited by Stallo, Tyndall declares that he discerns in matter " the promise and the potency of every form and quality of life," respecting which Stallo remarks: " Tyndall's words were little more tlian a new wording of an old thouglit of Francis Bacon, who said, more than two centuries ago: ' And matter, whatever it is, must be held to be so adorned, furnislied, and formed, that all virtue, essence, action, and natural motion may be the natural conse- quence and emanation tliereof ' ('Atque asserenda materia, qualiscunque ea sit, ita ornata et apparata et formata ut omnis virtus, essentia, actus h] NATURE IN THOUGHT AND LANGUAGE. 21 § 26. It follows, from what has been said, that the word physiology, as popularly limited to the fimctioRS of living beings, is made to include many phenomena which are not biotic, but are common ^o the organic and mineral kingdoms, and that we need some further definition to distinguish those which are characteristic of organic life. I therefore venture to designate the study of these by the distinctive name of Biophysiology, while those phenomena which are recognized as simply dynamic, or dynamic and chemic, whetlier manifested in organisms or in mineral species, may be included under the name of Abiophysiology. General physiology, comprehending these two divisions, will thus be restored to its original and proper significa- tion, as an inquiry into the reason of all things in the material universe, and as distinguished from physiography, whose province is the description of universal nature. Scientific precision demands a reform in our terminology, and requires us to extend the name of physiology once more to the processes and the activities of the three king- doms of nature. The inorganic, not less than the organic world, has its physiology. On the other hand, the study of mind and spirit, and the phenomena of consciousness, which Locke and Thomas Brown included under the head of physic and physiology, should be relegated to the domain of psychology. § 27. The kindred term physiography is now correctly employed in a general sense, wil-h a meaning co-extensive atqwe motiis naturalis ejus consecutio et emanatio esse posslt.' Baco, De Princ. atqiie Orlgg., 0pp. ed. Bohn, vol. ii., p. GDI). The same thing has been repeated many times since by the metaphysical evolutionists, in terms substantially like those of Schelllng: ' Matter is the general seed- corn of the universe wherein everything is involved that Is brought forth in subsequent evolution ' ( ' Die Materie ist das allgemeine Samenkorn des Universuras, warin Alles verhiillt ist was in spiiteren Entwickelungen sich entfaltet.' Schelling, Ideen zu einer Philos. der Natur, 2d ed., p. 315) " Stallo, The Concepts and Theories of Modern Physics, pp. 153, 154. Compare with this the view of W. B. Cai-penter cited in a note to § 20, supra, page 16.] KAXCBK ™ THOUGHT A^r> ^ASGUAGE. II. •I -^^ .n-v A great living ^Hh that wMch --'tv Ms'^v'^ "^ -'^- *''^ *'''^ "^^ teaoHer, Professor H»rie5^; has g we ^ ^^ ;, « Physiography ; an In'-^f "»j"'" , ° , describing the rocKs, „ elementary '^^''''^^V^''^;;!"; .Aich make up the in- the waters, and the ''f"^?^^!^' J\,,„eeeds to cons der organic portions f ^^''"tnimaU and their relaUons the development o P ™*;^^°^,^i kingdom, and cone udcs ::rrrnforthrar— irelationsofourplanet ^-:::lrs;iro^«.— -^-4p without which a trne -p^'^^^ZTs," a complete physi- Humboldt to attempt, .. h^J ^i,,eription of the ography, which -''''"''^^^^^'J things in the regions of universe, em.bracmg all "^^^^^j^t ^elsewhere speaks of space and in the earth. ^mnb associated w.th /the idea of vitality • • •/" '" ever-blending natural that of the OKistenee »f«-J Sphere," and, recallmg forces which animate t^e terrest' P .^^^^^^^^ „ the fact that the inorganic crust of t ^^^^ ^^ ^^ same chemical elements that «»'; „ ^^ physical cos- Til and vegetaWe ;r,a„.m ^a^a^^^^^^ | ,, to mography *o»"l^.'^^''ftLse {orce3,-and of the sub- omit a C''"»i'l^™*'"^ °* *ud and liquid combinations in stances that enter into ^^^^^''^^J^^,,,^ _ which, from organic tissues un^ "^^^n t--- ^ "^'T'" \ t. our ignorance of their actu j„,ai tendency of the vague term of vital forces. The n ^^^^ ^^y^^. Lman mind '"-f ™*S^Zgh all their varied series, cal phenomena of the e'''* *^f ^ ^ morphological evolu- r orvei:t";rmt 25 =elf.etermining powersof I.] NATURE IN THOUGHT AND LANGUAGE. 23 ophy of the material universe, or, in other words, a gen- eral physiology. The most complete attempt at thus systematizing nature is that of Lorenz Oken, who divided all philosophy into Pneumatophilosophy and Physiophilos- ophy, corresponding respectively to Spirit and to Nature. Physlophilooophy, as defined by him, is the scie'ice of the conversion of Spirit into Nature, and has for its object to show how, and in accordance »vith what laws, tlie material universe has been formed; to portray the first periods of the world's development from naught; to show how the heavenly bodies and the chemical ele- ments originated ; in what manner, Ijy self-evolution into higher and manifold forms, these generated mineral spe- cies became at length organic, and in man attained to self-consciousness. Physiophilosophy is therefore the generative history of the world, or, in other words, the history of the process of creation. It aims, in the language of Stallo, to describe " the genetic evolution of the material world ; therefore, also, its first origin in naught, and its subsequent develop- ment up to its limit, man, who is a complex of all j^reced- ing forms, includes all particular developments, and is, as it were, the focus where all the various tendencies of Nature converge. ... In man, all eternal activities, all divine ideas are gathered " ; and thus it is that, in the words of the poet, he is enabled " to think again the great thought of the creation " * § 29. The origin of matter itself, Hylogeny, belongs to Pneumatophilosophy. The genetic process in the primal undifferentiated matter, with wliieh Physiophilosophy first concerns itself, is by Oken considered under the two *" Schcin ist, Mutter Natur, deiner Erfindung Praeht Auf die Fluren verstreut ; schoner ein froh Gesicht Das den grossen Gedanken Deiner Schopf ung nocli einmal denkt." Klopstock, Ode, Dcr Zii.rcJicrsee. Compare this with the language of SchelHng, cited by Ilegel : " Uber die Natur philosophiren heisst die Natur schaffen." 24 NATURE IN THOUGHT AND LANGUAGE. n. m ill! HI! ! ' 11 heads of Ontology und Biology. The successive steps in the ontological process are, first, Cosmogony, or the fash- ioning of the heavenly bodies from the previously formed matter ; followed by the genesis therefrom of the chemical elements ; Stoichiogeny. These elements give rise to mineral species, which together make up the earth ; Geo- geny. Biology, wJiich has for its object the study of the organic world, is by Oken div'ded into Organogeny, with its sub-divisions, and Phytosophy and Zoosophy, treating respectively of the development of plants and animals. In the organism we have " a combination of all the activi- ties of the universe in a single individual body." The inorganic and the organic worlds are not only in harmony with each other, but are one in kind. Man, in whom self-consciousness or Spirit manifests itself, represents the whole universe in miniature.* § 30. The physiophilosophy of Oken, of which we have given an outline, is thus identical in its aim and its plan with the earlier attempts of the Greek philosophers to which the name of physiology was given, and the two terms are, in fact, synonymous. The study of nature, as has been shown, divides itself into physiography and phy- siology, and this division applies equally to each one of the three great kingdon.s of nature. Thus, for example, Physiographical Botany studies the relations of plants to each other as members" of the vegetable kingdom, and investigates the^r external forms and relationships, by which we arrive at Systematic and Descriptive Botany, with its classification and terminology. These together * Lorenz Oken, Physiophilosophy ; Introduction, pp. 1-3, of Tulk's translation, published by tlie Ray Society, London, 1847. See also an excellent analysis of the system by J, B. Stallo in his Philosophy of Nature, Boston, 184S, pp. 221-330, from which we have quoted above. Errors in detail, and defects, and obscurities, are to be foimd in the system of Oken, which even novices in science can to-ivine Spirit, which '" Lives through all life, extends through all extent. Spreads undivided, operates unspent.' " " The law of birth, growth and decay, of endless change and perpetual renewal, is everywhere seeA working throughout the Cosmos — in nebula, in world, and in sun, as in rock, in herb, and in man, all of which are but passing phases in the endless circulation of the universe, — in that perpetual new birth which we call Nature. This, it will be said, is the poet's view, but it is at the same time the one which seems forced upon us as the highest generalization of modern science." n. THE ORDER OF THE NATURAL SCIENCES. The system of classlflcation set forth in the preceding essay on Nature in Thought and Language was embodied in tlio following note presented to the Ameri< can Association for the Advancement of Science, at Minneapolis, August, 1883. This was published at the time in Science, in the Proceedings of the Association, and also the same year as an appendix to an address, by the author, on The Relation of the Natural Sciences, in the Transactions of the Royal Society nf Canada. Volume I., section iii., pages 7-8. § 1. The study of material nature constitutes wiiat the older scholars correctly and compreliensively termed physics (the words physical and natural being synony- mous), and presents itself in a twofold aspect, first as descriptive, and second as philosophical, — a distinction embodied in the terms Natural History and Natural Philosophy, or more concisely, in the words Phj^siography and Physiology. The latter word has, from the time of Aristotle, been employed in this general sense to desig- nate the philosophical study of nature, and will so be used in the present classification. § 2. The world of nature is divided into the inorganic or mineralogical, and the organic or biological kingdoms, the division of the latter into vegetable and animal being a subordinate one. The natural history or physiography of the inorganic kingdom takes cognizance of the sensible characters of chemical species, and gives us descriptive and systematic mineralog}'', which have hitherto been restricted to native species, but in their wider sense include all artificial species as vrell. The study of native mineral species, their aggregations, and their arrangement as constituents of our planet, is the object of geognosy and of geography. The physiography of other worlds gives rise to descriptive astronomy. 27 28 THE ORDER OF THE NATURAL SCIENCES. [II. H § 8. The natural philosophy of the inorganic kingdom, or mineral physiology, is concernetl, in the lirst place, with what is generally called dynamics or physics, includ- ing the phenomena of ordinary motion, sound, tempera- ture, radiant energy, electricity and magnetism. Dynam- ics, in the abstract, regards matter in general, without relation to species; chemism generates therefrom mine- ralogical or so-called chemical species, which, theoreti- cally, may be supposed to be formed from a single ele- mental substance, or materia prima, by the chemical process. Dynamics and chemistry build up our inorganic world, giving rise to geogeny and, as applied to other worlds, to theoretical astronomy. § 4. Proceeding now to the organic kingdom, its physi- ographical study leads us first to organography, and then to descriptive and systematic botany and zoology, two great subdivisions of natural history. Coming next to consider the physiological aspect of organic nature, we note, besides the dynamical and chemical activities mani- fested in the mineral, other and higher ones, wliich char- acterize the organic kingdom. On this higher plane of existence are found portions of matter which have become individualized, exhibit irritability, the power of growth by assimilation, and of reproduction, and, moreover, estab- lish relations with the external world by the development of organs, all of which characters are foreign to the mineral kingdom. These new activities are often desig- nated as vital, but since this word is generally "made to include at the same time other manifestations which are simply dynamical or chemical, I have elsewhere proposed for the activities characteristic of the organism the term biotics [Sioiiy.ug, pertaining to life). § 5. The philosophy of matter in the abstract is dynami- cal, that of mineral species is both dynamical and chemi- cal, while that of organized forms is at once dynamical, chemical, and biotical. The study of the biotical activi- ties of matter leads to organogeny and morphology, while m THE OKDER OF THE NATUUAL SCIENCES. 29 the relations of organisms to one another, and to the inorganic kingdom, give ns pliysiological botany and zoJJlogy. AVe thus arrive at a com[)reheiisive and simple scheme for the classification of the natural sciences, which is set forth in the subjoined table : — Natural Sciences. Inoroanic Nature. Organic Nature. DESCRU'TIVK. Mineral PiiYSioaRAPHY. BlOlMIYBIOGRAPHY. General Physiography Descriptive and Systematic Organography ; or Mineralogy ; Descriptive and Systema- Natural History. Geognosy ; Geography ; tio liotauy and Zoology. ' Descriptive Astronomy. PuiLosopnicAii. Mineral Physiology. BlOPUVSIOLOOY. General Physiology Dynamics or Physics; Biotics. or Chemistry. Organogeny ; Morphology; Natural Philosophy, Geogeny; Theoretical Physiological Astronomy. Botany and ZoOlogy. ■ 1^ *%-3C3), and that a furthur dlsciieNlon of 8ome of the questionM r.iised lieruli. will be found in the following essay on Celestial Chemistry fk'om the Time of >.OiVton. § 1. Questions concerning the condition of the terres- trial atmosphere in former periods of the eartii's history, and its geological relations, have occupied the attention of naturalists, physicists, and chemists. Brongniart long since suggested that the abundant vegetation of the coal period indicated the existence of a large proportion of carbonic acid in the air at that time. Ebelmen, howe/er, appears to have been the first to clearly understand the great geological significance of the atmosphere, and in his two remarkable memoirs on the decomposition of : >ck8, published in the Annales des Mines in 1845 and 1847,t treated the subject in its atmospheric relations with much research and philosophic breadth. Starting from the chemical changes of crystalline silicate rocks, he consid- ered both the conversion of feldspars into kaolin, and the * A summary of the views presented in this memoir was given at Dublin, in August, 1878, before the British Association for tlie Advance- ment of Science. An abstract thereof appeared in the Proceedings, and will be found in Nature for Aug. 29, 1878 (vol. xviii., p. 475). The principal conclusions of the memoir are also embodied in a communica- tion made by the author to the French Academy of Sciences, and pub- lished in the Comptes Rendus of Sept. 23, 1878 (vol. Ixxxvii., p. 4.'>2). They will moreover be found set forth in the preface to a second edition of the writer's Chemical and Geological Essays (pp. ix.-xix.) published in the spring of the same year. t Fourth Series, vols. vii. and xiii. These memoirs will also be found in the Receuil des Trav. Sclent, de M. Ebelmen; Paris, 1855, vol. ii., pp. 1-79. 80 III.] RELATIONS OF TIIK ATMOHIMIKIIK. 81 decay of protoxide-silicates, such as aniijliibolo and olivine. Tiio sub-aerial decoinjxKsition of the fehlspars had already been shown by IJerthier to result in the separation, in a soluble form, of the protoxide-base ; together with a por- tion of silica, from an insoluble aluminous silicate of definite composition. The analyses of Ebelmen now established the fact that the protoxide-silicates just men- tioned, lose, under similar conditions, the whole of their lime and magnesia, and ncirly the whole of their silica, leaving little behind save the higher oxides resulting from the fixation of atmospheric oxygen by the ferrous and manganous oxides of the silicates ; the soluble bases being in all rases removed by atmospheric waters in the form of carbonates. Such a decomposition of these sili- cates shows that the removal of silica in soluble form does not depend on the intervention of alkalies. § 2. The atmosphere of our earth, at a pressure of 760 millimetres, has a weight of 10,333 kilograms to the square metre, of which the oxygen equals 2376, and the carbonic dioxide (if we take Boussingault and Ldwy's determination of four and a half parts in 10,000 parts by weight) 4.64* kilograms. The alkali of 100 parts of orthoelase would require for its neutralization 7.8 parts of carbonic dioxide, so that a cubic metre of this silicate, of specific gravity 2.5, would, by the calculation of Ebel- men, fix, in the process of decay, 195 kilograms of the gas. From this it results that a layer of orthoelase over the earth of 0.0238 metre, or one of less than 1.0 metre over one-fortieth of its surface, would, in its decomposi- tion, absorb the whole amount of this gas now present in the atmosphere. Ebelmen further calculated that the formation of a layer of kaolin by this process, 500 metres in thickness, would require an amount of carbonic dioxide equal to many times the weight of the present atmos- phere. * This, by an error in Ebelmen's memoir, is given as only 1.24 kilo- gr.ims. 32 THE CHEMICAL AND GEOLOGICAL [in. iiiil I § 3. "We have repeated and extended these calculations, with revised molecular weights, and with the following results : A cuhic metre of orthoclase, with a density of 2.5, and containing theoretically 16.9 per cent of potash, equivalent to 7.89 of carbonic dioxide, would absorb in kaolinization 197.3 kilograms of this gas, while a cubic metre of albite of density 2.6, containing 11.8 of soda, equivalent to 8.37 of carbonic dioxide, would require not less th:i,;i 217.6 kilograms of the same. The figure of 195 kilogram is, adopted by Ebelmen, was thus below the truth, and we may, in view of the considerable proportion of soda-feldspar in the oldest crystalline rocks, convenie tly assume 200 kilograms as the amount of carbonic dioxide required to unite with the alkali from a cubic metre of orthoclase or albite, and form therewith a neutral car- bonate. § 4. In such a decomposition, 100 parts of orthoclase give theoretically about 46.5 parts of kaolin, so that 1.0 metre in thickness of orthoclase of the above density should yield 0.447 metre of kaolin of density 2.6. If we assume this process to have consumed for a cubic metre or 2500 kilograms of orthoclase, 200 of carbonic dioxide, we find that a layer of 51.66 metres of orthoclase, or its equivalent of quartzo-feldspathic rock, in undergoing the same change, would absorb 10,333 kilograms of this gas, equal to the entire weight of the present atmospheric column, and would yield a layer of pure kaolin 23.7 metres in thickness. The production of a stratum of kaolin 500 meters in thickness over the whole surface of the globe, would thus require an amount of carbonic dioxide equal to more than twenty-one times the entire weight of our present atmosphere. § 5. The absorption of this gas in the decay of silicates like hornblende, pyroxene, and olivine is far greater. If we assume, for convenience, a hornblende containing 20.0 per cent of magnesia, and 14.0 of lime, with a density of 3.0 (which figures are not above the average), we find ni.] RELATIONS OF THE ATMOSPHERE. 33 n ic ot 95 th, of tiy dde 3 of car- slase ti.o nsity f we netre ixide, ,r its ig the U gas, |)lierio 23.T km of [ace of •bonic lentire ticates 3r. If kg 20.0 [Sty of re find that it will require 33.0 per cent, or, in round numbers, one-third its weight of carbonic dioxide to convert these two bases into neutral carbonates ; so that a metre-cube of hornblende, weighing 3000 kilograms, would consume not less than 1000 Lllograms of carbonic dioxide. In other terms, the decay of 10^ metres of such hornblende (or its equivalent in hornblendic rock) would absorb 10,333 kilograms, or a whole atmosphere of this gas, being five times as much as is taken up in the kaolin izatiou of the same volume of orthoclase. § 6, The hornblendes in question are seldom without several hundredths of iron as ferrous oxide, which is peroxidized in the process of decay, and, with a little silica, is the chief insoluble residue in the case of non- aluminous hornblendes. In this connection, we revert to a farther calculation by Ebelmen, who pointed out that the conversion of 21,357 kilograms of ferrous oxide into 23,750 kilograms of ferric oxide would consume the whole of the 2373 kilograms of oxygen contained in the present atmosphere ; so that if we suppose the existence over the whole earth of 1000 metres of sediments derived from the decay of crystalline rocks, and containing only one per cent of ferric oxide thus formed, this amount would equal 25,000 kilograms per square metre of surface, requiring for its production from ferrous oxide the absorption of a quantity of oxygen more than equal to that now contained in our atmosphere. § 7. Ebelmen, at the same time, referred to the vrell- known deoxidation of carbonic dioxide by growing vege- tation, and also to the reduction, by decaying organic matters, of sulphates to sulphides, with reproduction of carbonic dioxide, through which the generation of metallic sulphides in ijature gives to the atmosphere, in union with carbon, a portion of the oxygen previously combined with sulphur and with the metals. The following calculations may serve to bring still more fully before us the great geological significance of these 34 THE CHEMICAL AND GEOLOGICAL [UI. k atmospheric changes. The weight of a layer of pure car- bon, with a density of 1.2p and a thickness of 0.7 metre, would require for its conversion into carbonic dioxide the whole of the oxygen of our present atmosphere. The separation of such an amount of carbon by the process of vegetable growth must therefore have liberated the same volume of oxygen. Again, a stratum of carbonate of lime of specific gravity 2.7, covering the earth with a thick* ness of 8.69 metres (or one of dolomite of 2.85, and 7.58 metres thick), would contain an amount of carbonic diox- ide equal in weight to the present atmosphere.* § 8. It was in view of these processes that Ebelmen declared, in 1845, that "the decomposition and the repro- d action of certain mineral species very abundant on the surface of the globe corresponds to important modifica- tions in the composition of the atmosphere." He farther said, "Many circumstances tend to prove that in ancient geological periods the atmosphere was denser, and more rich in carbonic acid, and perhaps in oxygen, than at pres- ent. To a greater weight of the atmospheric envelope would correspond a stronger condensation of the solar heat, anu atmospheric phenomena of a much greater intensity." f Similar conclusions with regard to the physical relations of a denser primeval ainosphere were subsequently announced by the late Edwi'i B. Hunt, in an essay on Terrestrial Thermotics, presented to the American Association for the Advancement of Science, in 1849, and published in its Proceedings for that year, page 135. § 9. We may get a clearer notion of the problem before us by inquiring into the probable amounts of carbonic dioxide which have, in past ages, been abstracted from the atmosphere. In a communication to the British * T. Sterry Hunt on the Primeval Atmosphere, Proc. Amer. Assoc. Adv. Science, 1860, and Can. Naturalist, II., iil., 118. t Ann. des Mines, IV., vii., 05; also Receuil dea Trav. Sclent, de M. Ebelmen. vol. ii., p. 65. ^ ui. III.] RELATIONS OF THE ATMOSPHERE. 35 tre, tlie The ss of same lime thick L7.58 ; diox- 3elmen ! tepro- onthe lodifica- ; farther ancient nd more 1 at pres- I envelope the solar crreater to the lere were Hxnat, in a to the £ Science, that year, ,lem heiore i carbonic vcted from \xQ British Amer. Assoc. Lv. Sclent, de Association for the Advancement of Science, in 1877,* Mr. J. L. Mott concludes, as the result of calculations, that the average amount of unoxidized carbon to a square mile of the earth's crust cannot be less, and is probably many times greater than 3,000,000 tons ; while a layer of 0.7 metres of carbon of density 1.25 (about that of coal), whicli we have calculated to be equal to the total atmos- pheric oxygen, would weigh only about 2,200,000 tons to the square mile. Mr. Mott rightly argues that the pres- ence in the atmosphere of so great an amount of carbon in the form of dioxide would imply a condition of things incompatible with the existence of animal life, and at the same time concludes that its deoxidation would yield an excessive amount of oxygen. He is hence led to assume the existence in the earth of a constant amount of carbon, which is subject to an annual subterranean oxidation equal to the amount of carbon annually removed by vege- tation; the source of the original amount of carbon being, in his hypothesis, left unexplained. § 10. While some have imagined an inorganic origin to the carbon found in the form of graphite, and even to petroleum and to coal, sound reasoning is, we think, on the side of those who, starting from the conception of an originally oxidized globe, see no evidence of any process of deoxidation therein which does not, directly or indi- rectly, depend upon vegetable life, and hence assign an organic origin to all carbons and hydrocarbons. When we take into account the vast amounts of these, from the graphite of Eozoic times to the coals, lignites, and petro- leums of the Tertiary, we can scarcely doubt that the total amount of carbon which has been reduced from car- bonic dioxide is equal to many times the equivalent of the oxygen now present in the atmosphere. Whether the great excess of oxygen thus liberated may perhaps have been absorbed in the production of ferric oxide, as above indicated, is a part of the problem before us. * Nature, vol. xvL, p. 406. 36 THE CHEMICAL AND GEOLOGICAL l! I; 11 ! PII. § 11. It may here be noted that in addition to the fossil carbonaceous bodies already mentioned, the rocky strata of the earth include great thicknesses of pyroschists, which are argillaceous sediments more or less impregnated with hydrocarbonaceous matters allied to coal in compo- sition. To give a single example, Newberry estimates the proportion of such matters diffused through the three hundred or four hundred feet of Devonian black shales which underlie the eastern half of Ohio, to equal fifteen per cent, and to be equivalent to a layer of coal fifty feet in thickness over the whole area.* In this connection it must be considered that the chem- ical composition of the various hydrocarbonaceous fossil substances implies a deoxidation not only of carbonic dioxide but of water. The amount of liberated oxy- gen from the latter would equal, for the different coals and asphalts, from one-eighth to one-fourth, and for the petroleums, one-half of that set free in the deoxidation of the carbon which these hydrocarbonaceous bodies con- tain. § 12. The amount of carbon removed from the atmos- phere in a deoxidized form by vegetation is, however, small when compared with that which has been absorbed during the decomposition of silicates, and is now fixed as insolu- ble carbonates, chiefly in the form of limestones and dolo- mites. That both the alkaline carbonates liberated in the decay of feldspars, and the magnesian carbonate set free in like manner fron magnesian silicates, must decompose the chlorid of calcium contained in the primitive ocean, tliereby giving rise to alkaline and magnesian chlorides on the one hand, and to carbonate of lime on the other, is a consequence which seems to have escaped Ebelmen, and was pointed out by the present writer in 1858. In 1862, however, there was opened a sealed packet which had been in 1844 deposited by Cordier with the French Academy of Sciences, and was found to contain views as * Geology of Ohio, vol. I., page 162. ItELATIONS OF THE ATMOSPHERE. 87 to the origin of limestones and of sea-salt similar to those just stated.* Thus, in the present state of our knowledge, we conclude that all carbonates of lime, whether directly- formed by the decay of calcareous silicates, or indirectly through the intervention of carbonates of magnesia or alkalies, derive their carbonic dioxide from the atmos- phere. The same must be said for the dolomites, magne- sites, and siderites. § 13. We have already shown that a weight of carbonic dioxide equal to more than twenty-one times that of our present atmosphere would be absorbed in the production from orthoclase of a layer of kaolin extending over the earth's surface with a thickness of five hundred metres, an amount which evidently represents but a small propor- tion of the results of feldspathic decay in the sedimentary strata of the globe. The aluminous silicates in the oldest crystalline rocks occur in the forms of feldspars and re- lated species, and are, so to speak, saturated with alkalies or with lime. It is only in more recent formations that we find aluminous silicates either free or with reduced amounts of alkali, as in the argillites and clays, in mica- ceous minerals like muscovite, margarodite, damourite, and pyrophyllite, and in kyanite, fibrolite, and andalusite, all of which we regard as derived indirectly from the more ancient feldspars.f § 14. It has been shown that the disengagement of the carbonic dioxide from a layer of limestone covering the * Hunt, Chem. and Geol. Essays, pp. 2 and 20. t These considerations, and tlielr stratigraphical bearings, first set forth In 1863 (Chem. and Geol. Essays, pp. 27 and 28), will be found further developed in the writer's report on Azoic Koeks, 2d Geol. Survey of Penn., 1878, p. 210. It is a question how far the origin of the various crystalline aluminous silicates named above is to be sought in a process of diagenesis in ordinary aqueous sediments holding the ruins of more or less completely decayed feldspars. Other aluminous rock-forming sili- cates, such as chlorites and magnesian micas, are however connected, througli aluminiferous amphiboles, with the non-aluminous magnesiau silicates, and to all these various magnesian minerals a vei7 dilferent origin must be assigned. [See in this connection Essay V., The Origin of Crystalline Kocks, etc.] 38 THE CHEMICAL AND GEOLOGICAL tm. ^ ' M II i earth's surface with a thickness of 8.69 metres, would double the weight of the atmosphere. The existence of vast formations of limestone and dolomite, often many hundred metres in thickness, throughout all geological periods, will, it is believed, justify the conclusion that the carbonates of the earth's crust are equal to a continuous layer of limestone 869 metres thick, and probably to more than double this amount. From this it would follow that the earth contains, fixed in the form of car- bonates, a quantity of carbonic dioxide, which, if liber- ated in a gaseous form, would be equal in weight to one hundred if not to two hundred atmospheres like the present. A considerable portion of this was doubtless absoibed at an early period in the history of our globe, since the limestones of the Eozoic age are of great thick- ness, and those of more recent times have been in part formed by the solution and re-deposition of portions of these older limestones. § 15. The question now arises, whence came this enor- mous volume of carbonic dioxide which, since the dawn of life on our planet, has been fixed in the form of carbon and carbonates ? The presence of even a small proportion of it at any one time in the terrestrial atmosphere is evi- dently incompatible with the existence of vegetable and animal life, and it may be added that the pressure of a column of this gas less than the minimum of 100 atmos- pheres which we have supposed, would suffice, at ordinary temperatures, for its partial liquefaction; the tension of liquid carbonate dioxide at 30°.7 C. being, according to Mareska and Donny, but eighty atmospheres. We are therefore forced to the conclusion that this gas was gradu- ally supplied from a source either within tlie earth or beyond our atmosphere. § IG. The difficulties of this problem were not over- looked by Ebelmen, though he apparently faile'l to recog- nize tneir full weight. He takes care to remark : " I do not pretend that this immense proportion of carbonic acid m>i KELAL WS OF THE ATMOSPHEUE. 89 ever made part, at any one time, of the terrestrial atmos- phere. ... I see in volcanic phenomena the principal agent which restores to the atmosphere t'he carbonic acid which the decomposition of rocks removes from it." He then inquires whether the carbonic acid (carbonic dioxide) evolved from the earth's interior, comes from the decom- position of carbonates at great depths and high tempera- tures by reactions with silicious matters, or whether we may imagine, with Elie de Beaumont, the existence of an immense reservoir of carbonic acid dissolved in the sup- posed liquid interior of the earth as oxygen is held in fused litharge or in molten silver. In either case, remarks Ebelmen, the cessation of volcanic phenomena would be followed by the removal from the atmosphere of the last traces of carbonic acid, a process which would entail the extinction of all vegetable and animal life. § 17. Of these two suggested sources of the terrestrial carbonic dioxide, a little reflection will show that although the first is doubtless a true one, and will serve to account for that which is so often disengaged from the earth, both in volcanic and non-volcanic regions (having a similar origin to the chlorhydric, sulphuric and boric acids evolved under analogous conditions — namely, the decom- position of saline compounds of aqueous origin),* it by no means meets the requirements of the problem. As pre- ceding calculations have shown, it is not a question of a small amount of carbonic dioxide alternately removed from our atmosphere by sub-aerial reactions and restored to it by subterranean processes, but of a vast quantity of this gas which, at one time or another, has existed in the terrestrial atmosphere, but is now removed from the aerial circulation and locked up in the form of carbonates. § 18. As regards the second source of carbonic dioxide, suggested by Ebelmen after ^lie de Beaumont, it is, un- like tlie last, purely hypothetical. That the globe has a molten interior is, in the present state of our knowledge * Hunt, Chem. and Geol. Essays, pp. 8 and 111. 40 THE CHEMICAL AND GEOLOGICAL [III. ;ii- , ,| of terrestrial physics, very improbable, and if such exists, the notion that it intervenes directly in volcanic phenomena is still more so. The suggestion that such a molten interior might liold dissolved a great volume of carbonic dioxide appears, moreover, to be inconsistent with what we know of the behavior of furnace-slags, which, though formed in atmospheres highly chnrged with this gas, do not, as shown by their behavior in cooling, hold it in solution. The tendencies of modern geological thought and investigation, it may be said, lead to the conclusion that the seat of volcanic phenomena is to be found in sedimentary strata,* and that although the earth's interior intervenes as a source of heat, tiie car- bonic dioxide disengaged from its crust is derived, as in the first hypothesis mentioned by Ebelmen, from the de- composition of carbonates previously generated by sub- aerial re-actions. § 19, The problem still before ns is then to find the source of the vast amount of carbonic dioxide continuously supplied to the atmosphere throughout the geologic ages, and as continuously removed therefrom, and fixed in the form of carbonaceous matters and limestones. We have shown reasons for rejecting the theory which would derive this supply either from the earth's interior or from its own primal atmosphere, tnd must therefore look for it to an extra-terrestrial source. The new hypothesis, which we here advance, starts with the assumption that our atmos- phere is not primarily terrestrial but cosmical, and that the air, together with the water surrounding our earth (whether in a liquid or a vaporous state), belongs to a continuous elastic medium which, extending throughout the interstellary spaces, is condensed around attracting bodies in amounts proportional to their mass and tempera- ture. This universal atmosphere (if the expression may be permitted) would then exist in its most attenuated form in the regions farthest distant from these centres * Chem. and Geol. Essays, pp. 59-67; also, farther, V. § 127. lU.] RELATIONS OF THE ATMOSPHERE. 41 of attraction ; while any change in the gaseous envelope of any globe, whether by the absorption or condensation, or by the disengagement of any gas or vapor, would, by the laws of diffusion and static equilibrium, be felt every- where throughout the universe. § 20. The precipitation of water at the surface of a cooling globe, and its chemical or mechanical fixation there, would thus diminish the proportion of gaseous water throughout all space. The oxygen liberated in the growth of terrestrial vegetation would be shared with the remotest spheres, while the condensatioi: of carbonic dioxide at the surface of our own or any other planet, would not only bring in a supply of this gas from the atmospheres of other bodies, but by reducing the total amount of it, would diminish, pro tanto, the baro- metric pressure at the surface of this and of all other worlds. § 21. The hypothesis here advanced is not wholly new. Sir William R. Grove, in 1842, suggested that the medium of light and heat may be " a universally diffused matter," and subsequently, in 1843, in his celebrated Essay on the Correlation of Physical Forces, in the chapter on Light, concludes, with regard to the atmospheres of the sun and planets, that there is no reason why these atmospheres "should not be, with reference to each other, in a state of equilibrium. Ether, which term wo may apply to the highly attenuated matter existing in the interplanetary spaces, being an expansion of some or all of these atmos- pheres, or of the more volatile portions of them, would thus furnish matter for the transmission of the modes of motion which we call light, heat, etc., and possibly minute portions of these atmospheres may, by gradual accretions and subtractions, pass from planet to planet, forming a link of material communication bettveen the distant monads of the universe.^' Subsequently, in his address as President of the British Association for the Advancement of Sci- ence, in 1866, Grove further suggested that this diffused 42 THE CHEMICAL AND GEOLOGICAL [III. ii U ' Jl matter might become a source of solar heat, inasmuch as the sun "may condense gaseous matter at, i,t travels in space, and so heat may be produced." § 22. Tl>is bold speculation of a universally diffused matter, consiituting an interstellary medium, though thus repeatedly insisted upon by Grove, lias passed almost un- noticed. It seems to have been unknown to Mr. W. Mat- tieu Williams, who, in 1870, published his very ingenious work entitled "The Fuel of the Sun,"* which is baseic^ni., i., 319. S since these pages were in type my attention has been called to a paper read before the Literary and Historical Society of Quebec in Janu- ary, 1870, by James Douglas, Jr., then President of the Society, and one of the Canadian expedition to observe the total solar eclipse of August 7, 18Gi). Therein, while discussing the spectroscopic observations made during the eclipse, he refers to those of Professor Young, who had sug- gested a comparison between certain lines in the spectrum of the solar corona and those observed by Winlock in that of the aurora borealis. With regard to these lines, Mr. Douglas then adds, " May they not there- fore belong to some imknown element ; — a gas lighter than hydrogen, which, like the hypothetical ether, fills space?" To this he adds the suggestion that electricity, both " in the auroral light of our own heavens and the corona of the sun, may render this hypothetical gas luminous." Trans. Lit. and Hist. Soc. of Quebec, New Series, part 7, p. 82. 1 1 P ^'1 ^!il 60 CHEMISTRY 01' 'x. "^. ATMOSPHERE. [UI. when it is considered that the firsb two of these are the only elements of which we have yet any certain evidence in the nebulie, it will be seen that the speculation of Lavoisier is really an anticipation of that view to which spectroscopic stacly has led the chemists of to-day. The three elements named by him are thoiie which, in the forms of air and watery vapor, make up nine hundred and ninety-nine thousandths of the atmosphere which, in ac- cordance with our hypothesis, constitutes the interstellary medium. It was in view of all these considerations that the writer in 1874 ventured to say that "the nebulsB and their resultant worlds may be evolved by a process of chemical condensation from this universal atmosphere ; to which they would sustain a relation somewhat analo- gous to that of clouds and rain to tiie aqueous vapor around us."* Such a speculation, which seeks for a source of the nebulous matter itself, is perhaps a legitimate ex- tension of the nebular hypothesis. * A Century's Progress, etc., cited above; also Chem. and Geol. Essays, preface to 2d ed., p. six. H ,„„ II! j rv. CELESTIAL CHEMISTRY FROM THE TIME OP NEWTON. This Essay, read before the Ph.'.i ophical Socletj of Cambridge, England, No- vember 28, 1881, and published in its Proceedings (Vol. IV., part iii.), was reprinted in the Chemical N^ws, and also in the Aniericau Journal of Science for February, 1882 ([111. J xxxiii., 123-133). A paper on "The Conservation of Solar Enerijy," by C. W. Siemens, was received by the KoyaJ Society of London, February 20, 18H2, and ])ublished in its Proceedings, Number 21!», and also in Nature, vol. xxv., p. 440. Its author therein called attention to my recent essay wiiioh had made known to him the ideas of Newton, and, after repeating my story of the " Hypothesis touching Light and Color," adds, " And now once more a philosopher on the other side of the Atlan- tic brings bacli to the birthplace of Newton his forgotten and almost despised work of two hundred years ago." Siemens admitted, with Grove, Williams, and myself, the existence of attenuated matter in space, which he sui)pased to include oxygen, nitrogen, hydrogen, aqueous vapor, and carbon compounds, besides solid materials, probably exhalations from the sun which constitute the so-called cosmic dust. In my review of the paper of Siemens (Nature, April 27, 1882 ; vol. xxv., p. C13) I have called attention to the fact that already in a communication to the French Academy of Science, September 23, 1878, cited below (Comptes Kcndus, >'ol. xxxviii., p. 452), on the subject of an Interstellary medium as afford; ut ■> means of material commu- nication between celestial bovill be found re- printed, with much other literature on the subject, in a volume by Siemens, in 1883, "On the Conservation of Solar Energy." In explanation of its concluding paragraph it should be said that the author, who had sketched the outli. of the present essay in Italy, some weeks before, and pro- posed to complete it in London, was unexpectedly called thence to Cambridge, a few days before the time which had been assigned for its presentation, and was a guest in quarters In the Master's Court of Trinity College, where, near the rooms formerly occupied by Newton, in the same court, he was obliged to finish the essay which had been promised to the Piiilosophical Society. § 1. The late W. Vernon Harcoiirt, in 1845,* called attention to the remarkable perception of great chemical truths which is apparent in the Queries appended to the third book of Newton's Optics, as well as in his Hypothe- sis touching Light and Color. With regard to the latter, Haicourt then remarked, "It has, I think, scarcely been quoted, except by Dr. Young, and its existence is but little known, even among the best-informed scientific * L., E. and D. Phllos. Magazine, III,, xxvili., 106 and 478; also xxix., 185. 61 i If 1 ' 1 * 111 T'l; f:' m ■a ii' ;iiiH ■ I > m\> M^ 'fill I Mlilllil 11 I 62 CELESTIAL CHEMISTRY PV. men." The essay in question was read before the Royal Society, December 9 and 16, 1675, but remained un- published till 1757, when Birch, at that time secretary to the Society, printed it, not without verbal inaccuracies, in the third volume of his History of the Royal Society ; a work inkiided to serve as supplement to the Philosophi- cal Transactions up to that date. In 1846, at the sugges- tion of Harcourt, the Hypothesis of Newton was again printed in the L., E. and D. Philosophical Magazine (vol- ume xxix.), and it subsequently appeared in the Appendix to the first volume of Brewster's Memoirs of Sir Isaac Newton, in 1855. The time has come for further inquiries into the science of Newton, and I shall endeavor to show that a careful examination of the writings of our great natural philos- opher, in the light of the scientific progress of the last generation, renders still more evident the wonderful pre- vision of him who already two centuries since had anti- cipated most of the recent speculations and conclusions regarding cosmic chemistry. § 2. As an introduction to the inquiries before us, and in order to show the real significance of the speculations of Newton, it will be necessary to review, somewhat at length, the history of certain views, enunciated almost simultaneously by the late Sir Benjamin Brodie, of Ox- ford, and the present writer, and subsequently developed and extended by the latter. In part I. of his Calculus of Chemical Operations, read before the Royal Society, May 3, 1866, and published in the Philosophical Transactions for that year, Brodie was led to assume the existence of certain ideal elements. These, he said, " though now re- ^vealed to us through the numerical properties of chemical equations only as implicit and dependent existences, we cannot but surmise may sometimes become, or may in the past have been, isolated and independent existences.'''' Shortly after this publication, in the spring of 1867, I spent several days in Paris with the late Henri Sainte- IV.] FROM THE TIME OF NEWTON. 58 Claire Deville, repeating with him some of his remarka- ble experiments in chemical dissociation, the theory of which we then discussed in its relations to Faye's solar hypothesis. § 3. From Paris, in the month of May, I went, as the guest of Brodie, for a few days to Oxford, where I read for the first time, and discussed with him, his essay on the Calculus of Chemical Operations, in which connection occurred the very natural suggestion that his ideal ele- ' ments might perhaps be liberated in solar fires, and thus be made evident to the spectroscope. I was then about to give, by invitation, a lecture before the Royal Institu- tion in London on the Chemistry of the Primeval Earth, which was delivered May 31, 1867. A stenographic re- port of the lecture, revised by the author, was published in the Chemical News of June 21, 1867, and in the Pro- ceedings of the Royal Institution.* Therein, I consid -red the chemistry of nebulae, sun, and L'tars in the combined light of (spectroscopic analysis and Deville's researches on dissociation, and concluded with the generalization that the "breaking-up of compounds, or dissociation of elements, by intense heat is a principle of universal appli- cation, so that we may suppose that all the elements which make up the sun, or our planet, would, when so intensely heated as to be in the gaseous condition which all matter is capable of assuming, remain uncombined; that is to say, would exist together in the state of chemical ele- ments ; whose further dissociation in stellar or nebulous masses may even give us evidence of matter still more elemental than that revealed in the experiments of the laboratory, where we can only conjecture the compound nature of many of the so-called elementary substances." § 4. The importance of this conception, in view of subsequent discoveries in spectroscopy and in stellar chemistry, has been well set forth by Lockyer in his late * See also Hunt's Chemical and Geological Essays, pp. 35-45. 11 tiL I' 54 CELESTIAL CHEMISTRY [IV. lectures on Solar Phj^sics,* where, however, the general- ization is described as having been first made by Brodie in 1867. A similar but later enunciation of the same idea by r^erk-M ' we'l is also cited by Lockyer. Brodie, in fact, ;; ti.:. yth Oi June, one week after my own lecture, gave i. 'ii.!Vi > on Ideal Chemistry before the Chemical Society >■ Lojdou, published in the Chemical News of June 14, in wIjs ... with regard to his ideal elements, in further extension of the suggestion already put forth by him in the extract above given from his paper of May 6, 1866, he says, "We may conceive that in remote ages the temperature of matter was much higher than it is now, and that these other things [the ideal elements] existed in the state of perfect gases — separate existences — un- combined." He further suggested, from spectroscopic evidence, that it is probable that " we may one day, from this source, have revealed to us independent evidence of the existence of these ideal elements in the sun and stars." During the months of June and July, 1867, I was absent on the continent, and this lecture of Brodie's remained wholly unknown to me until its republication in 1880, in a separate form, by its author,f with a preface, in which he pointed out that he had therein suggested the probable liberation of his ideal elements in the sun, referring at the same time to his paper of 1866, from which we have already quoted the only expression bear- ing on the possible independence of these ideal elements somewhere in time or in space. § 5. The above statements are necessary in order to explain why it is that I have made no reference to Sir Benjamin Brodie on the several occasions on which, in the interval between 1867 and the present time, I have reiterated and enforced my views on the great significance of the hypothesis of celestial dissociation as giving rise to * Nature, August 25, 1881, vol. xxiv., p. 396. t Ideal Chemistry, a Lecture. Macmillan, 1880. IV.] FROM THE TIME OF NEWTON. 66 forms of matter more elemental than any known to us in terrestrial chemistry. The conception, as at first enun- ciated in somewhat different forms alike by Brodie and myself, was one to which we were both naturally, one might say inevitably, led by different paths from our respective fields of speculation, and which each might accept as in the highest degree probable, and make, as it were, his own. I write, therefore, in no spirit of invidious rivalry with my honored and lamented friend ,it ">'mply to clear myself from the charge, which miglit ot- rwise be brought against me, of having on vari^ ^ oc a.nons within the past fourteen years put forth d enlarged upon this conception without mentioning Sii Benjamin Brodie, whose only publication on the subi -^t, so far as I am aware, was his lecture of 1867, unkno\s i > me until its reprint in 1880. § 6. It was at the grave of Priestley, in 1874, that I for the second time considered the doctrine of celestial dissociation, commencing with an account of the hy- pothesis put forward by F. W. Clarke, of Cincinnati, in January, 1873,* to explain the growing complexity which is observed when we compare the spectra of the white, yellow, and red stars ; in which he saw evidence of a pro- gressive evolution of chemical species, by a stoichiogenic process, from more elemental forms of matter. I next referred to the further development of this view by Lockyer in his communication to the French Academy of Sciences in November of the same year, wherein he connected the successive appearance in celestial bodies of chemical species of higher and higher vapor-densities with the speculations of Dumas and Pettenkofer as to the com- posite nature of the chemical elements.f I then quoted from my lecture ©f 1867 the language already cited, to the effect that dissociation by intense heat in stellar * Clarke [now of Washington, D. C], on " Evolution and the Spectros- cope," Popular Science Monthly, New York, vol. ii., p. 32. t Lockyer, Coniptes Kenilus, Xovembcr 3, 1873. ■#■: I 56 CELESTIAL CHEMISTRY i ii- ll{ j : i ; n ,/\ ■; t CI I [E?* worldi? might give us more elemental forms of matter than any known on earth, and further suggested that the green line in the spectrum of the solar corona, which had been supposed to indicate a hitherto unknown substance, may be due to a " more elemental form of matter, which, thougli not seen in the nebulte, is liberated by the intense heat of the solar sphere, and may possibly correspond to the priniary matter conjectured by Dumas, having an equivalent weight one-fourth that of hydrogen." § 7. The suggestion of Lavoisier, that " hydrogen, nitrogen, and oxygen, with heat and light, might be re- garded as simpler forms of matter, from which all others are derived," was also noticed in connection with the fact that the nebuloe, which we conceive to be condensing into suns and planets, have hitherto shown evidences only of the presence of the first two of these elements, which, as is well known, make up a large part of the gaseous envelope of our planet, in the forms of air and aqueous vapor. With this, I connected the hyj)othesis that our atmosphere and ocean are but portions of the universal medium which, in an attenuated form, fills the intcrstel- lary spaces; and further suggested, as "a legitimate and plausible speculation," that " these same nebulaj and their resulting worlds may be evolved by a process of chemical condensation from this universal atmosphere, to which they would sustain a relation somewhat analogous to that of clouds and rain to the aqueous vapor around us." * § 8. Tliese views were reiterated in the preface to a sec- ond edition of my Chemical and Geological Essays, in 1878, and again before the British Association for the Advance- ment of Science at Dublin,! '^^^^ before the French Acad- emy of Sciences in the same year.| They were still fur.^her * A Century's Progress in Theoretical Chemistry, being an address at the grave of Priestley in Northumberland, Penn., July 31, 1874; Amer. Chemist, vol. v., pp. 46-61 and Pop. S^ence Monthly, vi., p. 420. t Nature, August 29, 1878, vol. xviii., p. 475. J Comptes Rendus, September 23, 1873, vol. xxxviii., p. 452. ' iv.i FROM TH.'i: TIME OF NEWTON. 6T developed in an essay on the Chemiciil and Geological Relations of the Atmosphere, published in May, 1880 (^ante pages 30-50), in .ich attention was called to the impor- tant contribution to the subject by Mr. Lockyer in his in- genious and beautiful spectroscopic studies, the results of which are embodied in his " Discussion of the W(jrking Hypothesis that the so-called Elements are Compound Bodies," communicated to the Royal Society, December 12, 1878. It was then remarked that the already noticed "speculation of Lavoisier is really an anticipation of that view to which spectroscopic study lias led the chemists of to-diiy " ; while it was said that the hypothesis put forth by the writer in 1874, "which seeks for a source of the nebulous matter itself, is perhaps a legitimate extension of the nebular hypothesis." § 9. To show the connection of the above views with the philosophy of Newton, it now becomes necessary to give some account of the conception of the universal dis- tribution of matter throughout space, both as regards its dynamical relations and its chemical composition. Pass- ing over the speculations of the Greek physiologists, we conje to the controversies on this subject in the seven- teenth century, and find, in apparent opposition to the plenum maintained by Descartes and his followers, the teaching of Newton that "the heavens are void of all sensible matter." This statement is, however, qualified elsewhere by his assertion, that "to make way for the reg- ular and lasting movements of the planets and comets, it is necessary to empty the lieavens of all .natter, except perhaps some very thin vapors, steams, and effluvia, arising from the atmospheres of the earth, planets, and comets, and from such an exceedingly rare ethereal medium as we have elsewhere described," etc. (^Opties^ Book III. Query 28.) § 10. In order to understand fully the views of Newton on this subject, it is necessary to compare carefully his various utterances, including the Hypothesis, in 1675, the 58 CELESTIAL CHEMISTRY UV. first edition of the Princlpia^ i»i 1087, the second edition, in 1713, and tlio various editions of tlie Optics. This work iip[)eured in 1704, tjje tliird book, with its appended queries, having, according to its autlior's i)reface, been "put together out of scattered papers," subse([uent to the publicati(ni of the first edition of tlie Principia. Tiie Latin transhition of iha Optics, by Dr. Clarke, which was published in 1706, and the second English edition, in 1718, contiiin successive additions to these queries, which are indicated in the notes to Ilorsley's edition of the works of Newton, and are important in this connection. From a collation of all these, we learn how the concep- tions of the Hypothesis took shape, were reinforced, and in great part incorporated in the Principia. § 11. In the Hypothesis, he ir^agines "an ethereal me- dium much of the same constitution with air, but far rarer, subtler, and more elastic." "But it is not to be supposed that this medium is one uniform matter, but composed partly of the main phl3gmatic body of ether, partly of other various ethereal spirits, much after the manner that air is compounded of the phlegmatic body ot air intermixed with various vapors and exhalations." Newton further suggests in his Hypothesis that this com- plex L-pirit or ether, which, by its elasticity, is extended throughout all space, is in continual movement and inter- change. " For nature is a perpetual circulatory worker, generating fluids out of solids, and solids out of fluids, fixed things out of volatile, and volatile out of fixed, sub- tile out of gross, and gross out of subtile ; some things to ascend and make the upper terrestrial juices, rivers, and the atmosphere, and by consequence others to descend for a requital to the former. And as the earth, so perhaps may the sun imbibe this spirit copiously, to conserve his shining, and keep the planets from receding farther from him ; and they that will may also suppose that this spirit affords or carries with it thither the solary fuel and mate- rial principle of life, and that the vast ethereal spaces be- nil i Ijji iv.i FROM THE TIME OP NEWTON. 69 tween us and the stars are for a suflicient rcpositor^ for this food of the sun and phinets." § 12. The hiiiguage of tliis hist sentence, in whicli his Lite biographer, Sir David lirewster, regards Newt(jn as ''amusing himself with the extravagance of his s[)ecuhi- tions," at which "we may bo aUowed to smile," * was not apparently regarded as unreasonable by its author when, more than ten years 1 ter, ho quoted it in the poststi-ipt of his letter to Ilalley, dated Cambridge, Juno 20, 1080. The views therein contained, with the single exception of the suggestion regarding gravitation, have not wanted ad- vocates in our own time, and many of them were endjodied in the Principia, which Newton was then engaged in writing. § 13. But this was not all : Newton saw in the cosmic circulation and the mutual convertibility of rare and dense forms of matter a universal law, and rising to a still bolder conception, wiiieh completes his Hypothesis of the Universe, adds : " Perhaps the whole frame of nature may be nothing but various contextures of some certain ethe- real spirits or vapors, condensed, as it were, by precipita- tion, much after the same manner that vapors are con- densed into water, or exhalations into grosser substances, though not so easily condensible ; and after condensation wrought into various forms, at first by the immediate hand of the Creator, and ever since by the power of na- ture, which, by virtue of the command 'increase and mul- tiply,' became a complete imitator of the copy set lier by the g- 'at Protoplast. Thus, perhaps, may all things be originated from ether." § 14. If now we look to the third book of the Prin- cipia, we shall find in Proposition 41 the remarkable chemical argument by which Newton was led to reg.ird the interstellary ether as affording "the material principle of life" iuid "the food of planets." Considering the ex- halations from the tails ot comets, he supposes thai: the * Brewster's Memoirs of Newton, vol. i., pp. 121 and 404. Ill iiili \v\ M jiBmSSmOam 60 CELESTIAL CHEMISTRY ipr. vapors thus derived, being rarefied, dilated, and spread through the whole heavens, are by gravity brought within the atmospheres of the planets, where they serve for tlie support of vegetable life. Inasmuch, moreover, as all veg- etation is supported by fluids, and subsequently, by decay, is, in part, changed into solids, by which the mass of the earth is augmented, he concludes, that if these essential matters were not supplied from some external source, they must continually decrease, and at last fail. This vital and subtile part of our atmosphere, so important, though small in amount, might, he then supposed, come from the tails of comets.* § 15. This appeared in the first edition of the Prineipia, in 1687. It was not until later that the conception of exhalations from other celestial bodies took shape in the mind of Newton, as we may learn from the Optics. Thus, in the first edition of this work, in Query 11, the sun and fixed stars are spoken of as great earths, intensely heated, and surrounded with dense atmospheres which, by their * " Vapor enim in spatiis illis lib idrais perpetub rarescit ac dilatatur. Qua ratione fit ut cauda omnis ad extremitatem superiorem latior sit quam juxta capita cometae. Ea autem rarefactione vaporem perpetu6 dilatatura diif undi tandera et spargi per coelos imiversos, deinde paulatim in planetas per gravitatem suani attrahi et cum eorum atmospliaeris rais- ceri, ration! consentaneum videtur. Nam quemadmodum maria ad con- stitutionem Terrae liujus omnino requiruntur, idque ut ex iis per calorem Solid vapores copiose satis excitentur, qui vel in nubes coacti decidant in plu\ iis, et Terram omnera ad procreationem vegetabilium irrigeut et nutriant; vel in frigidis montium verticibus condensati (ut aliqui cum ratione phi'osophantur) decurrant in fontes et flumina: sic ad eonserva- tionem marium et humorum in planetis requiri videntur cometaj, ex quo- iTira exhalationibus et vaporibus condensatis, quicquid liquoris per vege- tationem et pntrefactionera consumitur et in Terram aridam convertitur, continu5 suppleri et refici possit. Nam vegetabilia omnia ex liquoribus omnino crescunt, dein magnii ex parte in Terram aridam per putrefac- tionem abeunt, et limus ex liquoribus putrefactis perpetuo decidit. Hinc moles Terrae aridae in dies augetur. e' uores, nisi aliunde augmentum sumerent, perpetuo decrescere deberem,, ac tandem deficere. Porro sus- picor spiritum ilium, qui ai-ris nostri pars minima est, sed subtillissima et optima, et ad rernm omnium vitam requiritur, ex cometis praecipue venire."— iVeio ten, rrincipia, lib. ill., prop. xli. I I rv.] FROM THE TIME OF NEWTON. 61 weight, condense the exhalations arising from these hot bodies. To this Query is added, in 1706, the suggestion that the weight of such an atmosphere " may hinder the globe of the sun from being diminislied except by the emission of light " ; while in the second English edition, in 1718, we find a further addition in the words, " and a very small quantity of vapors and exhalations." A simi- lar change of view appears in the Query now numbered 28, wherein we read of "places [almost] destitute of matter," and also that " the sun and planets gravitate towards each other without [dense] matter between." In these quotations the two words in brackets are wanting in the edition of 1706, and first appear in that of 1718 ; while the language which we have in a previous page quoted from • this same Query, is found in the edition of 1706. § 16. The Queries now numbered 17-24, appeared for the first time in the edition of 1718, and herein we find, in 18, the ethereal medium spoken of as being "by its elas- tic force expanded through all the heavens.". Of this medium, "which fills all space adequately," he asks, "may not its resistance be so small as to be inconsiderable," and scarcely to make any sensible alteration in the movements of the planets?* This complex ether of the interstellary space wab thus, in the opinion of Newton, made up in part of matter common to the planetar}^ and stellar atmos- pheres, the origin and importance of which is concisely stated in the paragraph which appears for the first time in 1713, in the second edition of the Principia, in the third book, at the end of Proposition 42, here much augmented. In this statement, which serves to supplement and com- plete that already made in 1687, in Proposition 41, we read that the vapors which arise alike from the sun, the fixed stars, and the tails of comets, may by gravity fall into the atmospheres of the planets, and there be con- densed, and pass into the form of salts, sulphurs (^id est, * Compare this with Prop, x., Book III., of the Principia. I' CELESTIAL CHEMISTRY [IV. combustible matters), tinctures, clay, sand, coral, and other terrestrial substances.* § 17. The conception of Newton, who, while rejecting alike the plenum of the Cartesians, with its vortices, and an absolute vacuum, imagined space to be filled with an exceedingly attenuated matter, through which a free cir- culation of gaseous substances might take place between distant worlds, has found favor among modern thinkers, who seem to have been ignorant of his views. Sir Wil- liam Grove in 18-12 suggested that the medium of light and heat may be "a universjilly diffused matter," and sub- sequently, in 1843, in the chapter on Light, in his "Es- say on the Corrc'ution of Physical Forces," concluded with regard to the atmospheres o! the sun and the planets, that then is no reason "why these atmosplieres should not be, wiih reference to each other, in i state of equilib- rium. Ether, which term we may apply to the highly attenuated matter existing in the interplanetary spaces, being an expansion of some or all of these atmospheres, or of the more volatile portions of them, would thus furnish matter for the transmission of the modes of motion which we call light, heat, etc. ; and possibly minute portions of the atmospheres may, by gradual accretions and subtrac- tions, pass from planet to planet, forming a link of mate- rial communication between the distant monads of the universe." Subseqaently, in his address as Pi-esident of the British Association for the Advancement of Science, in 1866, Grove further suggested that this diffused matter may become a source of solar heat, " inasmuch as the sun may condense gaseous matter as it travels in space, and so heat may be produced." § 18. Humboldt, also, in his Cosmos, considers the ex- * " Vapores,autem, qui ex Sole et stellis fixls et caudls eometarum oriuntur, incidere possunt per gravitatom siiam in atinospliaeras planeta- ruin, et ibi coiulensari et couverti in aquam et spiritos luuniilos, et siiliinde per lentiun calorcm in sales, et sulphura, et t.nt'turas, et liniiun, et Intiun, et argillani, et arenaui, et lapiiles, et coralla, et substantias alias terrestres paulatini inigrare." — Newton, Principia, lib. iii., prop. xlil. IV.] FROM THE TIME OP NEWTON. 63 istence of a resisting medium in space, and says " of tliis impeding ethereal and cosmical matter," it may be sup- posed that it is in motion, that it gravitates, notwith- standing its great tenuity, that it is condensed in the vicinity of the great mass of the sun, and that it may include exhalations from comets ; in which connection he quotes from the 42d Proposition of the third book of of the Prlncipia. He further speaks comprehensively of "the vaporous matter of the incommensurable regions of space, whether, scattered without definite limits, it exists as a cosmijal ether, or is condensed in nebulous masses and becomes comprised among the agglomerated bodies of the universe." * Humboldt also cites in this connection a suggestion made by Arago in the Annuaire du Bureau des Longitudes for 1842, as to the possibility of determining, by a comparison of its refractive power with that of terrestrial gases, the density of " the ex- trer. ely rare matter occupying the regions of s^mce." f § 19. In 1854, Sir William Thomson published his note on the Possible Density of the Luminiferous Ether,J wherein he remarks, "that there must be a medium of material communication throughout space to the remotest visible body, is a fundamental conception of the undu- latory theory of light. Whether or no this medium is (as appears to me most probable) a continuation of our own atmosphere, its existence cannot be questioned." He then attempts to fix an inferior limit to the density of the luminiferous medium in interplanetary s})ace, by con- sidering the mechanical value of sunlight, as deduced from the value of solar radiation and the mechanical equivalent of the thermal unit. He concludes " that the luminiferous medium is enorijously denser than the con- tinuation of the terrestrial atmosphere would be in inter- * Cosmos, Otte's translation, Harper's ed., vol. 1., pp. 82, 86. t Ibid., vol, lii., p. 40. t Trans. Koy. boc, Edinburgh, vol. xxi., part 1 ; and Phil. Mag^, 1855, vol. ix., p. 36. '"'"'™- '■"■"" 64 CELESTIAL CHEMISTRY [IV. ' I ' 1 1 '''11 i i ^1 liHI planetary space if rarefied according to Boyle's law always, and if the earth were at rest in a state of con- stant temperature, with an atmosphere of the actual density at its surface." The earth itself in moving through space "cannot displace less than 250 pounds of matter." [ P. Glan, who has since examined the ques- tion, concludes that the lower limit of density would be more than 7000 times greater than that calculated by Thomson.*] § 20. In 1870, W. Mattieu Williams published his very ingenious work entitled "The Fuel of the Sun," in which, apparently without any knowledge of what had been written before with regard to an interstellary medium, he attempts to find tlierein the source of solar heat — -the "solary fuel" of Newton. To quote his own huiguage, "the gaseous ocean in which we are iunnersed is but a portion of the infinite atmosphere that fills the whole solidity of space, that links together all the elements of the universe, and diffuses among them light and heat, and all the other physical and vital forci::s which heat and light are capable of generating." (Loc. cit. p. 5.) § 21. [In 1872, ap] t,j; d the remarkable work of Zoll- ner, '■'•Uber .nc Natuy n.rr i mieten^'' in which tlie view of an interstellary atmospnere is set forth with great clear- ness. His conclusions may be gathered from the follow- ing extracts from an extended review and analysis of the volume, published in the same year. Reasoning from known facts it is maintained that even such fixed bodies as the metals, at very low temperatures, are constantly giv- ing off vapor, though in amount too small to be recognized by ordinary tests; whence "it follows that a mass of mat- ter in space will ultimately surround itself with its own vapor, and the tension of the latter will depend upon the mass of the body — that is, upon its gravitative energy — * Ann!»i.;n der Physifc unci Chemie, No. vlii., 1870; cited by the author in a revitnv of Siemens, in " Nature " for April 27, 1882 ; nlso in " Conser- vation of Solar Energy," by yiemena, 1883, p. 33. IV.] FROM THE TIME OF NEWTON. 65 iw )n- iial ing Lies- . be , by very hich, been m, lie — the i-uage, but a whole jments it and which p. 5.-) f ZoU- iew of It cleav- follow- of the from bodies itly giv- oo-nized of mat- its own Jipon the bergy — Ithe author " Conser- Irv and the temperature. If the mass of the body is so small that its attractive force is insuthcient to give to the en- veloping vapor its maximum tension for the existing tem- perature, the evolution of vapor will be continuous, until the whole mass is converted into it." § 22. [" Then comes the question whether a mass of gas or vapor under these circumstances would be in a state of stable equilibrium. The analytical discussion of this point leads to the result that in empty and unlimited space, a finite mass of gas is in a condition of unstable equilib- rium and must become dissipated by continual expansion and consequent decrease of density. A necessary conse- quence of this result is that tlie celestial spaces, at least within the limits of the stellar universe, must be filled with matter in the form of gas, preeminently that of the terrestrial "^"mosphere. Any solid body in space must, by virtue of its gravitative energy, condense the gas, to form an atmosphere upon its surface, and the density of this gaseous envelope can readily be calculated when the size and mass of tlie body are known." Zollner then proceeds to discuss the density of the atmospheres sur- rounding the various bodies of the solar system, and to calculate that of the interstellary sj)aces, where he con- cludes that the gaseous medium would be so ra tliat it " could have no appreciable effect either iq:)on i rays of light or upon the motion of bodies in space." *] § 23. Since the days of Newton, however, no one, so far as I am aware, had liitherto considered the iiiiorstellary matter from a chemical point of view. In 1874, as al- ready shown, the writer had, in extension ol lie concei)- tio.n of Humboldt that its condensation gives rise to nebula), ventured the suggestion that from an ethereal medium having the same composition as our own atmos- pliere, the chemical elements of the sun and tlie planets have been evolved, in accordance with the views of Brodie, Clarke, and Lockyer, by a stoichiogenii ^)rocess : * Amer. Jour. Science, 18" 2, iii., 47G. I i' 66 CELESTIAL CHEMISTRY liv. 80 that, in the language of Newton's Hypothesis, " all things may be origin-ited from ether." § 24. It was not, however, until 1878, that, from a consideration of the chemical processes which have gone on at the earth's surface within recorded geological time, I was led to another step in this inquiry. That all the deoxidized carbon found in the earth's crust in the forms of coal and graphite, as v/ell as that existing in a diffused state, as bituminous or carbonaceous matter, has come, through vegetation, from atmospheric carbonic acid, ap- peal's certain. To the same source we must ascribe the carbonic acid of all the limestones v/hich, since the dawn of life on our earth, have been deposited from its waters It is through the sub-aerial decay of crystalline silicated rocks, and the direct formation of carbonate of lime, or of carbonates of magnesia and alkalies which have reacted on the calcium-salts of the primeval ocean, that all lime- stones and dolomites have been generated. These, apart from th3 coaly matter, hold, locked up and withdrawn from the aerial circulation, an amount of carbonic acid which may be probably estimated at not less than 200 atmo,«pheres equal in weight to our own. That this amount, or even a thousandth part of it, could have existed at anyone time in our terrestrial atmosphere since the beginning of life on our planet is inconceivable, and ihht it could be supplied from the earth's interior is an hj^iothesis equally untenable. § *'.5. I was therefore led to admit for it an extra-ter- rettriai .source, and to maintain that the carbonic acid has thei!?e giadually come into our armosphere to supply the deti neiicies created by chemical processes at the earth's S'u ace. Since similar processes are even now remov- iii from our atmosphere this indispensable element, and fix g it in solid forms, it follows that, except volcanic agency, which can only restore a portion of what was pri- marily derived from the atmosphere, there are on earth, besides organic decay, only the artificial processes of IV.1 FROM THE TIME OP NEWTON. 67 all human industry wliioh can furnish carbonic acid ; so that but for a supply of this ing back to the specula- tions of the ancient philosophers, we find those of the last two centuries, Newton, Descartes, Leibnitz, and Buffon, among others, accepting the hypothesis of a former igneous condition of our planet. Starting from this basis, the phenomena of volcanoes, and the resem- blances between their consolidated lavas and many of the crystalline rocks, naturally gave rise to the notion of the igneous origin of these, which was formulated in the hy- pothesis that all such rocks, whether massive or schistose, were directly formed during the cooli ig and consolidation of a molten globe. § 2. Playfair, in his " Illustrations of the Huttonian Theory of the Earth," tells us that it was Lehman, who, V,] THE ORIGIN OF CRYSTALLINE ROCKS. 69 ill 17r>6, first distinguislied by the name of Primitive the ancient crystalline rocks, described by him as arranged in beds, -ertical or highly inclined in attitude, and overlaid by horizontal stiata of secondary origin. These primitive rocks were by Lehman regarded " as parts of the original nucleus of the globe, which had undergone no alteration, but remained such as they were first created." This view was shared by Pallas and by De Luc, the latter of whom at one time considered the primitive rocks "as neither stratilied nor formed by water," though, as Playfair in- forms us, De Luc subsequently admitted " their formation from aqueous deposition, as the neptunists do in general." * Pallas held a similar view, and, according to Daubrde, both Pallas and Saussure " admitted, as Linnajus had done, that all the terranes have been formed by the agency of water, and that volcanic phenomena are but local accidents." Pallas published his " Observations on Mountains " in 1777, and Saussure the first volume of his " Voyages dans les Alpes " in 1779. It was about 1780 that the celebrated professor of Freiberg began, in his lectures, the exposition of his views, called by Playfair " the Neptunian system as improved by Werner " ; though his Classification of the Rocks, in which these views were finally embodied, dates only from 1787. § 3. According to Werner, the materials which now form the solid crust of the globe were deposited from the waters of a primeval ocean, in which the elements of the crystalline rocks were at one time dissolved, and from which they were separated as chemical precipitates. The granite, which he regarded as the fundamental rock, was first laid down, and was closely followed by the gneisses and the hornblendic and micaceous schists. When the dissolving ocean covered the whole globe to a great depth, * John riayfair, loc. cit., pp., 160, 102. The Theory of the Earth, by James Ilittton, first appeared in 1785, and in a second edition in 1705. riayfalr's celebrated exposition of it, here quoted, was published in Edin- burgh in 1802. oaiUn.Ul.Mru, ^ mmt TO THE ORIGIN OP CRYSTALLINE ROCKS. [V. ' (I ■li-i and its waters were tranquil and pure, the rocks depos- ited were exclusively crystalline, and, like the ocean, they were universal. These he distinguished as the Primitive rocks. At r. later period, the depth of the ocean was supposed to have been diminished by the retreat of a portion of the waters to cavities within the globe ; a notion apparently borrowed from Leibnitz, who imagined caverns, left by the cooling of a formerly fused mass, to have subsequently served as reservoirs for a part of the universal ocean. In this second period, according to Werner, a chemical de- position of silicates still went on, but dry land having been exposed and shallows formed, currents destroyed portions of the previously deposited masses, which were also attacked by atmospheric agents. By these actions were formed mechanical sediments, which became inter- stratified with those of chemical origin. It was during this period of coincident chemical and mechanical deposi- tion that were formed the Intermediate or Transition rocks of Werner, which, from the conditions of their for- mation, necessarily covered portions only of the universal Primitive series. At a still later period, marked by a farther diminution of the superficial waters, were laid down the Secondary rocks of Werner, at a time when the sea no longer produced mineral silicates, and had assumed essentially its present composition. § 4. The Primitive rocks, according to this hypothesis, were those composed entirely of chemical deposits, which are either crystallized or have a tendency to crystalliza- tion, and in which the action of mechanical causes cannot be traced. In the Transition series, the products of chemical and mechanical processes are intermingled, and materials derived from the disintegration and decay of Primitive rocks are present ; while the rocks of the Secondary series were formed from the ruins alike of the Pi'imitive and the Transition series. During the process of their consolidation, the various strata having been v.i THE ORIGIN OF CRYSTALLINE ROCKS. 71 broken, fissures were formed through wliich the surplus waters retired to the internal cavities, depositing on the walls of the fissure^" through which the}' descended the various matters still iield in solution. In this way were formed metalliferous and other mineral veins. The aqueous solution in which all these crystalline rocks were at first dissolved was described by Werner and his disciples as a chaotic liquid, and he even desig- nated the rocks themselves as chaotic, " because they were formed when the earth's surface was a chaos." These Primitive rocks, consisting of the granite and the over- lying crystalline sciiists, covered the whole earth, and their geograi)hical inequalities were due to the original deposition, which did not yield a regular surface, but presented elevations, upon the slopes of which were sub- sequently laid down the Transition strata. Such, according to Werner, was the origin of all rock- masses except recent alluvion^, deposits of obviously organic origin, and the ejections of volcanoes, which he conceived to be due to the subterraneous combustion of carbonaceous deposits. In the earlier ages of the world there were, according to him, no volcanoes and no evi- dences of subterranean heat. Neither in the formation of granite, of basalt, of the crystalline schists, or of mineral veins, or in the displacements of the strata to be seen in the deposits of various ages, did he recognize any mani- festations of an internal activity of the earth.* § 5. We now pass to the consideration of the rival geological theory of Hutton, which was developed at the same time with that of Werner. Saussure, as early as * In preparing the foregoing synopsis of the views of Werner, I have followed ill part, the exposition of his system given by Mnrray in his Re- view of 'Playfair's Illnstrations of the Iluttonian Tlieory, pnblishecl anonymously in Edinburgh in 1802; in part the statements to be found in Playfair, in Bakewell, in Lyell, and in Naumann; and also the excel- lent analysis given by Daubree in his fctudes et Experiences Synthetiques 8ur le Metamorphisme, et sur la Formation des Roches Cristallines; Paris, 1860. e> vO. % IMAGE EVALUATION TEST TARGET (MT-3) 1.0 I.I ^ 1^ III 2.2 !t 1^ lllllio 1.8 25 lilU il.6 V] y^ % y. % y ^J y PhotDgraphic Sciences Corporation \ ;V •NJ \ \ rv ^^^ >^1j 6^ ■^ 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716)S73-4S03 '9) :\ <\ 72 THE ORIGIN OP CRYSTALLINE ROCKS. [V. 1776, had ascribed to aqueous infiltration the granitic veins in the Valorsine, and others near Lyons — a view which was shared by Werner, who, from their similar constitution, conceived that the formation of massive and stratiform granitic rocks had taken place under conditions •like tliose which gave rise to the veins in question, and then extended this view to other veins and masses of what we must regard as injected or irrupted rocks, including not only granites but dolerites and basalts. Hutton and his interpreter, Playfair, on the other hand, regarded all granitic veins as having been filled by injec- tion with matter in a state of igneous fusion, repudiating the notion of Saussure and of Werner that such materials could be formed by crystallization from aqueous solutions. Granitic veins, according to Hutton, are in all cases bit ramifications of great masses of granite, themselves often concealed from view. "In Hutton's theory, granite is regarded of more recent formation than the strata incum- bent upon it ; as a substance which has been melted by heat, and which, forced up from the mineral regions, has elevated the strata at the same time." * From this con- dition of igneous liquidity, he supposed, had crystallized alike quartz and feldspar, as well as tourmaline and the other minerals sometimes found in granitic veins. Granite is elsewhere declared by him to be matter fused in the central regions of the earth. § 6. With Werner, granite was the substratum under- lying all other known rocks, simply because it had been the first deposit from the chaotic watery liquid, and it was said to pass into or to alternate with the distinctly strati- form or schistose crystalline rocks. In this view of its geognostical relations, Werner was strictly correct if by granite we understand the massive or indistinctly strati- form aggregate which makes up what some call granite and others fundamental granitoid gneiss. This is what I have called an indigenous rock, which may be with or * Playfair, Illustrations, etc., p. 89. ' If ^ ■ ■ ■ v.] THE ORIGIN OF CRYSTALLINE ROCKS. 73 without apparent stratification. We must, however, dis- tinguish, besides this first type of crystalline rock, — the underlying granite of Werner, — two others which, though mineralogically similar, and often confounded, are geognostically distinct. Of these, what I have called EXOTIC rocks consist apparently of softened and displaced portions of aggregates of the first type, and are met with alike in dikes and in masses of greater or less size, in- truded or irrupted among the stratified or indigenous rocks. These are the typical granites of Hutton. The third type includes those concretionary masses of granitic material formed in fissures or cavities, which are evidently deposits from aqueous solutions. These are the infiltrated veins of Saussure and of Werner, and are what I have designated endogenous rocks. § 7. By keeping in view this threefold distinction between indigenous, exotic, and endogenous granitic aggregates, as I have long since endeavored to show, the obscurities and apparently contradictory views of different observers are easily explained. These distinctions are recognized in other crystalline rocks than granite. Under the name of crystalline limestones, as is well known, have been included both indigenous and endogenous masses. The question whether or not certain crystalline silicated rocks are to be regarded as eruptive, is seen to be of minor importance, when we consider that it is possible for indigenous crystalline deposits to appear in the rela- tion of exotic masses, whether displaced in a softened and plastic condition, as more generally happens, or else forced, in rigid masses, among softer and more yielding strata, as appears, from the observations of StaptT, to be the case of the serpentines of Mount St. Gothard.* § 8. Werner argued, and, as we shall endeavor to show, correctly, from their analogies with concretionary granitic veins, that all granitic rocks were deposited from water, and are consequently indigenous or endogenous in origin. * See Essay X.,§12&-1S0. .' 74 THE OKIGIN OF CRYSTALLINE ROCKS. [V. He denied the existence of exotic and of igneous rocks. Hutton, on the contrary, from the phenomena of exotic granites, and the analogies observed between these and basalts and modern volcanic rocks, was led to assume an igneous and exotic origin for all save the clearly strati- form crystalline rocks. Metalliferous lodes, also, he sup- posed to have been formed, like granitic veins, by igneous injection from below. While the disciples of Werner denied the igneous origin of basalts, and even of obsidian, Hutton and his school, on the other hand, maintained that the agates often found in erupted rocks were formed by lire. Playfair reasons : — " The fluidity of the agate was therefore simple and unassisted by any menstruum"; that is, it was due to heat, and not to solution ; while, in the case of mineral veins, their closed cavities were held to " afford a demonstration that no chemical solvent was ever included in them." * These cavities were regarded as due to the contraction consequent on the cooling of injected igneous material. i ^ , ' § 9. The basic rocks, included by Hutton under the common names of basalt and whinstone, are regarded by him as similar in origin to granite, and called " unerupted lavas." He elsewhere says that " whinstone is neither of volcanic nor of aqueous, but certainly of igneous origin," that is to say plutonic. Playfair distinguishes between what he calls the volcanic and the plutonic theory of basalt. But while Hutton ascribed a plutonic origin to basalt and to granite, he did not, as some have done, assign a similar plutonic origin to gneiss and other crystalline schists. These were by Werner declared to result from a continuation of the same process which gave rise to granite, and to graduate into it. Gneiss is held both by Wernerians and by modern plutonists to be but a strati- form granite, and both of these rocks are believed by the one school to be aqueous and by the other to be igneous in origin. ♦ Playfair, Illustrations, etc., pp. 79 and 260. v.i THE ORIGIN OP CRTSTALLINB ROCKS. 75 In the system of Hutton, however, a wide distinction is made between the two rocks. Gneiss was no longer a prim- itive or original rock, as taught by Lehman and by Wer- ner, but, like the other crystalline schists, designated by Hutton as Primary, was supposed to be " formed of mate- rials deposited at the bottom of the sea, and collected from the waste of rocks still more ai)cient." In his sys- tem " water is first employed to arrange, and then fire to consolidate, mineralize, and lastly to elevate the strata ; but with respect to the unstratified or crystallized sub- stances the action of fire alone is recognized." * Hutton also conceived the pressure of the waters of a superincum- bent ocean to exert an important influence in the consoli- dation of the sediments. He was thus a plutonist only so far as regards granite and other unstratified rocks, while in maintaining a detrital origin for the crystalline schists he, as Naumann has remarked, may be regarded as the author of the so-called metamorphic hypothesis of their origin. Playfair himself declares of Hutton's system: "We are to consider this theory as hardly less distin- guished from the hypothesis of the vulcanists, in the usual sense of this appellation, than it is from that of the nep- tunists or disciples of Werner." f § 10. It was no part of Hutton's plan to discuss the origin of those more ancient rocks, which had, according to him, furnished oy their disintegration the materials for the primary stratified rocks. It was, in the language of Playfair, a system " where nothing is to be seen beyond the continuation of the present order." " His object was not . . . like that of most other theorists — to explain the first origin of things." This system, as interpreted by his school, asserts the conversion of detrital rocks into masses indistinguishable from those of truly igneous origin, which were the sources of the first detritus. The changes which it assumed to be wrought by the alternate action of water * Playfair, Illustrations, etc., pp. 12, 131. t Biography of Hutton; Playfair' s Works, vol. iv., p. 52. !'J I'i 76 THE ORIGIN OF CRYSTALLINE ROCKS. [V. and fire on the earth's crust were not supposed to be lim- ited by any external conditions in the nature of things, and were compared by Playfair to the self-limited pertur- bations in the movements of the heavenly bodies, in which, as in the geological changes of the earth's crust, " we dis- cern no mark either of the commencement or termination of the present order." ^ 11. Hntton's system is thus concisely resumed by Daubrde : — " The atmosphere is the region in which the rocks decay; their ruins accumulate in the ocean, and are there mineralized and transformed, under the double influ- ence of pressure and the internal heat, into crystalline rocks having the aspect of the older ones. These re-fo'^med rocks are subsequently uplifted by the same internal heat, and destroyed in their turn. The disintegration of one part of the globe thus serves constantly for the reconstruc- tion of other parts, and the continued absorption of the underlying deposits produces incessantly new molten rocks, which may be injected among th j overlying sedi- ments. AVe havs thus a system of destruction and reno- vation of which we can discern neither the beginning nor the end." * § 12. It was this perpetual round of geological changes, which took no account either of a beginning or an end, that led the theologians of his day \,^ oppose" the system of Hutton. On the other hand, in the system of Werner, which taught the fashioning of the present order of our globe from a primeval chaos beneath the waters of a uni- versal ocean, they saw a conformity with the Hebrew cosmogony, which recommended to them the neptunian hypothesis. Hence the theological element which, as is well known, entered so largely'' into the controversies of the vulcanists and the neptunists at the beginning of this century, and the suspicion with which the partisans of Hutton were then regarded by the Christian world. The extreme neptunian views of Werner, however, soon * Daubr^e, Etudes et Experiences, etc., p. 12. v.] THE ORIGIN OF CRYSTALLINE ROCKS. TT fell into disfavor. The visible evidences of the extrusion of trappean rocks in a heated and softened state, observa- tions showing the augmentation of the temperature in mines, and the phenomena of thermal springs and volca- noes, soon turned the scale in favor of Button's views. There were not wanting those who attempted to unite the Wernerian hypothesis with that of an igneous globe, and who supposed a primeval chaotic ocean, to the waters of which, heated by the mass below, and kept at a high boil- ing-point by the pressure of an atmosphere of great den- sity, was ascribed an exalted solvent power. § 13. Such a modified neptunian view was advanced by De la Beche. In his " Researches in Theoretical Geol- ogy," published in 1837, he favored the notion of an unoxi- dized nucleus, as suggested by Davy, and held to a solid crust resting on a liquid interior, and presenting, from the first, irregularities of surface. He then speaks of "the much debated question " whether the crystalline stratified rocks " have resulted from the deposit of abraded portions of pre-existing rocks mechanically suspended in water, or have been chemically derived from an aqueous or an ign'.- ous fluid in which their elements were disseminated." We have in this paragraph three distinct hypotheses pre- sented. Two years later he clearly declared for the sec- ond of them. While admitting the crystallization of detrital matter in proximity to intrusive rocks, De la Beche objected to what he called the "sweeping hypothesis" of Hutton and his school. He supposed that, in the cooling of our planet from an igneous fluid state, " there must have been a time when solid rock was first formed, and also a time when heated fluids rested upon it. The latter would be condi- tions highly favorable to the production of crystalline sub- stances, and the state of the earth's surface would then be so totally different from that which now exists, that min- eral matter, even when abraded from any part of the earth's crust which may have been solid, would be placed 78 THE ORIGIN OF CRYSTALLINE ROCKS. [V. under very different conditions at these different periods." He suggests that there would be " a mass of crystalline rocks produced at first, which, however they may vary in minor points, should still preserve a general character and aspect, the result of the first changes of fluid into solid matter, crystalline and sub-crystalline substances prevail- ing, intermingled with detrital portions of the same sub- stances abraded by the movements of the heated and first- formed aqueous fluids. In the gneiss, mica-slate, chloritic- slate, and other rocks of the same kind, associated together in great masses, and covering large areas in various parts of the world, we seem to have those mineral bodies which were first formed. The theory of a cooling globe, such as our planet, supposes a transition from a state of things highly favorable to the production of crystalline rocks, to one in which masses of these rocks would be more rarely formed. Hence we could never expect to draw fine lines of demarcation between the products of one state of things and those of the other," * § 14. Still later, in 1860, we find a similar view sug- gested by Daubrde as a probable hypothesis. He goes back in imagination to a time when the waters of our planet, as yet uncondensed, surrounded the globe with a dense envelope estimated to possess a weight equal to 250 atmospheres. " The surface of the earth was at this time at a very high temperature, and if silicates then existed they must have been formed without the co-operation of liquid water. Later, however, when it began to assume a liquid state, the water must have reacted upon the pre- existing silicates upon which it reposed, and then have given rise to a whole series of new products. By a veri- table metamorphic action, the water of this primitive ocean, penetrating the igneous masses, caused their primi- tive characters to disappear, and formed, as in our tubes, crystallized minerals from the matters which it was able * De la Beclie, Geology of Cornwall and Devon, pp. 33-34; also Re- searches in Theoretical Geology. T.1 THE ORIGIN OP CRYSTALLINE ROCKS. 79 to dissolve. These matters, formed or suspended in the liquid, would then be precipitated, and give rise to depos- its presenting different characters as the temperature of the liquid diminished." He then inquires, " Were these cUfferent periods of chemical decomposition and recompo- sition, in which aqueous action Qa vote humide') intervenes under extreme conditions which approach those of igneous action (la voie scehe), the era of the formation of granite and of the azoic and crystalline schists? We cannot affirm this in an absolute manner, but we may presume it, especially when we consider that on this hypothesis there must have been formed two principal products, the one massive and the other presenting evidences of sedimenta- tion, passing into each other gradually, as is the case with granite and gneiss. In any case, it cannot be contested that if there was a time when the rocks were exclusively under the dominion of fire, they passed under that of water at an epoch much more remote than we liave hith- erto admitted. The influence, now established, of water in the crystallization of silicates, no longer permits any doubt on this point. We cannot perhaps now find any- where upon the globe rocks of which it may be affirmed with certainty that they have been formed by igneous action, without the intervention of water." * § 15. To give some notion of the temperature of the first water precipitated on the earth's cooling surface, Daubrde calculates that the waters of the present ocean, estimating their mean depth at 3500 metres, would, if spread uniformly over the earth's surface, have a thick- ness of 2563 metres, which, if converted into vapor, would correspond to a pressure of 248 atmospheres, a weight which would be augmented by the presence of other vapors and gases. "No liquid water could there- fore rest upon the earth until its temperature had fallen below that which would give to the vapor of water a ten- sion of 250 atmospheres" at least. When we consider * Daubr^e, Etudes et Experiences Synth^tiques, etc., pp. 121, 122. 80 THE ORIGIN OF CBYSTALLINB ROCKS. 1*^ |i that a tension of only fifty atmospheres of steam corre- sponds, according to Arago and Dulong, to a temperature of 265''.89 centigrade, we can form some conception of the temperature corresponding to a tension five times as great ; which, on this hypothesis, would have been that of tho first waters precipitated on the cooling planet, re- alizing many of the conditions attained by this ingenious experimenter when he subjected mineral silicates to the action of water in tubes, at temperatures of from 400° to 600° centigrade. It is unnecessary to point out that Daubrde here at- tempts to adapt Werner's neptunian hypothesis to that of a once fused and cooling globe, and to find, like De la Beche, in the highly heated primeval ocean, the chaotic liquid which, according to the Uiaster of Freiberg, was the menstruum which at one time held in solution the ele- ments of the primitive rocks. The experiments of Dau- br^e in liis tubes, above referred to, are of great impor- tance in this connection, and will be considered farther on, in the third part of this paper. § 16. The Huttonians early borrowed the notion of a granitic substratum from Werner, and supposed the earth when first cooled to have had a surface of granite. Hutton, true to his thesis, avoided the question of the primal rock. His reasonings, according to Playfair, " leave no doubt that the strata which now compose our continents are all formed from strata more ancient than themselves ; * while, as we have seen, ihe intruded gran- ites were looked upon as but fused and displaced portions of underlying strata. The granitic character of the rocks which antedated aqueous disintegration was, however, a matter of legitimate inference, and his disciple, Maccul- loch, supposed the earth when first cooled to have been " a globe of granite." Later, in 1847, filie de Beaumont, * Playfair's Biography of James Hutton, in Playfair's complete works, 4 vols., Edinburgh, 1822; see vol. iv., pp. 33-81. His lllustnitions of the Iluttouian Theory will there be found reprinted in vol. i. [V. jrre- ,ture m o£ es as that jt, ve- luious ;o the 00° to ▼.] THE ORIGIN OF CRYSTALLINE ROCKS. 81 ere ftt- tliat of De la chaotic rg, was the ele- of Dau- ,t impor- . farther lion of a ihe earth granite, m of the Playfair, mose our . ient than [led gran- i portions the rocks jowever, a 1 Maccul- [lave been ieaumont, '8 complete 1 lUusti-ations il. 1. starting from the hypothesis of a cooling liquiil globe, imagined it " a ball of molten matter, on the surface of which the first granites crystallized." * § 17. It sliould here be mentioned that Poulett Scrope, in 1825, put forth what lie called "A New Theory of tho Earth," in which he supposes " the mass of the globe, or at least its external zone to a considerable depth, to have been originally (that is at or before the moment in which it assumed the position it now holds in the planetary sys- tem) of a granitic composition, composed probably of the ordinary elements of granite, and having a very large grain ; the regular crystallization having been favored by the circumstances under which it pre\ ously took place, though, as to what these circumstances were, I do not venture to hazard a supposition." He farther says, "If then we imagine a general intumescence of an intensely heated bed of granite, forming the original surface of the globe, to have been succeeded by a period in which the predominance was acquired by the repressive force occa- sioned by the condensation of the waters on its surface, and the deposition from them of various arenaceous and sedimental strata (the transition series), the structure of the gneiss-formation is at once simply explained. This structure may have been subsequently increased by the friction of the different laminte against one another as they were urged forward in the direction of their plane surfaces, towards the orifice of protrusion, along the ex- panding granite beneath; the laminse being elongated, and the crystals forced to arrange themselves in the direc- tion of the movement." This implies an exoplutonic ori- gin of gneiss. Later in the same essay, however, Scrope supposes an intensely heated ocean, holding in solution great amounts of silica, and having, at the same time, suspended in its waters, feldspar, quartz, and mica, derived from the disin- * Sur les Emanations Volcaniques et Metallifferes. de Fr. (2) iv. Bull. Soc. Geo!. 82 THE OIIIGIN OF CRYSTALLINE BOCKS. It. tegration of the underlying granite. These suspended materials were deposited and consolidated into gneiss, and later, the dissolved silica precipitating, with some enclosed mica, as the ocean cooled, gave rise to mica-schists. In this last, we see the germ of the therraochaotic hypothe- sis, while in preceding statements of Scrope we have out- lined the volcanic and metamorphic hypothesis of Dana, to bo noticed farther on.* § 18. That such a primitive granite had been the source of gneiss, was taught by Beroldingen, "who main- tained that all the rocks of granitic character having an ap- pearance of stratification, are granites of secondary forma- tion, or regenerated granites, similar in their origin to sandstones"; a notion which was vigorously combated by Saussure,t who held, as we have seen, to the neptunian theory of the origin of these rocks. The detrital hypoth- esis, which he opposed, was however strenuously defended by Ilutton and his school, and especially by Boue and by Lyell. To the former belongs the first definite attempt to explain how uncrystalline sediments like graywacke and clay-slate might be changed into crystalline rocks such as gneiss and mica-schist. Of his views, put forth in 1822 and 1824, Naumann remarks, " Boue first understood how to bring this theory into more decided harmony with the details of geological phenomena, and besides invoking the internal heat, brought to his assistance emanations of gases and vapor from the earth's interior to explain the alteration of sedimentary slates into gneiss and mica- schist." He imagined under these conditions " a sort of * Scrope, Considerations on Volcanoes, etc., 1825, pp. 225-228. Tlie cosmogony of Scrope was fantastic in tlie extreme; he conjectured tlie solid granitic earth to have been detached from the sun as an irregular mass, and compared it to an aerolite. [In rewriting his book on Volca- noes for a new edition, in 1802, Scrope omitted his Theory of the Earth, and did not attempt a cosmogony, but maintained the views already expressed by him as to the granitic nature of the exterior of the primitive earth, which he supposed to be intensely heated, and solid to the centre. (Ed. of 1872, pp. 300, 305. )] t Voyages dans les Alpes (1796), vol. vili., pp. 55, 64. v.] THE ORIGIN OF CRYSTALLINE ROCKS. 88 jn the ( main- ; an ap- forma- igm to (Hibated iptuuian hypoth- leiended j and by •,tempt to [icke and ;s such as . in 1822 itood how Avith the •oking the lations oi :plain the ind mica- u a sort of igneous liquefaction, followed by a cooling process, which permitted a crystalline arrangement and a development of new mineral species without destroying or deranging notably the original laminated structure." * § 19. These views were adopted in 1838, in his "Prin- ciples of Geology," by Lyell, who designated strata sup- posed to have been thus transformed by the name of "hypogene metamorphic rocks"; a title intended to indi- cate a metamorphism which took place in the depths of the earth's crust, and proceeded from below upwards. Under this name, Lyell first popularized the Huttonian view as extended by Boue, which may be conveniently des- ignated as the METAMORPHIC hypothesis of the origin of crystalline rocks. Its plausibility has led to the adoption of this theory by many geologists during the past fifty years. Some, un- willing to admit the influence of a high temperature in such change, have imagined it to result from causes operating at ordinary temperatures during very long j)e- riods. As regards " the nature of these transforming pro- cesses, Gustaf Bischof and Haidinger were inclined to suppose that a long-continued percolation of water through the rocks produced an alteration of their substance and a recrystallization, in the same way as must have taken place in the production of certain pseudombrphs by alter- ation." f Hence the significance of the often repeated dictum that "metamorphism is pseudomorphism on a broad scale." By a further application of the notions derived fiom the study of epigenic or replacement-pseudomorphs, which show in many cases the partial or even the total replace- ment of the original elements of a mineral species, consti- tuting what has been appropriately designated metasoma- • Boue, Annales des Sciences Naturelles, August, 1824, p. 417, cited by Naumann. t Naumanu, Lehrbuch der Geognosie (1857), 2d ed., vol. ii., pp. 160-170. We shall have frequent occasion in these pages to quote from this section of Naumann's Lehrbuch. S4: THE OETGIN OF CRYSTALLINE KOCKS. P. tism, a METASOMATio hypothesis of the origin of crystal- line rocks has been arrived at, to which we shall revert farther on. § 20. Regarding the metamorphic hypothesis, we may remark, as Naumann has done, that the very transforma- tion assumed, namely, that of mechanical sediments into crystalline rocks, remains to be proved. In his "Lehr- buch der Geognosie " in 1857, while still admitting the metamorphic origin of certain limited areas of crystalline schists, Naumann declared that the facts were " not all fa- vorable to the baseless iiypothesis which is now carried to extremes." Such an origin of crystalline rocks was denied by the neptunians, who held to the direct crystallization of these rocks from a chaotic watery liquid, for which reason we may conveniently and appropriately call their view the chaotic hypothesis. It is also denied by those who hold these rocks to be of simple igneous oiigin, the first products of a cooling globe, a view which we may call the ENDOPLUTONIC hypothesis ; and in part by those who advocate what we shall call the exoplutonic or vol- canic hypothesis of their origin. We have already noticed at length the chaotic hypoth- esis, both as originally held by Werner, and modified by intervention of internal heat, as taught by De la Beche and by Daubr^e, constituting what we may call the ther- MOCHAOTic hypothesis. It remains to notice first the two plutonic hypotheses just named, and finally to consider the metasomatic hypothesis, both as applied to rocks con- sisting of crystallir j silicates, and to limestones. § 21. Reasoning, as Naumann has said, from "the great resemblance which gneiss and most of the rocks ac- companying it bear to granite and to other eruptive rocks ; the probability that most of these eruptive rocks have been solidified from a state of igneous fluidity ; the almost unavoidable assumption that our planet was originally in the same state, and was only later covered with a solidi- fied crust; finally the fa "t that in the primitive gneissic irystal- revert ye may sforma- its into "Lehr- ing the jrstalline nt all fa- arried to IS denied allization or which call their lay those nigin, the 1 we may t by those [ic or VOL- VO THE ORIGIN OP CRYSTALLINE ROCKS. 85 series, granite and gneiss are found regularly interstrati- fied with each other," we are led to what we have desig- nated the endoplutonic hypothesis, which is, that the primitive rocks form the "first solidified crust of our planet." Naumann remarks of this, that although it has " not found so many supporters as that of the metamor- phic origin of the primitive rocks, the objections against it are probably neither greater nor more numerous than against the latter." Of this hypothesis, he adds that " it leads necessarily to the inference that the succession of the primitive rooks downward corresponds to their age from oldest to youngest, because it was, of course, through a solidification from without inward that the strata in question were formed." Those who would maintain, on the contrary, that the succession of these in age is from below upward, must suppose, as he explains, that the ma- terial of the younger crystalline rocks "has been pro- truded from the interior, through the earth's crust, in an eruptive form." For these two opposite modes of forma- tion, both essentially plutonic, we may properly adopt the names of ' endoplutonic,' already used above, to designate the hypothesis which supposes the rocks to be generated within tl.cj first-formed crust; and 'exoplutonic' for that which conceives them to have been formed outside of the same crust, by eruptive or what are popularly called vol- canic processes. § 22. The endoplutonic hypothesis has not wanted de- fenders, among whom are some of the most distinguished names of geology, lu 1882, we find Hubert, the emi- nent professor at the Sorbonne, declaring of the ancient crystalline schists : " These mineral masses appear to be due to a crystallization in plc^e, consequent upon the cooling of the fluid terrestrial globe." " The absence from these of rolled masses or of detritus of pre-existing rocks" — assumed by him — " indicates that water did not at that time as yet exist in the state of a liquid mass." This series, including various jji^eisses, micaceous, hornblendic ■'".I ", THE ORIGIN OF CRYSTALLINE ROCKS. [V. and chloritic schists, with crystalliue limestones, " should form a group clearly distinct from all others. It is ante- rior to granite, and constitutes a truly primitive series, which is neither eruptive nor sedimentary, but is due to a third mode of formation, which, borrowing the name from d'Umalius d'Halloy, we may call crystallophylliany * It is difficult to conceive that this can be any other than that imagined by Naumann, which we have called endoplu- tonic. § 23. Thomas Macfarlane, in a learned essay in 1864, on " The Origin of Eruptive and Primary Kocks," f has developed the hypothesis of the endoplutonic origin of the primitive rocks with much ingenuity, and defends a view already suggested by Scheerer, that the laminated struc- ture of these rocks may have been caused by currents in the molten mass of the globe. He further suggests that the first-formed crust may have had a different rate of rotation from the liquid below ; J from which also would result a stratiform arrangement in the elements of the solidifying layer, such as is seen in many slags, and in certain eruptive rocks. But while he applies this view to the primitive rocks, he proposes for the later crystalline schists one which is essentially the thermochaotic hypoth- esis of De la Beche and Daubrde, ascribing their origin to the action of a highly heated primeval ocean on the previ- ously formed crust. The chief difficulties with V'^hich this endoplutonic hypothesis has to contend, according i.j Naumann, "arise from the structural relations of the primitive series, and the mineralogical characters of c^r- * Bull, Soc. Ge'ol. de Francft (3), xi., 30. t Canadian Natuialist, vol. viii. X It is worthy of note in this connection that Halley was long ago led, from the study of terrestrial magnetism, to adopt a similar hypothesis with re^/ard to the earth's interior. "He supposed the existence of two magnetic poles situated in the earth's outer crunt, and two others in an interior mass, separated from the solid envelope by a fluid medium, and revolving by a very small degree slower than the outer crust. The same conclusion was subsequently adopted by Ilansteen." (Hunt, Chem. and Geol. Essays, p. 60.) v.i TFm ORIGIN OF CRYSTALLINE ROCKS. 87 ould aiite- eries, J to a from * It a. that ioplu- 1864, 'I has of the a view 1 struc- ents in 3ts that rate of would , of the and in view to ystalline ! hypoth- origin to he previ- ;li w'hich Drding ■«-J ,s of the rs of cer- tain rocks belonging to it. Whether these difficulties can be explained away by the supposition of a hydro-pyrogen- ous development of the outside of the first solidified crust, as indicated by Angelot, Rozet, Fournet, Scheever, and others, we must leave undecided in the meantime." Such a hyclro-pyrogenous process is more clearly defined by Daubr^e, when he refers the formation of granites and crystalline schists " to aqueous action intervening under extreme conditions which approach igneous action," as explained in § 14. Any modificati ms of the heated crust through the intervention of water must come under the categories of what we have called the thermochaotic and the metasomatic hvpotheses, or else of that one which remains to be described in the present essay. § 24. In the paper already cited, Macfarlane has, more- over, discussed at length the probable condition of the earth's interior, beneath the crust of primitive straiiform rocks, with especial reference to the origin of the different types of eruptive rocks. Already in the last century we find Dolomieu maintaining the existence, beneath the granitic substratum, of a liquid layer fiom which come what he called basaltic lava-flows. A similar view was developed later by Phillips, Durocher, Bunsen, and Streng, who have imagined a separation of the liquid matter at the surface of the cooling globe into two layers, an upper, acidic one, corresponding to granites and tra- chytes, in which, besides alumina and an excess of silica, lime, magnesia, and iron-oxyd are present in very small quantities, and potash and soda abound; and a lower, basic one, corresponding to dolerite and basalt, in which lime, magnesia, and iron-oxyd abound, with an excess of alumina, and but little alkali. These two constitute the trachytic and pyroxenic magmas of Bunsen, who endeav- ored to determine what he conceived to be their normal composition, and, as is well known, sought to show that there exists such a relation between the proportions of these various bases and the silica, that it is possible to ijiimiiSili THE ORIGIN OF CRYSTALLINE ROCKS. [V. y ' ' 'i. 1 calculate the composition of any given eruptive rock from the amount of this element which it contains. He thence concluded that various intermediate rocks have been produced by a mingling or amalgamation, in different proportions, of these two separated magmas. For the composition of these, see farther a note to § 6Q. I have elsewhere discussed the history of this hypothesis, and have given reasons for its rejection.* Sartorius von Waltershausen has also objected, from another point of view, to this hypothesis, and has main- tained that while there is no such distinct separation of the liquid interior as was imagined by Phillips, Durocher, and Bunsen, there is nevertheless a gradual passage down- ward from a lighter, acidic to a denser and more basic liquid stratum ; beneath which still heavier metallic min- erals are supposed by him to be arranged in the order of their respective densities. This view has been adopted and extended by Mr. Macfarlane in his paper above cited. We shall however attempt to show in the second part of this memoir that the observed relations of acidic and basic eruptive rocks admit of a widely different interpretation to those above given, and one more in accordance with known chemical and mineralogical facts.f § 25. Returning from this digression on hypothetical notions of the earth's interior, we propose to consider the exoplutonic or volcanic hypothesis of the origin of the crystalline stratified rocks,- accoi*ding to v/hich, as con- cisely stated by Naumann, the material composing them " has been projected from the interior, through the earth's crust, in an eruptive form." Inasmuch as the matter dis- charged in sub-aerial or submarine eruptions appears in part as flows of molten lava, and in part as disintegrated * On the Probsble Seat of Volcanic Action, Geological Magazine, June, 1860, and Chem. and Geol. Essays, p. 66. t For a discussion of the views of Phillips, Lorocher, Bunsen, and Streng, see Hunt, Chem. and Geol. Essays, pp. 3-6, 66, and 284. See also farther Bunsen, Ann. de Chim. et de Phys., 1853 (3), vol. xxxviii., pp. 215-289. v.] THE ORIGIN OF CRYSTALLINE ROCKS. 89 solid materials which, like other detritus, may be arranged by water, it is evident that this hypothesis connects itself with that of the Huttonian school, to which, considering the mineralogical resemblances between volcanic and other crystalline rocks, it would make little difference whether the sediments required for the metamorphic process came from the disintegration of older crystalline strata, from a primeval granite, or from volcanic products. The vol- canic hypothesis, except so far as consolidated lava-flows are concerned, thus becomes, as we shall see, a metamor- phic or plutonic-detrital hypothesis. As an illustration of this view, we find J. D. Dana in 1843 propounding a general theory of crystalline rocks, which is essentially volcanic. In this he endeavors to show (1) that the schistoiie structure of gneiss and mica- schist is not a satisfactory evidence of sedimentary origin, inasmuch as exotic or eruptive rocks may sometimes take on a laminated arrangement; (2) that granites without any trace of schistose structure may have had a sedimen- tary origin ; and (3) that the heat producing metamorphic changes in sediments did not come from below, as sup- posed by the Huttonians, but through the waters of the ocean, heated by the same eruption which brought to the surface the materials of the metamorphic rocks, which were spread* over the ocean's bottom in a disintegrated form. Their comminution was supposed by Dana to be effected in one of three ways : (1) they were ejected as pyroclastic material, in the form of a sand or ash-eruption, or (2) were disintegrated by coming in contact with water while in a fused condition, or (3) were broken by abrasion after consolidation. In any case, the detrital matter, as in the Huttonian hypothesis, was supposed to be trr isformed into a crystalline rock by the action of heated \»aters. § 26. After assigning such an origin to certain rocks called by him metamorphic porphyries and basalts, with regard to which he supposes " every eruption produced a heated sea around it, which hardened " the disintegrated 90 THE ORIGIN OF CRYSTALLINE ROCKS. [V. porphyry, and recrystallized the comminuted materials, Dana proceeds to say that " granite, like porphyry, is an igneom rock. In its era, granite-sands were formed like porphyry-sands, and restored by heat to metamorphic granite, like metamorphic porphyry. ... I use the word granite here as a general term for this and the associated rocks, mica-slate, syenite, and hornblende-slate, etc., which, I have shown, may also have an igneous origin. These granite-sands, like porphyry-sands, were formed about the regions of eruption, in one of the modes pointed out, and in all probability were never clays like the alluvial deposits of the present day With regard to primary limestones, a general survey of the facts seems to evince that some of these were of igneous origin like granite. If this were the case, there must have been others, formed at the same time with the deposits of granite-sand, and through the action of the same causes. These were re- crystallized by the next discharge of heated waters." * Dana, forgetting the effects of the law of convection ia liquids, here makes the suggestion that "at no great depth the waters might be raised to the heat of ignition before ebullition will begin, and if the leaden waters of a deep ocean . . . are for days in contact with the open fires of submarine volcanoes, we can scarcely fix a limit, to the temperature which they would necessarily receive." We have thus presented a complete exoplutonic or vol- canic hypothesis, and at the same time a complete' meta- morphic or volcanic-detrital hypothesis, alike for porphyry, granite, syenite, gneiss, mica-schist, and crystalline lime- stone ; each and all which are assumed to have a twofold origin, and to appear alike in an eruptive and in a second- ary sedimentary form. A reference to the previous specu- lations of Scrope, already set forth in § 17, will show to what extent Dana was his disciple. * Dana, On the Analogies between the Modem Igneous Rocks and the so-called Primary Formations. Amer. Jour. Science, 1843, vol. xlv., pp. 104-129. v.i THE ORIGIN OF CRYSTALLINE ROOKS. 91 § 27. Dana has since abandoned this hypothesis, so far as regards the eruptive origin of the detrital matters. In his later writings, he sets forth the familiar view of a liquid interior ccvered with a solid crust, which latter was the supposed source of the Archasan or primitive rocks. " These Archaean rocks are the only universal formation ; tliey extend over the whole globe, and were the floor of the ocean, and the material of all the emerged land, when life first began to exist." These rocks of the first crust, disintegrated by submarine and sub-aerial agencies, yielded beds of detritus, which, being consolidated by the action of the heated waters, gave rise to new rocks, which would "be much like those that resulted from the original cool- ing, because chiefly made out of the latter by reconsoli- dation and recrystallization." " Igneous rocks have a close resemblance to granite, diorite, and other crystalline kinds, and hence may have proceeded from the fusion of older kinds. But these older kind? derived their material from an older source, and originally from the fused mate- rial of the globe, so that the proof of such an origin by refusion is not established beyond a doubt." § 28. It is not clear whether, according to Dana, we have anywhere this hypothetical primitive or truly Archaean rock exposed, since, speaking of the Laurentian series, which he also calls Archaean, he says at the same time : — " These Laurentian rocks are made out of the ruins of older Laurentian, or of still older Archaean rocks ; that is to say, the sands, clays, and stones made and distributed by the ocean, as it washed over the earliest-formed crust of the globe. The loose material, transported by the currents and the waves, was piled into layers, as in the following ages, and vast accumulations were formed ; for no one estimates the thickness of the recognized Laurentian beds as below thirty thousand feet." Lest he should be supposed to hold to his former theory of the volcanic origin of these supposed detrital matters, which formed the Laurentian, he now declares, 02 THE ORIGIN OP CRYSTALLINE ROCKS. IV. "They have no resemblance to lavas or igneous ejec- tions." * These crystalline stratified rocks are thus not that universal Archtean terrane which was the first-formed crust of the cooling globe. The imagination is at a loss, however, to understand the nature of the disintegrating process, or the source of the materials wliich in the Lau- rentian period were, according to this hypothesis, spread over vast areas to a depth of not less than thirty thou- sand feet, and seeks in vain for the site of the vanished Atlantis which furnished this enormous amount of me- chanically disintegrated rock. § 29. Clarence King, in 1878, gave us a clear and admirable discussion of the same detrital metamorphic theory, and argued, as Dana had done before him, that the depression of sedimentary strata below the surface of the earth, even to gieat depths, is not sufficient to effect their crystallization; since basal paleozoic beds which have been buried beneath 30,000 feet or more of sedi- ments are now seen, when exposed by great movements of elevation, and by erosion, to present no evidences of crystallization or so-called alteration. King, however, did not reject volcanic action as a source of detritus, for in discussing the origin of the great beds of serpentine and of olivine-rock which are often met with in the older crystalline schi^.ts, he says, " olivine-bearing rocks are among the oldest eruptive bodies," and then asks, " may not olivine-sands, like those now seen on the shores of the Hawaiian Islands, have been then, as now, accumu- lated by the mechanical separation of sea-currents, and subsequently buried by feldspathic and quartz-sands." He thus looks to volcanic eruptions for the source of olivine and serpentine beds, and adds, "I see no reason to ask for a different origin for the magnesian silicates than for the aluminous minerals," f the eruptive source of which is thus implied. A similar hypothesis of the for- * Dana, Manual of Geology, 3rd. ed., 1879, pp. 147, 154, 155, also 720. t Geology of the Fortieth Parallel, vol. i., p. 117. ▼J THE ORIGIN OP CRYSTALLINE ROCKS. 93 mation of beds of olivine-rock and serpentine from accumulations of volcanic olivine-sand, has since been maintained by Julien, whose paper is mentioned further on, § 37. § 30. Other geologists, besides King, have in later times advocated a similar volcanic hypothesis of the origin of crystalline rocks. A. Kopp, in 1872, taught that granite is an altered trach^tic lava, and that gneiss may be derived from the detritus of trachyte or of gran- ite, while doleritic lavas in like manner give rise to the various greenstones. The transformation of these is sup- posed to be effected through the intervention of heated waters, at great depths in the earth.* All this is but a repetition of the hypothesis put forward forty years sinca by Dana, and subsequently abandoned by him. Tornebohm has also lately advanced a similar hypothe- sis to explain the origin of the primitive granite, and of the gneiss into which it seems to graduate. The material of these rocks came up as lava now does, and a portion of it, disintegrated, re-arranged by water and recrystallized, assumed the form of gneiss. Reusch, in like manner, ac- cording to Marr, supposes that the gabbros, diorites, and dioritic and hornblendic schists of the Bergen district, in Norway, are but altered tufas and erupted rocks. § 31. Mr. Marr, in a recent paper, urges the claims of the volcanic hypothesis to explain the origin of the ancient crystalline rocks, seemingly unaware of its earlier advocates. It is apparent that if we accept the doctrine of the permanence of continents and of oceanic depres- sions, the metamorphic-detrital theory of the Huttonians, which builds up series of crystalline rocks beneath the sea from the ruins of an older land, which had itself been formed beneath the sea, is no longer tenable. The diffi- culty of getting the thirty thousand feet of sediments required to spread over a continent, as in Dana's later hypothesis, is, as Marr perceives, overcome if we suppose « Neues Jahrbuch fiir Mineralogie, 1872, pp. 388 and 490. 11 |: I.I I \h U.I m '.1.1 f 1*1 If; K^ 94 THE ORIGIN OF CRYSTALLINB ROCKS. w this material to have been derived, not by the superficial waste and disintegration of former land, but by ejection from reservoirs beneath the earth's crust. Hence, with the advocates of the doctrine of the permanence of conti- nents, the volcanic or exoplutonic hypothesis is again com- ing into favor.* Similar considerations appear to have led C. H. Hitch- cock, in 1883, to a like conclusion. The' continents, in his scheme, are built up from beneath the waters of a univer- sal ocean. He writes : — " We start with the earth in the condition of igneous fluidity. It cools so as to become encrusted and covered with an ocean. Numerous volca- noes discharge molten rock, building up ovoidal piles of granite [beneath the ocean], which change gradually into crystalline schists. When the hills are high enough to overlook the water, they constitute the beginnings of dry land." This is intelligible, but it seems strange to one familiar with the geological literature of the last forty years to read, in this connection, the remark of Hitchcock that few "have ventured to spp"k of anything like vol- canic action, except as it has been manifested in the for- mation of dikes, in the early periods." f To all of these speculations as to the exoplutonic or volcanic origin of the crystalline rocks, the language of Naumann, in criticising the original volcanic hypothesis of Dana, is applicable. "The perfect and thoroughly crystalline character of the gneiss, the enormous extent which the primitive formations occupy in so many dis- tricts, the architecture of these great gneissic regions, and their occurrence wholly independent of larger granitic masses, are all incompatible with this idea." § 32. The view of the igneous and eruptive origin of crystalline limestone, admitted in Dana's former scheme, was familiar to the geologists of forty years since. Em- * Marr, The Origin of Archisan Rocks; Geological Magazine, June, 1883. t Hitchcock, The Early History of the North American Continent. — Proc. Amer. Assoc. Adv. Science, 1883. v.i THE ORIGIN OP CRYSTALLINE ROCKS. u mons and Mather in America, and Von Leoiihard, Rozet, and Savi in Europe, among others, then lieUl to the belief that many crystalline limestones were igneous, and Savi had even attempted to point out the centres of eruption of the Carrara marbles.* It is hardly necessary to recall the fact that serpentines, and great deposits of magnetite and specular iron, are still by some authorities considered as eruptive rocks, and that the hypothesis of the igneous origin of metalliferous lodes, taught by Hutton, is not yet wholly obsolete. In 1858, H. D. Rogers wrote of "the great dikes and veins of auriferous quartz " supposed to have issued " in a melted condition, through rents and fis- sures in the earth's crust. Outgushing bodies of this quartz," chilled by contact with the cold waters of the ocean, were supposed by him to have furnished the mate- rial for the Primal quartzites of Pennsylvania.f Still later, in 1874, we find Belt raaintahiing with learned in- genuity the igneous origin and the injection of auriferous quartz veins. He insists, as I have elsewhere done,| on the transition from veins of quartz, often metalliferous, to others containing feldspar, and thence to true granitic veins ; but instead of regarding these as aqueous and con- cretionary, assumes them to be igneous, and thence con- cludes that the gold-bearing quartz lodes were filled with liquid quartz by " igneous injection," though admitting that in these, as in granites, water helped to impart liquidity. § § 33. In farther illustration of the extension of the plutonic doctrine to other rock-masses than those already mentioned, I quote from an essay by Daubr^e, published * See for references, Hunt, Chem. and Geol. Essays, p. 218; also Boue, Guide du G^ologue Voyageur, ii., 108. t Geology of Pennsylvania, ii., 780. t Chemical and Geological Essays, pp. 192-208, and infra § 58. § Belt, The Naturalist in Nicaragua, 1874, pp. 97-100. In the pages here referred to, my friend, whose premature death was a great loss to science, has set forth with clearness the Huttonian theory of metallifer- ous veins. THE OlllGIN OF CRYSTALLINE UOCICS. IV. in 1871.* "The hypothesis advanced by Lazzaro Moro, in 1740, attributing an eruptive origin to rock-salt, as well as to uulpluir and bitumen, was again tal^on up and applied by De Cliarp'intier (1823) to the salt-mass at Hex, whicli is associated with anhydrite ; and D'Alburti, in the chissic study made by him of this terrano, maintained the same hypothesis for all the rock-salt found in the trias. More- over, the examination of the deposits of pisolitic iron-ore had, in 1828, conducted Alexandre Brongniart to a similar conclusion, which was soon after applied to the siliceous deposits which constitute the buhrstone of the tertiary. A like origin was by D'Omalius (1841 and 1855) ascribed to other substances, particularly to certain clays and to certain sands, which, especially in Belgium, appear to be connected with the formation of calamine, and which Du- mont in 1854 called geys^rian deposits." " It was thus," adds Daubr . mi ill 'if S 'IS ■11 112 THE ORIGIN OF CRYSTALLINE KOCKS. tv. M';!,-! I !'! II. — THE DEVELOPlSrENT OP A NEW HYPOTHESIS. § 47. The history of the beginning and the growth of the new hypothesis here proposed to explain the origin of crystalline rocks is necessarily to a great extent personal, since it covers the work of many years of the author's life. The lines of investigation which have led to this hypothesis may be described as, first, that of the order and succession of the crystalline stratified rocks of the earth's crust; secondly, their mineralogy and lithology; thirdly, their history, considered in the light of phj'sics and chem- istry, involving an inquiry into all the chemical relations of existing rocks, waters, and gases, including the trans- formations and decay of rocks, and the artificial production of mineral species ; and fourth and lastly, the probable condition of our planet before the creation of the present order. The adequate discussion of all these themes, which would include a complete system of mineral physiology, is impossible within the limits of the present essay, but a brief outline of some of the chief points necessary to the understanding of the hypothesis will here be attempted. § 48. As regards the order and succession of the crys- talline rocks, the author's studies of them, begun in New England forty years since, and continued in Canada from 1847 onwards, were for many years perplexed with the difficulties of the Huttonian tradition (then and for many years generally accepted in America), that the mineral character of these rocks was in no obvious way related to their age and geological sequence, but that the strata of paleozoic and even of cenozoic times might take on the forms of the so-called azoic rocks. It was questioned by the partisans of the Huttonian school whether to the south and east of the azoic rocks of the Laurentides and the Adirondacks, in North America, there were any crys- talline strata which were not of paleozoic or of mesozoic age, although many of these are undistinguishable from the rocks of the Laurentides. %1 THE ORIGIN OF CRYSTALLINE ROCIiS. 113 As I have elsewhere said, tlie metamorphic and the metasomatic, not less than the exoplutonic hypothesis of the origin of the crystalline rocks, by failing to recognize the existence and the necessity of an orderly lithological development in time, have powerfully contributed to dis- courage intelligent geognostical study, and have directed attention rather to details of lithology and of mineralogy, often of secondary importance.* That a great law pre- sided over the development of the crystalline rocks, was from the first my conviction; but until the confusion which a belief in the miracles of metamorphism, metaso- matism, and vulcanism had introduced into geology was dispelled, the discovery of such a law was impossible. § 49. Convinced of the essential trutli of the princi- ples laid down by Werner, and embodied in his dis- tinctions of Primitive, Transition, and Secondary rocks, I sought, during many years, to define and classify the rocks of the first two of these classes, and by extended studies in Europe, as well as in North America, succeeded in establishing an order, a succession, and a nomencla- ture, which are now beginning to find recognition on both continents, f While the succession of the various groups of crystal- line rocks was thus being established, not without the efficient aid and co-operatiou of other workers in late years, mineralogical and chemical studies were teaching us much of the true nature of the differences and resem- 1 * Amer. Jour. Science, 1880, xix., 298. t I have elsewhere given tlie history of the progress of inquiry in tliis direction in Report E of tlie Second Geological Survey of Pennsylvania (Azoic Rocks) 1878; in brief, in an essay on Pre-Cambrian Rocks, etc., in the Amer. Jour. Science, 1880 (xiv., 2G8); and later in a study of the Pre-Cambrian Rocks of the Alps, in the Trans. Roy. Soc. Canada, pout, Essays X and XI. See also in this connection the late address of Dr. Hicks, President of the British Geologists' Association, in its Proceed- ings, vol. viil., 1883, On the Succession of the Archsean Rocks, etc.; and the still more recent paper of Prof. Bonney, President of the Geological Society of London, on The Building of the Alps, in Nature iov May 18 and 25, 1884; also the Geological Magazine for June, 1884, p. '280. i 114 ^n^ OBIGIN OF CRYSTALLINE BOCKS. IT. ill a.»— . «.ll as of the natural relations Uancesot these grours.;.s wen as ^.^.^^^^^ ^^^^ and modes of «""!^*™4 -^.^ the composition of the mineral speeies ^^!"»|^^f ! ^^.^ tig.tions of physiosts and crystalline rocks. !'« '"„ fo^m and consistence to Jronomers had, ^^T^^^^ origin of our planet and the ancient theory o the >gne b ^._^^^ ^^ j^^^^^j^g^. the eonem'rent workmg J ■ the way tor ^ new tion above indicated wa» «" » l^^Pii^^^'^eks - a hypothe- '■>Tf*whioif l'^^^^^^^^^^^ "'^^'■■'"""" *T^"":s in January «58 - ac;nt«rysince,thatlventar dtopu ^^^^ ^^^^^^ ^ to the «l-"f' y °' *,^°t vii which geognosy makes Considering only that « "« j ^^at at a very early „, accinainted It was — .,^ ^„,„,„t, were um ed period tlie whole of '» ""^J'^tvich included the metalUo i„ a fused mass o£ s'>"'»,*.^,"' J" , ;„ the ocean's waters; bases of the salts ^^^f'ffXtUme was charged with „hile the dense ''f '^P^'^^X ne, combined with oxy- all the carbon, f P'^«' ^^y *,vhich were present watery gen or with ^^y^^o^'^^'^ZhMe excess of oxygen. The vapor, nitrogen, and a P *^» " j,.„„ tfe atmosphere, Jt precipitated and ^^^ ^l^^^^ ,,„,t, would, it was falling on the hot ^-^^^J\ a,, protoxyd bases giv- «aid, soon become "«"f ^ J^V,^ates of the primeval sea i„g rise to the chlonds "d sulp ^^i„ed silica, at wife the probable sep— of tj^ ^^ .,„ „g. ■ that high t«"r ; 'a are of the primitive atmosphere, gestion as to the ae.d "^'"^ °\,u,h were obvious deduo- Snd its first «l'«™''=f Jf 'X; y, W. ^» I ^''^"™'^ tions from the ^g"''"™ * Q^nstedt. « ^ ^ #EpocbeuderNatur,p.20. v.] THE ORIGIN OF CRYSTALLINE ROCKS. 115 William Hopkins. The subsequent sub-aerial decay of exposed portions of the earth's primitive crust in a moist atmosphere, now purged of the acid compounds of chlo- rine and sulphur, but still holding carbonic acid, was tlien set forth as resulting in the transformation of feldspathic silicates into clays, and the transference to the sea of the lime, magnesia, and alkalies of the decayed roc^'. in the form of carbonates, the latter of wdiich, reacting on cal- cium-chlorid, would 3deld carbonate of lime and chlorids of sodium and magnesium. It was then said that by this hypothesis " we obtain a notion of the processes by which, from a primitive fused mass, may be generated the various silicious, argillaceous, and calcareous rocks which make up the greater part of the earth's crust." Of this it was declared, " the earth's solid crust of anhydrous and primi- tive igneous rock is everywhere deeply concealed beneath its own ruins, which form a great mass of sedimentary strata, permeated by water," and subjected to heat from below, changing them to crystalline metamorpliic rocks, and at length reducing them to a state of igneo-aqueous fusion, th^'ough which they yield eruptive rocks. Of this primitive rust it was farther asserted that it " probably approached dolerite in composition." The principal points in this hypothesis, as presented in 1858, were thus the solid condition of the earth's interior, and the derivation of the whole of the rocks of the known crust, by chemical transformations, from the origi- nal superficial and last-congealed layer of the cooling globe, which was considered to have been a basic rock, not unlike dolerite. All of these positions are fundamen- tal to the present hypothesis. § 52. These views Avere again repeated in a paper read before the G ological Society of London in June, 1859, with some farther developments as to the origin of the various crystalline rocks derived from the primeval crust. This, it was claimed, was necessarily quartzless, and far removed in composition from the supposed granitic sub- M ^ If) ,HE OKlom O^ CMST^LUSr, BOCICS. IIQ THE OBIGI-N ui. All attempt was, liow- stratum, or tl>^P"'";«;\S»^ fte quart;, derived from the ever, made to show that ""^ "' '1 , primitive igneous slpilosed first d^^^^P^t " td mente resulting from ■oei. by acid «»*<=«' "J"^ *d sub-aevial decay, coarse' subsequent di^'»*fS'"*l\Xs permeable, would result, and finer sediments, moie « '''^^ | „f infiltrating waters Uich by the natural o^-^-^lJ^vs, divide themselves „iglrt, in »c«»dance vv th know ^^^^^.^^^ ^^ ,^ into two great c asse • «- « , ,^„,, „£ potash, and by g,eat classes, "tire <»"; ^^f 'otash, and by eess of slica, ^^X t^^^P^^t: • -^^°'^ ""' T counts of Inne, magn ^^,^;j^ ;„ the other m 11 amounts of -• f 8' ^ e^ wbUe » the other ,,„ted by the gramtes a.^ t^arf'jt . _^_^^^ ^^^_^^ .. ^^^^^ ^^^ silica and potash are less abun , ^_^^^ ; magnesia prevail, gmng "f *"„] ^ii.piace' aent of sueh feldspars. The ™^*-»"^"~^p,ain the origin of the sediments -^^Y th- «naUe^: ocks without calling to our geneo«sundifEerent,ated rus,w ho ^^^^^.^^^ ^^ ^^^ ^^„ llutonie matters , "^ *';yf;,ystalline rocks; gnersses, great types of acrd.c and basic cy ^^^^^ ^^^ ^ Iranites, and trachytes on the ^^^^^,, ,j^g,,aed focks, greenstones, and ^^sato °n ^^^ ,,,,;, to the as an attempt to ''dapt *hoJ:lu ^ ^^^, ,t • growing demands of *"" „t in time, and a po^r- lad hitherto lacked a ^^rtrng P" ^_^^ ^^^.^ fe, ble explanation of the W" t^f ^ ^j^^ j^i^tory of geology, tuis scheme d-»d- f J^ts author, it must share the although, m the 3uag j^^^_ . for the references to this early «^a»t the f ^ ^^^^^^^^^ ^^ Geological Essays, pp. l-l"^- .v.i THE ORIGIN OF CRYSTALLINE ROCKS. 117 fate of all other forms of the metamorphic hypothesis. In recognizinj^ the adequacy of a primitive undifferen- tiated layer of igneous rock as the source of the materials of the future order it, however, effected a great stej) to- wartUi a more satisfactory hypothesis. § 54. The nature and history of this primitive layer were farther discussed by the author in a lecture on " The Chemistry of the Primeval '""-arth," given at the Royal Institution in London, in June, 1867.* Therein it was said : " It is with the superficial portions of the fused mineral mass of the globe that we have now to do, since there is no good reason for supposing that the deeply seated portions have intervened in any direct manner in the production of the rocks which form the superiicial crust. This, at the time of its first solidification, pre- sented probably an irregular diversified surface, from the result of contraction of the congealing mass, v/hich at last formed a liquid bath of no great depth, surrounding the solid nucleus." It was further insisted that this mate- rial would contain all of the bases in the form of silicates, and must have much resembled in composition certain furnace-slags or volcanic products. Of this primary lava- like rock, it was said, that it is now everywhere concealed, and is not to be confounded with the granitic substratum. That granite was a secondary rock, formed through the intervention of water, waa then argued from the presence therein, as a constituent element, of quartz, " which, so far as we know, can only be generated by aqueous agen- cies, and at comparatively low temperatures." The meta- morphic hypothesis of the origin of granite was then maintained. In 1 869, in an essay on " The Probable Seat of Volcanic Action," t a further inquiry was made into the probable * Proceedings of the Royal Institution, and also Chemical and Geo- logical Essays, pp. .35-45. t Geological Magazine for June, 1809, and Amer. Jour. Science, for July, 1870 (vol. i., p. 21). See also Chemical and Geological Essays, pp. 59-67. >4':\ I--- I rnii I' 111;-! i!l|i| i II 118 THE ORIGIN or CRYSTALLINE ROCKS. nr. nature "nd condition of what had been spoken of in 1858 as "the ruins of the crust of anhydrods and primitive igneous rock." This, it was now said, " must by 'contrac- tion in cooling have become porous and permeable, for a considerable depth, to the waters afterwards precipitated upon its surface. In this way it was prepared alike for mechanical disintegration and for the chemical action of the acids . . . present in tlie air and the waters of the time. . . . The e?.rth, air, and water, thus made to react upon each other, constitute the first matters, from which, by mechanical and chemical transformations, the whole mineral world known to us has been produced." It was farther argued, from many geological phenomena, that we have evidence of the existence between the solid nucleus and the stratified rocks of "an interposed layer of par- tially fluid matter, which is net, however, a still unsolidi- fied portion of the once liquid globe, but consists of the outer part of the congealed primitive mass, disintegrated and modified by chemical and mechanical agencies, im- pregnated with water, and in a state of igneo-aqueous fusion." * * Prestwich, in a memoir presented to the Royal Society of London, April 16, 1885, of wliicli an abstract appears in Nature for April 23, after considering (1) the flexibility of the earth's crust as shown In fold- ings and corrugations, and in secular depressions and e;3vations of con- tinental areas; (2) the increase of temperature in the depths, and (3) the volcanic phenomena of the present day, and the outpouring of vast sheets of trappean rocks during late geological periods, and after discuss- ing the bearing of these upon various of^ ,r geological hypotheses, enounces a similar view to that set forth above, and farther, in § 127. He concludes that all these phenomena "are most compatible with the movement of a thin crust on a slowly yielding viscid body or layer, also of no great thickness, and wrapping around a solid nucleus. The viscid magma is thus compressed between the two solids, and while yielding in places to compression, it, as a consequence of its narrow limits, expands in like proportion In conterminous areas." It would be difficult to ex- vess more concisely or more correctly the view of the earth's interior already set forth by the author in 1860, in his discussion of " The Prob- able Seat of Volcanic Action." The intervention of water in this primary plutonic magma, which Prestwich appears to reject, is, in the writer's opinion, inevitable. ' v.] THE CKENITIC HYPOTHESIS. 119 § 55. Although in 1858 I had, as ah-eady shown, sought to give a more rational basis to the metamorpliic hypothesis of the origin of crystalline rocks, the tradi- tions of which, as expounded by Lyell, weighed so heavily on tlie geologists of the time, other considerations soon afterwards led me to seek in another direction for the solution of the problem. The examination of the mineral silicates deposited during the evaporation of many natural waters, that of the Ottawa river among others, and the study which I had made of the hydrous magnesian silicate found in the tertiary strata of the Paris basiii, induced me, as early as 1860, to inquire "to wiiat extent rocks composed of calcareous and magnesian silicates may be directly formed in the moist way " ; and again, in the same year, to declare with regard to the latter, " it is evident that such silicates could be formed in basins at the earth's surface, by reactions between magnesian solutions and dissolved silica " ; a consideration which was then applic.d to the generation of serpentine and of talc. Again in 1863 and 1864, I ventured to conclude that " steatite, ser- pentine, pyroxene, hornblende, and, in many cases, garnet, epidote, and other silicated minerals, are formed by a crystallization or molecular re-arrangement of silicates generated by chemical processes in waters at tlie earth's surface." * § 56. While natural waters hold in abundance both lime and magnesia, alumina is, under ordinary conditions, insoluble in them, and, moreover, is not found vuicom- bined with silica. The problem of the genesis of the alu- minous double silicates, so abundant in the rocks, was therefore a more difficult one than that of tlie simple prot- oxyd-silicates, with which they are often intimately asso- ciated. Many facts In the history of r.eolitic minerals, however, soon led me to recognize in the conditions under which these aluminous double silicates are formed, a clew * For citations and references see Chemical and Geological Essays, pp. 206, 297, and 300. 120 THE ORIGIN OP CHYSTALLINE ROCKS. m f '.;! «|ll|!i to the 8L,lution of the problem. Thus it was that, in an essay read before tlie Geological Society of Dublin, in April, 18G3,* I called attention to the observations of Daubrd'e on the production, during the historic period, of the zeolites, chabazite and harmotome (phillipsite), by the action of thermal waters at a temperature not above 70° C, on the masonry of the ancient Roman baths at Plombieres. The mode of the occurrence of these miner- als showed that the aluminous silicate of the burned bricks had been changed into a temporarily soluble com- pound, which had crystallized in cavities as zeolites, — species which differ in composition from feldspars only by the presence of combined water. I also called attention, in this connection, to the experiments of Daubrde, who, by operating at higher temperatures in sealed tubes, had succeeded in producing crj-stallized quartz, pyroxene, and, apparently, feldspathic and micaceous minerals. § 57. The aqueous origin of feldspars, and their inti- mate relations to zeolites and other hydrous minerals, were farther noticed by the author, in the "Geology of Canada," in 1863, in which he cited the observations made by J. D. Whitney on the frequent occurrence of orthoclase in the copper-bearing veins in the melaphyres of Lake Superior. The crystals of this mineral, which had been mistaken for stilbite, are there found under con- ditions which show their formation contemporaneously with the zeolites, analcime and natrolite ; while elsewhere in the same region, the associates of the orthoclase are epidote, calcite, native copper, and quartz, upon which, as well as upon saponite, the crystals of the feldspar were found implanted.f Whitney recalled in this connection the occurrence of a variety of orthoclase, the weissigite of Jenzsch, with chalcedony, in cavities of an amygdaloid. [Mr. George F. Kunz has since discovered orthoclase in * The Chemistry of Metaraorphic Rocks; DubUn Quarterly Journal for July, 1863; reprinted in Chemical and Geological Essays, pp. 18-34. t Wliitney, Amer. Jour. Science, 1869, vol. zxviii., p. 16. ; 'ii ; v.] THE CRENITIC HYPOTHESIS. 121 the mesozoic diabase of New Jersey. Tlie specimei. ately shown to the New York Academy of Sciences are de- scribed by him as " compact veins and crystals of tlesh-red primitive orthoclase, formed directly on the diabase," and sometimes in a granular form making up the chief part of veins traversing this rock. "The veins are usually per- pendicular, running east and wost, varying in thickness from half an inch to four inches, and were evidently formed by deposition directly on the walls of diabase. On each side of the orthoclase, milky quartz, either massive or at times in imperfect crystals, is implanted. On the orthoclase and quartz alike, calcite, massive and crystal- lized, and also apophyllite, datolite, pectolite, and the zeolitic minerals are often deposited. Scattered through the orthoclase and quartz are found pyrite, chalcopyrite, and occasionally galenite — all of these in perfect isolated crystals. The veins are frequently made up entirely of quartz, both massive and crystalline, neither calcite nor any zeolitic minerals having been deposited upon them. The zeolitic minerals are usually deposited directly upon the diabase." The careful observations by Mr. Kunz of these veins, which according to him are frequently found in the excavations in Bergen Hill, at Weehawken, throw much light on the relations of the zeolites to feldspathic and granitic aggregates.] [Garnet is found in similar associations, Mr. Charles Robb, having, in 1882,* noticed its occurrence in a vein in the diabase of St. Ignace Island, Lake Superior, implanted in prehnite, with laumontite, quartz, calcite, barite, mag- netite, native silver, copper-glance, and a chloritic matter. A specimen of this received from him shows dodecahedral garnets, reddish brown in color, f two or three millimetres in diameter, and small octohedrons of magnetite on prehn- ite with laumontite.] § 58. The facts noted by Whitnej'^ were insisted upoii, * Communication to the New York Academy of Sciences, May 25, 1885. t Cauadiau Naturalist, x., 176. 122 \. THE ORIGIN OF CRYSTALLINE ROCKS. [V. i v^ in connection with my own observations, to prove the aqueous origin of the feldspar found in veins among crys- talline schists in the province of Quebec, wiiero "a llesh- red orthoclase occuis so intermingled with white quartz and chlorite as to show the contemporaneous formation of the three species. The orthoclase generally predominates, often reposing upon or surrounded by chlorite, and at other times imbedded in quartz, which covers the latter. Drusy cavities are also lined with small crystals of the feldspar, and have been subsequently filled up by a cleav- able bitter-spar," often with crystaUized hematite, rutile, and copper-sulphids. It was shown that among these veins, then described as of aqueous origin, there was to be seen a transition, from those " containing only quartz and bitter-spar, with a little chlorite or talc, through others in which orthoclase appears, and gradually predominates, until we arrive at veins made up of quartz and feldspar, someiimes including mica, and having the character of a coarse-grained granite ; the occasional presence of copper- sulphids and hematite characterizing all of them alike." There was also described the occurrence, in the same region, of a dark-colored argillaceous and schistose rock, having in parts the aspect of a chloritic greenstone, which is rendered amygdaloidal by the presence of numerous spherical or ovoidal masses of quartz, or more commonly of reddish orthoclase, often with a nucleus of quartz. In schistose varieties of this rock the feldspar extends from these centres in such a manner as to give a gneissoid aspect to the mass. All of these facts were regarded as showing the aqueous origin of orthoclase, and its secretion from the adjacent rock.* § 59. With the feldspar in the above-mentioned veins may be compared the similar occurrence, observed in 1872, in the great quartz lodes with chalcopyrite which traverse the Huronian greenstones at the Bruce Mines, on Lake Huron, of bands one or two inches wide of a brick-red * # Geology of Canada, 1863; pp. 470 and 606. V v.] THE CllENlTIC HYPOTHESIS. 128 ortlioclase, mingletl with a little quartz and a small ainouiit of a greenish, apparently hornblendic element, furming an aggregate which can hardly be distinguished from some of the older granitic rocks, but is clearly inter- banded with the metalliferous quartz and the bitter-spar of the lode. In this connection may also be quoted a description of the vertical parallel veins found cutting at right angles the Montalbau gneisses hi Northbridge, near Worcester, Massachusetts. These veins, as described by the writer, " may be traced iov considerable distances, and are ordinarily but a few inches in thickness. The vein- stone of these is generally a vitreous quartz, which in some parts exhibits selvages and in others bands of white orthoclase, by an admixture of which it passes elsewhere into a well characterized granitic vein. The quartz veins, in places, hold cubic crystals of pyrite, together with chal- copyrite and pyrrhotite, the latter in considerable masses, sometimes accompanied by crystals of greenish epidote imbedded in the quartz, and occasionally associated with red garnet. In one part there is found enclosed in the. wider portion of a vein, between bands of vitreous quartz, a lenticular mass three inches thick, of coarsely granular pink calcite, with imbedded grains of dark green amplii- bole, and on one side small crystals of olive-green epidote and red garnet ; the whole mass closely resembling some crystalline limestones from the Laurentian," and evidently endogenous.* I have also described remarkable examples of similar associations of zoisite, garnet, hornblende, pyr- oxene, and calcite in the metalliferous quartz-lodes in the Montalban series, at Ducktown, Tennessee, f § 60. The question of the aqueous origin of concretion- ary veins was resumed by the author in 1871, in an essay On Granites and Granitic Veinstones, when it was main- tained that the relation of granitic veins with metalliferous * Azoic Fiocks, Report E, Second Geological Survey of Pennsylvania, p. 247. t Chemical and Geological Essays, p. 217. ■I I I I 1'! c I II •' lli! r 124 THE ORIGIN OF CRYSTALLIITE ROCKS. quartz lodes, on the one hand, and with calcareous veins carrying the ordinary minerals of crystalline limestones, on the other, is such that to all these veins must be assigned a common aqueous origin. It was farther shown that the endogenous granitic masses or veinstones in the Montalban or younger gneissic series in New Eng- land often attain breadths of sixty feet or mcne, and that they present great varieties in texture, from coarse aggregates of banded orthoclase and quartz, often with muscovite (from which these various elements are mined for commercial purposes), to veins in which the concre- tionary character is not less marked, including beryl, tour- maline, garnet, cassiterite, and other rare minerals ; while others still of these great veins are so fine-grained and homogeneous in character as to have been quarried as granites for architectural uses. These endogenous masses are included alike in the gneisses, the quartzites, the stau- rolitic mica-schists, ami the indigenous crystalline lime- stones of the Montalban series, and, though generally transverse, are sometimes, for a portion of their course, coincident with the bedding of the enclosing rock.* It was clear that these endogenous granitic veins of posterior origin were mineralogically very similar to the older gneisses and the erupted granites. From a pro- longed study of all these phenomena, the conclusion was then reached that we have in the action which gen-- erated these endogenous granitic rocks a coniinuation of the same process wliich gave rise to the older or funda- mental granitoid gneisses, which were hence of aqueous origin. § 61. This process of reasoning was in fact identical with that by which Werner, in the last century, was led to as.sign an aqueous origin to the primitive granite and the crystalline schists. In a description, in 1874, of some examples of these banded veinstones from Maine and Nova • Amer. Jour. Science (3), vol. i., pp. 88 aud 182, and vol. ill., p. 115; also Cheni. and Geol. Essays, pp. 183-209. ▼4 THE CUENITIC HYPOTHESIS. 125 Scotia, it wan said that their structure is "tluo to suc- cessive (lepof.its from water of erystalliiio matter on the walls of the /eiu, and results from a process which, though operating in later times and in subterranean fissures, was probably not \ery much unlike that which gave rise to the indigenous granitic gneisses." * The same ideas as to the origin of the ancient crystalline rocks, and their relations to granitic and to zeolitic veins, were still farther defined by me, in 1874, when it was said : ''The deposition of im- mense (quantities alike of orthoclase, albite, ind oligoclase in veins which are evidently of aqueous origin shows that conditions have existed in which the elements of these mineral species were abundant in solution. The relation between these endogenous dei'osits and the great beds of ortlioclase and triclinic feldspar-' ucks is similar to that between veins of calcite and of (|uartz, and beds of mar- ble and of travertine, of quartzite and of hornstone. But while the conditions in which these latter mineral species are deposited from solution have been perpetuated to our own time, those of the deposition of feldspars and many other species, whether in veins or in beds, appear to belong only to remote geological ages, and, at best, are represented in more recent times only by the production of a few zeolitic minerals." f § 62. A farther and more particularized statement of the author's conclusions as to the origin of the crystallina rocks was embodied in a paper read before the American Association for the Advancement of Science at Saratoga, in August, 1879, containing the three following proposi- tions :$ — "1. All gneisses, petrosilexes, hornblendic and mica- ceous scliists, olivines, serpentines, and, in short, .all sili- cated crystalline stratified rocks, are of neptunian origin, * Proc. Boston Society of Natural History, xvi., 237, p. 198. t Cliemical and Geological Essays, p. 298. t The History of Some Pre-Cambrian Kocks, etc. Proc. A. A. A. S., for 1879, and Amer. Jour. Science (1880), six., p. 270. Also farther on, Essay VUI. ,(1 ^ a J>^ --ir-vja«^ua-«K;»tai.j 126 THE ORIGIN OP CRYSTALLINE ROCKS. PR s and are not primarily due to metamorphosis or to meta- somatosis, either of ordinary aqueous sediments or of vol- canic materials. " 2. The chemical and mechanical conditions under which these rocks were deposited and crystallized, whether in shallow waters or in abyssal depths (where pressure greatly influences chemical affinities), have not been re- produced to any great extent since the beginning of paleo- zoic time. " 3. The eruptive rocks, or at least a large portion of them, are softened ani^ displaced portions of these ancient neptunian rocks, of which they retain many of the min- eralogical and lithological characters." § 63. In a subsequent paper, in 1880, it was said, with reference to the sub-aerial decay of rocks : " The alumi- nous silicates in the oldest crystalline rocks occur in the forms of feldspars, and related species, and are, so to speak, saturated with alkalies or with lime. It is only in' more recent formations that we find aluminous silicates either free or with reduced amounts of alkali, as in the argillites and clays, in micaceous minerals like muscovite, margarodite, damourite, and pyrophyllite, and in kyanite, fibrolite, and andalusite ; all of which we regard as derived indirectly from the more ancient feldspars." In connec- tion with this important point, which I had already dis- cussed elsewhere, I added the following note, referring at the same time to the propositions of the ^receding para- graph : * " It is a question how far the origin of such crystalline aluminous silicates as muscovite, margarodite, damourite, pyrophyllite, kyanite, fibrolite, and andalusite, is to be sought in a process of diagenesis in ordinarj'- aqueous sediments holding the ruins of more or less com- pletely decayed feldspars. Other aluminous rock-forming * The Chemical and Geological Relations of the Atmosphere, ante, page 37. See farther, for the stratigraphical relations of the various aluminous silicates (which were first set forth by the author in 1863), Chem. andGeol. Essays, pp. 27 and 28; also Report E, Second Geological Survey of Pennsylvania (1878), p. 210. ;l' •■' v.] THE CRENITIC HYPOTHESIS. 127 silicates, such as chlorites and magnesian micas, are, how- ever, connected, through aluminiferous ampliiboles, with the non-aluminous magnesian silicates, and to all of these various magnesian minerals a very different origin must be ascribed." In a farther discussion of this subject, in 1883, it was noted "that decayed feldspars, even when these are reduced to the condition of clays, have not, in most cases, lost the whole of their alkalies." * This was shown by the analyses made by Sweet of the kaolinized granitic gneisses of Wisconsin, from which it appears that " the levigated clays from these decayed rocks still hold, in repeated examples, from two to three hundredths or more of alkalies, the potash predominating." § 64. The question of the source of the matters in aqueous solution, which, according to the hypothesis be- fore us, gave rise to granitic veinstones, naturally comes up at this stage of our inquiry. As we have seen, the granitic substratum of igneous origin, the existence of which is postulated by most modern geologists, is, since the time of Scrope, Scheerer, and Elie de Beaumont, gen- erally conceived to be impregnated with a portion of water, conjectured by Scheerer to equal perhaps five or ten hundredths of its weight ; and through the interven- tion of this to assume, at temperatures far below the point of liquefaction of the anhydrous rock, a condition which has been designated one of aqueo-igneous fusion. This interposed water, under the influence of great heat and pressure, we may suppose, with Scheerer, to consti- tute a sort of " granitic juice," which, exuding from the mass, might fill fissures or other cavities, alike in the granite and in the adjacent rocks, with the characteristic minerals of granitic veins. This seems to have been essentially the view of Elie de Beaumont, who described the elements of the pegmatites, the tourmaline-granites, * The Decay of Rocks Geologically Considered, Amer. Jour. Science (1883), xxvi., 194. Also post, Essay VII., § 10. I la f ^/te universal basic have gone on in the « ly 3' J,„„ the surface of the ,„.V, which we have ^•^Pr"?^" ^■^ .,„d penetrated by .o'.li'^^^ tachvlite-basalts), and tne ny ^ fo the vitreous matter ofJ^ieU^^^^^ ^ ^-^^^^^?J^ffered f-m the (1853) (3) xxxviii., 215-iay.) v.] THE CRENITIO HYPOTHESIS. 131 globe, consolidating at the centre, left a superficial layer of matter which has yielded all the elements of the earth's crust. This last-cooled layer, mechanically disinte- grated, saturated with water, and heated by the central mass, furnished in aqueous solution the silicates which were the origin of the ancient gneisses and similar rocks." * § 68. The transformation of the primary basic layer, judging from the phenomena seen in basic exoplutonic rocks, would give rise not only to quartz, feldspars, and zeolites, but to such aluminous silicates as prehnite and epidote, and to non-aluminous silicates like pectolite, okenite, and apophyllite. These silicates are all non- maguesian, but the reactions of many of them, while in a soluble condition, with dissolved magnesian salts would give rise to various natural magnesian silicates, both aluminous and non-aluminous. § 69. The cooling of the surface of the earth by radia- tion, and the heating from below, would establish in the disintegrated, porous, and unstratified mass of the pri- mary layer a system of aqueous circulation, by which the waters penetrating this permeable layer would be returned again to the surface as thermal springs, charged with various matters there to be deposited. The result of this process of upward lixiviation of the mass would be the gradual separation of the primary undifferentiated layer into an upper stratum, consisting chiefly of acidic silicates, such as feldspars with quartz, and a lower, more basic, and insoluble residual stratum, charged with iron- oxyd and magnesia; the two representing respectively the overlying granitic and the underlying basaltic layers, the presence of which beneath the earth's surface have generally been inferred from exoplutonic i^henomena. The intervention of the argillaceous products of sub- a|3rial decay was considered, and the reactions between * From a repor, a lecture by the author before the Lowell Institute, Boston, Mass., Feb. 29, 18S4, in the Boston Daily Advertiser of March 1. 132 THE OKIom OF CBVBXALI..^ BOCKS. * , them and mineral ^"'fZ^Z^Z^l^^-^^'^''"' ' ■ tured, might give nse to c^"» ^^^ j s 70. Tliat tlie great sl>rmKing ^j ^^^ eolequent upon the re-ov^J-»^{;% ^,^ „,,,,yi„g vast amount o£ matter wtoon ^^ ^^^ ; auitic ana gneissic «e™^;,7„^t.i;i:g deposit, and tl>at : general corrugation of tto ove y S ^^^^^^^_ ^j^^^^ ^ this would P''°»'=''l*y''.'',,fw^ magma, constituting the fissures, of the ^-''"^'y^^.^Zl^^e^e among the most . first eruptive or ^^»f "'"^Jhyp^thesis. These various Obvious deductions fom this W ^^^^^ ^^^ ;„ April points were ~°:;;^tlhthf suggestion that this nevvly Ld May of 1884 with the sugg ^^ ^ ^^^^.^^ ^^^^s, proposed explanation of the o g ^^ ^,„ i . Lough the ac ion o J^/<,,u,a the ceesihc hy- matters from below, might « ^^ g , pothesis, from the Greek J^'J.^ ^^gical ^.^^^^ „f the § "• Tl;' -flhi* !e have BketcLd in the preceding "^^^-^tZ^Z::^ doctrine of ^ I. -1858. An attempt to deduce .^^^ a solid incandescent — 'J^tle^s^^^ 'asic, through ous rock, supposed *" jl' J^V , t^,, jutinct and un- niechanical and 't^era'^^^g^^^^^;;, ^hjch, when subse- liUe classes of/ed'^^^f ^ ^rrane^n heat, should give quently transformed by «""*"* ^^^^^^ rocks. This le two types of acidic and ^s'" -^^^^^i^ nietamorphro woB an attempt to »d^* *he ^4 ,„ silicates. »„,^nt to extend this last conception ^ 111.-1863. An attempt to ex ^^^^^^ ^ . on «« Origin o. 'h« C^^"- ^:rica?N.tur.U.. .or ^-=; i' v.] THE CRENITIC HYPOTHESIS. 133 e g e- h in- 36- ve his hie to by tyd- tion ly of June; ience, to double aluminous silicates, by a consideration of the formation of zeolites at the earth's surface in rocks of secondary age, and also in more recent times, through the action of thermal waters ; it being shown, from the asso- ciation of zeolites with feldspar and quartz in nature, tiiat all these are sometimes formed contemporaneously from aqueous solutions, and also that many feldspathic veins and masses have probably had a similar aqueous origin. IV. — 1871. The subject of granite veins farther dis- cussed, and the mineralogical similarity between these endogenous masses and the indigenous gneissic and gran- itic rocks insisted upon. V. — 1874. The argument reiterated that the condi- tions under which the primitive granitic and gneissic rocks had been produced were essentially similar to those of the granitic veins of the later crystalline schists, and that these conditions are reproduced to a smaller extent, in later times, in the formation of zeolitic minerals: finally, that the gneisses and bedded granites are to gran- itic veins what beds of chemically-deposited limestone and travertine are to calcareous veins. VI. — 1880. The definite assertion of the aqueous origin of stratified crystalline rocks, coupled with the rejection of the doctrines of metamorphisra and meta- somatism in explaining their origin, and the assertion of their pre-paleozoic age. At the same time, the probable intervention of clays from the sub-aerial decay of feld- spars, as a source of certain crystalline aluminous silicates is suggested. VII. — 1884. The definite assertion is made that the ancient crystalline rocks were generated either directly from materials brought to the surface by subterranean springs from the primary igneous rock, or, as was the case in later times, by the reactions of these materials with the products of sub-aerial decay. These latter included clays from feldspars, and dissolved magnesian salts formed by the action upon sea-wate^' of magnesian carbonate set i m m n m 184 THE ORIGIN OP CRYSTALLINE ROCKS. tv. li t. free in the atmospheric decomposition of basic rock erup- ted from tlie primary stratum. Thus, while what may be called the Primitive crystalline rocks were wholly crenitic in their origin, the soluble and insoluble results of the sub-aerial decay, alike of basic exoplutonic matter and of the older crenitic rocks, contributed to the formation of the later or Transition crystalline schists. III. — ILLUSTRATIONS OF THE CRENITIC HYPOTHESIS. § 72. The crenitic hypothesis, which has been proposed in the second part of this essay to account for the origin of the granites and crystalline schists, conceives them to have been derived, directly or indirectly, by solution from a primary stratum of basic rock, the last congealed and superficial portion of the cooling globe, through the inter- vention of circulating subterranean waters, by which the mineral elements were brought to the surface. This view not only comjiares the generation of the constituent minerals of the primitive rocks with that of the minerals formed in the basic eruptive rocks of later times, but supposes these latter rocks to be extruded portions of the primary plutonic stratum which, though more or less modified by secular changes, still exhibit after eruption, though on a limited scale, the phenomena presented by that stratum in remoter ages. The study of these rocks, ^nd of their accompanying secondary minerals, which may be properly described as the secretions of these rocks, will therefore be found very important as illustrations of the crenitic hypothesis. § 73. Without here entering into the details of their geognosy or their lithology, it is sufficient to recall the fact that such basic eruptive rocks abounding in zeolitic minerals are found, with many characters in common, from the time of the Cambrian or pre-Cambrian Keweenian series of Lake Superior to that of the trias of eastern North America, the tertiary of Colorado and the British Islands, and the recent lavas of Iceland. The secreted v.] THE CRENITIC HYPOIHESIS. 135 minerals of these rocks often occur in closed cavities in tufaceous beds, constituting amygdaloids, and, at other times, in veins or fissures of considerable size. They are not, however, confined to the tufaceous or recomposed detrital exoplutonic rocks (which are sometimes them- selves hydrated and transformed into palagonite, as de- scribed by Bunsen in Iceland), but occur in veins and cavities in massive rocks, as is well seen in the diabasu of Bergen Hill, New Jersey, and the massive basalt of Table Mountain, near Denver, Colorado, both remarkable for their zeolitic minerals. § 74. The accumulations of secreted minerals in these conditions are often considerable in amount. Among other examples, it may be noticed that the zeolitic masses in the amygdaloids of the Faroe Islands are sometimes three or four feet in diameter, and constitute a large por- tion of the rock. Veins of laumontite in Nova Scotia attain breadths of a foot or more, while some veins on Lake Superior, which are made up to a great extent of zeolitic and related species, are two and three feet or more in breadth, and often of considerable extent. The history of the chemical composition of the zeolite-bearing rocks of Lake Superior, and of the changes which hi,ve taken place in their degradation from the original eruptive mass, have been studied in detail by Pumpelly, with the help of the previous analyses of Macfarlane, but cannot here be discussed.* . - § 75. We may here notice the modes of occurrence of the zeolites of Table Mountain, Colorado, as described in 1882 by Messrs. Cross and Hildebrand.f The upper forty feet of a great flow of basalt, one hundred feet or more in thickness, show many cavities, large and small, de- * T. Macfarlane, Geological Survey of Canada, 1866, pp. 149-164; Pumpelly, Geology of Michigan, 1872, part 2; also the same, on The Metasomatic Development of the Copper-bearing Rocks of Lake Supe- rior, Proc. Amer. Acad., Boston (1876), vol. xiii., pp. 253-309. t Cross and Hildebrand, American Journal of Science, xxiii., 452, and xxiv., 129. r 136 THE ORIGIN or CUYSTALLINB liOCKS. [v; •.;!y: !i;;l scribed as more or less flattened and drawn out. Some of these cavities are empty, wliile others a'o more or less completely filled by various zeolites, which are also found in fissures m the mass and, in the case of analcite, in a < conglomerate made up of pebbles of basic eruptive rocks, underlying the bed of basali", The zeolitic deposit often appears as "a reddii'.h-yellow sandstone-like material, which occurs in many of the cavities. In the larger ones it takes the form of a floor, the upper surface being hori- zontal, and the deposit may be several inches in thickness. Small cavities have been completely fllled with it, and it is clear that the deposition has taken place from the bottom of each cavity, upward. In parts of Sorth Table Mountain, especially, the same material has filled fissures. Usually the lower part of such masses is composed of a reddish-yellow mineral in irregular grains, which form a compact aggregate, in which lie isolated spherules of a similarly colored radiated mineral. These spherules are seldom more than two millimetres in diameter, and are very perfect spheres. They increase in number upwards, and finally form the greater part of the deposit. In one cavity, six or eight feet in horizontal diameter and about two feet high, the deposit is quite different. Here the main mass is loosely granular, and is formed chiefly by a bright greenish-yellow mineral, while a stratified appear- ance is produced by layers of a white or colorless mineral. Some of the white layers are chiefly made up of easily recognized stilbite, and the same mineral, in distinct tablets, forms the upper layer of the whole deposit. There are also irregular seams of white running through the yellow mineral." „ The greenish-yellow crystalline mineral was found to consist of laumontite, and the other layers were mixtures of stilbite and laumontite, with some of which were found spherules of thomsonite. This, in other cavities, formed layers by itself, without admixture of the other zeolites mentioned. The presence of these zeolites in cavities 01 ■I - ▼0 THE CRENITIC HYPOTHESIS. 137 side by side with other cavities v.hich were entirely empty, is, according to the writers whom we have quoted, appar- ently due to the fact that the former communicated witli fissures which were channels for the percolating waters that deposited the zeolites. Such fissures, filled up with similar zeolites, were in many cases found leading to these cavities. § 76. The eruptive rocks which break through the Trenton (Ordovician) limestone at and near Montreal, in Canada, are of various ages and unlike composition. Some of these are highly basic, and have been described as dolerites and diorites, while some have been found to contain analcite, and others again much nephelite, and have been refei'red +,o teschenite and nepheline-syenite. In some fine-grained amygdaloidal varieties of these basic rocks, which have been designated dolerites, I long since described the occurrence of heulandite, chabazite, anal- cite, and natrolite, with quartz and epidote.* These zeo- lites are not abundant, but in certain of the basic doleritic rocks on Mount Royal I have found remarkable veins of orthoclase with quartz and other minerals, which merit a notice in this connection. Included in veitical dikes of these rocks, themselves cutting the horizontal limestones which appear at the base of the mountain, are frequent granitic veins, sometimes twelve inches or more in breadth, parallel with the walls of the inclosing dike, often distinctly banded, and exhibiting a bilateral sym- metry which, together with their drusy structure, shows them to be endogenous. The most characteristic of these veins are made up of white, coarsely crystalline orthoclase with a little quartz which, in druses, presents pyramidal forms. In somo of the veins. Dr. Harrington has since detected, besides orthoclase and quartz, nephelite, sodalite, cancrinite, amphibole, acmite, biotite, and magnetite. All of these minerals are seemingly secretions from the en- closing basic exotic rock. * Hunt, in Geology of Canada, 1863, pp. 441, 655, and 668; also Har- rington, Report Geol. Survey of Canada, 1877-78, p. 43, G. r m *■! I m 138 THE ORIGIN OP CRYSTALLINE ROCKS. [V. § 77. The mineral secretions of tlio basic eruptive rocks may bo conveniently grouped under seven heads, as follows : — 1. The aluminous silicates, including the zeolites prop- erly so called, to which we append the related hydrous species, prehnite, and the associated species, orthoclase, garnet, and epidoto, which are found in the amygdaloidal rocks of Lake Superior. To these we must add albite, axinite, tourmaline, and sphene, observed by Emerson, in 1882, in a diabase dike in the trias at Deerlield, Massa- chusetts,* and also the various aidiydrous aluminous sili- cates found with orthoclase in the veins on Mount Koyal, just described. 2. The group of hydrous protoxyd-silicates, the bases of which are lime and alkalies, and of uiiich pectolite may be taken as the type. These 8i)eeies are sometimes wrongly spoken of as belonging to the class of zeolites. As an appendage to this group, we note the hydrous boro- silicate of lime, datolite, frequently found in these rocks. Mention should here also be made of the anhj^drous pro- toxyd-silicates, amphibole and acmite, in the feldspathic veins of Mount Royal. We have already called attention to the occurrence of amphibole and pyroxene in granitic veins under other conditions (§ 57). 8. Quartz in its various crystalline and cryptocrystal- line forms, as rock-crystal, amethyst, chalcedony, agate, and jaspery varieties, is found both alone and associated with the minerals of the preceding groups. Hyalite of very recent origin has also been observed by Emerson at Deerfield. 4. The oxyds, magnetite and hematite, are frequent in the zeolite-bearing rocks of Nova Scotia, where both of these species form /eins in amygdaloid, and where mag- netite moreover occurs in drusy cavities with quartz, lau- * Emerson, Amer. Jour. Science, xxlv., pp. 195, 270, and 329. We re- serve for another occasion the discussion of tlie paragenesis of the min- erals of this locality, so carefully studied by Emerson. v.] THE CUENITIC HYP0TIIi:SI8. 189 of at in of inontito, and calcito. Homntito, in tho form of plates of specular ore, is also found there in veins with lauinontite, and niangiinese-oxyd is sometimes associated with these iron-oxyds. Small crystals of hematite on prehnite, with a little niiiiiganese-oxyd, have been observed by Emerson at the Deerfield locality, as also cuprite on datolite, and malachite on prehnite. In bimilar associations lie found moreover small portions of various sulphids, such as chal- copyrite, pyrite, sphalerite, and galenite. (ylnfe.page 121.) 5. The presence of native co[)per, and occasionally of native silver, associated with the various silicates already named, should also be noticed. The former metal is com- mon to the zeolitic rocks of Lake Superior and Nova Scotia. 6. Mention should here be made of the saponite often found in amygdaloidal rocks, which, in its purer form, is a hydrous silicate of magnesia with but little alumina or iron-oxyd. Matters, apparently of this class, fill, or more frequently line, amygdaloidal cavities which are filled with other species. This magnesian hydrous silicate is per- haps distinct in origin from the delessite or iron-chlorite which is a frequent constituent of many basic rocks, such as the melaphyres of Lake Superior, and is probably not a secretion but a residual product of the transformation of the rock. . 7. Calcite in various forms is a common species in the rocks in question, and fluorite and barytine may also be mentioned as accidental minerals therein. It is principally with *he first two classes of minerals, the zeolitic group with its appendages, and the pectolitic group that we have to do. These two, as is well known, though chiefly found in the eruptive rocks already noticed, are not confined to them. Some species of zeolites occur occasionally in veins in gneiss and other crystalline rocks, aild even in limestones and other sedimentary deposits. These occurrences are the more readily understood when we consider that the same minerals have in various locals Hi 'ill mamm >. |!(. I 'L.^miLm^it^mimmaii'mgimmtmmmi Hlf'M li 140 THE ORIGIN OP CRYSTALLINE ROCKS. [V. ties been recently formed by the action of thermal waters, and are even generated iu submarine ooze. Many of the species of these two groups have also been formed artifi- cially in the chemist's laboratory. § 78. It is our present purpose to consider, first, the zeolitic, and secondly, the pectolitic group, both as regards their chemical composition and their relations to various anhydrous silicates. We shall then proceed to notice the action of water at high temperatures on glass and similar bodies, in giving rise to various crystalline species, includ- ing quartz. In this connection will also be discussed some facts relating to the chemistry of tne alkaline sili- cates. We shall next notice the action of thermal waters in historic times, aiiv' the occurrence of zeolites in the clays of the deep sea, and then pass to the experiments on the artificial reproduction of zeolitic species in the labora- tory of the chemist, and discuss the relations of hydrous and anhydrous species. From this, we shall proceed to a consideration of the reactions of the hydrous species of the two groups with magnesian s?its. The origin of these salts through sub-aerial decay of exoplutonic magnesia- bearing silicates, and their relation to the primeval sea, will then claim our notice ; after which will be considered the probable relations of the clays from the sub-aerial decay of feldspathic rocks to other classes of rock-making silicates. The conditions of crystallization of mineral matter will next be considered in relation to the formation of rocks, after which the conclusions of our present study will be briefly summed up in the fourth and last part of this essay. § 79. In the accompanying table of zeolites and related species, are placed, in the first column, the names of hydrous species, and in the second column the ox3'gen- ratios between the protoxyd-bases, the alumina, the silica, and the water, represented respectively under R, r, Si, and H. In the fourth column are given the names of corre- sponding anhyd' ous species. In this and the succeeding V.1 THE CEENITIC HYPOTHESIS. 141 TABLE OF ZEOLITES AND BELATED SPECIES. HYDRO0S. E : r : SI : H R Anhydeous. Thomgonite ...... 1:3:4:2 Ca, Ka. Anortbite,etc. Oismondite 1 : 3 : 4i : 4} Ca. Nephelite. Fsmarkite Fahlunlte 1:3:5:1 1:3:5:2 Mg. Mg. Barsowite, lolite, eto. Natrolite Scolecite Mesolite Levynlte 1:3:6:2 1:3:6:3 1:3:6:3 1:3:6:4 Na. Ca. Ca, Na. Ca, Na. Labradorlte. Analcite Kudnopliite Laumontite Herachclite PhiUipsite Ckabazite Gmelinite 1:3:8:2 1:3:8:2 1:3:8:4 1:3:8:5 1:3:8:5 1:3:8:6 1:3:8:6 Na, Ca, K. Na. Ca. Na,K. Ca.K. Ca, Na, K. Ca, Na. Andesite, Hyalophane, Leuoito. * Faujasite Hypo8tilbite Puflerite 1:3:9:6 1:3:9:6 1:3:9:6 Ca, Na. Ca, Na. Ca. Oligoclase. Harmotome 1 : 3 : 10 : 6 Ba. ? Heulandite EpiBtilbite Brewsterite ...... Stilbite 1 : 3 : 12 : 6 1 : 3 : 12 : 5 1 : 2 : 12 : 6 1 : 3 : 12 : 6 Ca. Ca. Ba, Sr. Ca. Orthoclase, Microoline, Albite. Prehnite 2:3:6:1 Ca. ? Jollyte 1:2:3:2 Mg, Fe. ZoiHito, eto. :f ' ywaw— 142 THE ORIGIN OP CRYSTALLINE ROCKS. [V. tables I have generally followed the terminology and adopted the formulas given in the fifth edition of Dana's "System of Mineralogy." In the line with the most basic zeolite known, thom- sohite, is placed the feldspar, anorthite, and with nephe- lite is coupled the hydrous species gismondite, a true zeolite. The recent analyses, by Cross and Hildebrand, of the zeolites of Table Mountain, Colorado, give for the zeolites having the characters of thomsonite a proportion of silica greater than corresponds to the formula of that mineral given by Rammelsberg, which we have placed in the table. Some of their analyses, while yielding almost exactly the other ratios of the formula, give for silica, instead of 4.0C the numbers, 4.65, 4.76, and even 5.17 ; showing a composition more silicious than that of gismon- dite, and approaching that of a zeolite corresponding to fahlunite, barsowite, and bytownite. These chemists, while believing the specimens analyzed by them to repre- sent a pure and unmixed mineral, leave undecided the question of its real composition. § 80. The feldspar which has been called barsowite and bytownite, according to several concordant analyses is as distinct from anorthite .s it is from labradorite, and apparently as much entitled to form a distinct species as the latter feldspar, or as andesite or oligoclase. The composition of a lime-barsowite, with the ratios, 1:3:5, would be silica 48.54, alumina 33.33, and lime 18.13 = 100.00. With barsowite has been placed iolite, which is a magnesia-iron silicate, giving the above ration, and, as I long since pointed out, is from its atomic volume enti- tled to be regarded as a feldspathide. These various anhydrous species would appear to correspond very nearly with the so-called thomsonite of Cross and Hilde- brand. With this anhydrous group we have placed two hydrous magnesian species, the one, esmarkite, also called praseolite and aspasiolite, and the other fahlunite, which includes what have been called auralite and bonsdorffite. and i hyc thej dial spoj COM Joii v.] THE CRENITIO HYPOTHESIS. 143 These species are often associated in. nature with iolite, from which they differ only in the presence of Avater, and they have been by most minerak)gists regarded as formed by subsequent hydration from this mineral. This view, however, was contested by Scheerer, who regarded the association of the hydrous and anhydrous minerals as due to a simultaneous crystallization of two isomorphous species.* The relations of the silicates of the natrolite section to labradorite are obvious from the table. The same may be said of the relations of the numerous silicates of the analcite section to andesite, hyalophane, and leucite, and of the faujasite section to oligoclase. It is to be noted that the well-defined zeolite, harmotome, has as yet no corresponding anhydrous silicate. Of the heulandite sec- tion, and the corresponding feldspars, orthoclase and al- bite, it is to be remar!:ed that orthoclase and albite are the only feldspars hitherto found associated with zeolites, and the only feldspars as yet artificially produced in the wet way. The observations of Whitney already noticed (§ 57) have since be.en fully confirmed by Pumpelly, who finds orthoclase very common with the zeolitic minerals on Lake Superior, where its deposition is shown to be posterivjr to laumontite, prehnite, analcite, apophyllite, quartz, calcite, copper, and datolite ; the only species superimposed upon it being calcite, chlorite, and epidote, which latter also occasionally occurs between laumontite and prelmite in order of superposition, f § 81. We have placed at the end of the table the two hydrous silicates, prelmite and jollyte, though neither of them presents the ratios for protoxyds and alumina which characterize the zeolites. Prehnite has no known corre- sponding anhydrous silicate, while jollyte, though a less common species, is interesting inasmuch as it affords the I .;£■ ■\i. :• i. * Amer, Jour. Science (1848), v. 385, from Pogg. Annalen, Ixviii., 319. t See Pumpelly, Geology of Michigan, already cited, § 74; also Amer. Journal Science (1871), ill., 254. in ill I THE ORIGIN OF CBYSTALI-INE ROCKS. tv. Ill I oxygen-ratios of the anhydrous zoisite and th°i nearly anhydrous species epidote. It has also the o'cygen-ratios of meionite of the scapolite group, an anhyd)0U8 silicate which, however, belongs to a much less coudensed ^^ype than zoisite, as is indicated by its inferior density and hardness, and its ready decomposition by acids. I have elsewhere discussed the relations of these two silicates, and have shown that the density, hardness, and chemical indifference of epidote and saussurite assign them a place with garnet and idocrase, in the grenatide group ; while meionite, though lacking the proper feldspar-ratio between protoxyds and alumina, belongs to the feldspathides.* [The recent conclusions of Tschermak as to the oxygen- ratios of the scapolites are set forth in Essay VIII., on "A Natural System in Mineralogy, etc.," § 75-78, in which essay, under Tribes 6, 7, and 8, will be found dis- cussed at length the chemical constitution and history of the principal aluminous double silicates here noticed.] § 82. It is to be noted that the p.rotoxyd-bases of the zeolites and their related feldspathides are either alkalies or lime, baryta or strontia, ii we except the partially magnesian zeolite, picrothomsonite, and iolite and some related hydrous species, which, besides magnesia, include ferrous oxyd. The latter base enters also to some extent into epidote and prehnite. It should also be remarked that small portions of ferric oxyd are frequently found in the analyses of zeolites, amounting, in the red varieties of laumontite to three or four, and in some natrolites to one and two hundredths. Some part of this, however, is dis- seminated in the form of hematite, giving color to the zeolites, and recalling the association alike of hematite and magnetite with zeolites, as already noticed, as also a similar occurrence of these oxyds crystallized in many granitic veins. § 83. We next come to the hydrous silicates of lime and alkalies, which we have called, for convenience, the * Chemical and Geological Essays, pp. 445-447. TJ its fenoi and secoi ill lii'ii I y.j THE CRENITIC HYPOTHESIS. 145 pectolitic group, and which are correlated in the accom- panying table with other protoxyd-silicates having similar oxygen-ratios, chiefly magnesian, and partly hydrated and partly anhydrous. We have indicated in the second column, for the known silicates of the pectolitic group, the oxygen-ratios of R, Si, and H, as in the former table, and have left a blank under H, where, as in the first three terms, for example, no nou-maguesian species is known. TABLE OF PROTOXYD SILICATES. Pectolitic. R : SI : H 4:3:- Chondrodite, Humite, etc. ? . . . 1:1: — Chrysolite, Monticellite, Phenaclte, etc. ? . . . 3:4: — Serpentines, Leucophanite. Gyrollte, etc. . 2:3:1 Deweylite, Genthite. Xonaltite . . Plomblerite . 1:2: J 1:2:2 WoUastonite, Amphibole, Rhodonite, Pyroxene, Enstatite, Cerolito. Pectollte . . . 6 : 12 : 1 Amphibole in part, Spadaite. ? . . . 2:6: - Talc in pa. ? . . . 1:3: — Sepiollte, Talc in part. (Unnamed) . . Okeuite . . . Apophyllite . 1:4: J 1:4:2 1:4:2 Titanite, Ouarinite. The first place in the table is given to chondrodite, with its sub-species, the most basic natural protoxyd-silicates known, and remarkable for the replacement of a small and variable proportion of oxygen by fluoriuQ. In the second line, besides the chrysolites (including the pure yj I'l . ni^IGlN OF CEYSTAI^M BOCKS. tV. il6 THE OBlGIIi UJ= ^- nuo nnd phenacite, ^agnesian specie, f"— iCf ud^r^pe-S' to- There uamea, belong *^ ''^ ^J / rfes, viUa.-site, the trandite, the hydrous '^^^^^^ ^ i,,, wiUemite and anhydrous rincie and ^"f ^X the third Une are to be tephroite, with roany f"'^-'„^i^„ silicates generaUy placed the various hydrous "^^'^ussed f-.ther on m Lown as serpentine; ^^^r^Z^ ies, including the Essay Vin- 'h'' ^'"'T -nlUe the foliated thermophyl- , smatiechrysoUteandp^r hte^th^t^^ eoUoid species, lite and marraoUte, and the ""^ '. ^1,^ nameof serpen- ';i,;alite, and that ^^ ^^^J^ i^lTphU, P«sents the tine. The anhydrous SP^"^^' '^ /^ 3iiiea as the seipen- tle atomic ratios for protoxydsand^^^^^^ tine group, for ^''"«l>,f "^'X oJ note that, as Daubr^e nragnesian speeds " ''. '^"JXdrated and (used, hreata has shown, serpentnie, when eny ^„a enstatite, np into a mixture of "-^'^^i^^/iUs intermediate in com- t^een ^vWcl. exclutog wat^^^^^^^^ „Hte, may be position* With the hydrous U ^^^ ^^^ ^^lon- ■ the niccolio species, genthite^ ^^^^.^^ ^j ^i^jaeates, s 84. We come next to tne g woUastomte, en- .epresented among ^M-^^ "nd the manganesian ■ statite, pyroxene, '^ny amphAo ^^^^^ sub-specres. ,necie3 rhodonite, with i-^l*""' "l .„ UsiUcates, picros- ^^ these are the y*^""' "S^r with hydrorhodo- mine, aphrodite, and "=«■»''*?' '"^These various bisilicates Te 'dioptase. a-^ ^Xp iolil'cl^oup by plombier te are represented among tepec^t^^^ ^^^^^ ^y Dau- and xonaltite; the «"»" ^^^^^^ at the hot spring o£ b,fe in the process «« £°™ a,, o^ygen-ratio, 1.2.2. plorabiSi-es in France, »* "\ H ^^y „£ vemark that, Ofthelesshydratax— .t^^^^^ ♦ ComptesRendusdelAcaa-u T.l THE CRENITIC HYPOTHESIS. 147 as observed by Rammelsberg, it occurs in concentric layers with the anhydrous species, rhodonite (bustamite), and the hydrous quadrisilicate, apophyllite. While many amphiboles have the ratio of a bisilicate, others are believed to have a ratio (excluding a little water) of 4:9, not far from that of pectolite, with which we have placed them. Here also comes the hydrous mag- nesian species, spadaite. Different analyses have assigned to talc the ratios for the fixed bases of 2 : 5 and 1 : 3 (the water being variable), — the latter corresponding to sepio- lite, 1:3:1. For neither of these do we know any cor- responding pectolitic silicate. § 85. We come, in the last place, to the quadrisilicates, which have no known representatives among hydrous magnesian species, or among anhydrous silicates, if we except the titanosilicates, titanite, and guarinite. They are, however, represented in the pectolitic group by no less than three species, okenite, apophyllite, and an un* named species got artificially by Daubr6e. It is fibrous, like okenite, is decomposed by acids, and is a hydrous sili- cate of lime, with six per cent of soda, giving the ratios, 1:4:^. Pectolite, it will be recollected, contains in like manner about nine per cent of soda, while apophyllite contains five per cent of potash and a little fluorine. § 86. The process by which this unnamed pectolitic silicate was obtained by Daubrde is very instructive, as showing, in many ways, the action of heated water on an undifferentiated silicate of igneous origin. He took for the subject of his experiments a common glass, the analy- sis of which gave silica 68.4, alumina 4.9, lime 12.0, mag- nesia 0.5, and soda 14.7 = 100.5. Tubes of this glass were sealed up, with many precautions, in tubes of iron, with about one-third their weight of pure water, and exposed during several weeks to a temperature not less than 400° C. At the end of this time the glass was found to be completely disaggregated and changed into a white fibrous or lamellar substance, composed in great part of il I n^GIK OF CET8TA1.UNE MOM. f». 11« THE ORIGIN Ui! f lime and soda in tte fusible reotoWc '/.,j,re also included mi- Z cryslls of tWB latV^ouu.. -1 were ^^^^ ttc/l-ic g^i- ° ; *t:..„er. "he iron of these mm- nr Tiicotite, probably tne lo ^^^^^ : als ^vas perhaps derived fr- the v ^^^.^_^ ^^ ,^ ^ S 87. The net result of the P J ^.^.^^^^ „p water on the glass ^™.s *at f e v .^^ ^^^j^_ ^„4 L per cent of its siUca, 6*.« P^°%it,i the remaining ttn Tr cent of its alumina; the lime, w j^^^i^edths) mea' and soda and -^^"^ Z ^pa-'ed silica ming the pectolitic sUiea e U ^^ ^^^^ „j.,t,Uaed fhe larger part separated in the r ^^^^ ^^^,^,3 corresponding to an "^yS^^/,^ ,„,i rfs, 85.0 per cent of But as, according to D»ubr6e s j ^^^ ^^^^^ „^e the alumina had P»f ^^f ° Jthan 9.7 parts of alum>n«- for 63 parts of soda n°' '"'^^i<,„.ai„minate in solution a which should give for the sUieo ^^ ^^^^^ ji- ■ cance wmcn n'stW^iasrecora^expen^^^^^^ above made to 'i«te™"'%f' 'tl such as obsidian and wTtcr upon vitreous volcanic r»*8, b ^^^„g,„ ,„- Lute, which gave similar res^^s g^ ^j ^^. cording to him, not so we^ detoe ^^ ,, din, of oligoclase, of P^t^^'-^t^t change, though m- the'se tubes suffered J "W-en^^^ed from the glass, crusted with crystals ot quari v.] THE CRENITIC HYPOTHESIS. 149 This stability was to have been expected from the fact that crystals of pyroxene are formed under similar con- ditions, and, as we shall see, both albite and orthoclase have since been crystallized at high temperatures in pres- ence of solutions of alkaline silicates. Another experi- ment, mentioned by Daubrde in this connection, is important. By heating in a glass tube with water a refractory clay (probably under similar conditions to the preceding exi)eriments), this became filled with white pearly hexagonal scales, resembling a mica. They were fusible, attacked by hydrochloric acid, and contained both silica and alumina, being seemingly a product of the ac- tion of the alkaline silicate from the glass upon the infus- ible kaolin.* Daubrije recalls in this connection the observations of Fr%- i'-\ ) It . IV. "ISO TUcJ vi**""' rfnot taken, a i.ovUoa of -"J^on by peroxyaato. which is not scpavatea t'o™ t le ^^^^^^^ ^^^^^ of Zl but -n-eviectly ^ »^1>"^^^^ ^^.j^,i„„, but the liquA the water-glass is ^ '" "7„„,^.%iear again on concentia- ^- '■t:ts::w\i"- : fe.{ a.p» ^^ -^ :Xuon":f a metallic ^"J^Z I -dissolved wa evglass, the P™7'"^,tJe thus takes up no incon- ; .giition. " A bcimd ^'>;'=;f J\,„„, ,i„c, manganese, denvble amount of *!>« °^y^^J„„,y." By agitatmg a tin, antimony, copper, a^>d mere y ^^^^^.gi,^, ^^ a ,tiou ot ferrous sulphate wi» ° . . which, after ;tser;rrtly fined with «^-;^XV This solubility of Xtion, has a very deepj-l-jl of alkaline siUcates f..ilip oxvds m aqueous »u obscure tactb :« Va rational e.;— » "^ J^ ..^ presence to mineralogical '='^^™'''y> ''^Lev-oxyds, and of metalhc b Je and others on the e""*'^^ „i„eral species, by e stalline zeolites, -"'^ J^^^liwaters on the bricks Z slow action of «™»^ *rmasonry in France and , and mortar of ancient Koma ^^ ^^at his Atoria. It was at ^ '""^''^'Xhot water, here rising tst of^evvations were made Th ^^^^^^^^ ^ j^^,^ f om a fissure in a 8'™*° J" "fe superficial waters, the 'gCel, and to protect i rom tl.J^ P ^ ,, e^e lomans had capped the splig.^^ ^^^ P'"**^ "C er a •-n Cm bCa^h^hl concrete, exten mg -- rghof-rethanahundredmeu^^.-/^ I^^ Ses in thickness>e wf - w ^_^^_ The w.ter, hav- through vertical ctann^^^^^^^^^^^^^,_^^,, 33, . v.] THE CRENITIC HYPOTHESIS. 151 ing at its outlet a temperature of 70° C, fills the gravel beneath the roof of concrete, and a portion filters slowly upward through this. The concre^ itself was made of fragments of burnt red brick, with others of sandstone and of a friable granite, the whole in a calcareous cement. Repairs having required cuttings to be made in this mass, it was found to contain numerous crystallized mineral species, foi-med through the action of the water, which were examined by Daubr(3e, with the aid of De Senarmont for the crystallographic determinations, and first described in 1858. § 91. The substance of the fragments of brick was found to be altered to some depth, while the numerous cavities therein were lined or filled with various matters, often distinctly crystallized. Among these were identi- fied chabazite and phillipsite (christianite), gismondite, implanted on the chabazite, scolecite, and what is desig- nated by Daubrde as mesotype (thomsonite or natrolite). In the calcareous cement were well defined crystals of apophyllite, containing, as usual, a little fluorine ; while in cavities in the lower part of the concrete, near the gravel, was found an abundant gelatinous matter, which was detected in the act of deposition in recent cuttings in the mass through which the water was still oozing. This matter elsewhere had consolidated into a white mammillary concretionary fibrous substance, which was found to be a hydrous silicate of lime, with but 1.3 hun- dredths of alumina, and constitutes the pectolitic species, plombierite, already noticed (§ 84). With the various minerals in the concrete were also found an abundant deposit of silica in the form of hyalite, and, more rarely, crystals of tridymite, and globules of chalcedony, together with calcite in well defined crystals, arragonite, and fluo- rite. The chabazite was often found adherent to frag- ments of wood enclosed in the concrete, recalling, as observed by Daubrde, the similar occurrence of zeolites with fossil wood in lacustrine limestone in Auvergne. 162 THE OUIOIN OF CUY8TALLINE ROCKS. :ir^i; The variouM minerals named were absent from the frag- ments of friable jj;ranite, while in the underlying gravels the only matter deposited was an amorphous aluminous silicate, compared to halloysite, and found also in the con- crete. § 92. The fragments of red burnt brick in the cement had undergone an alteration from their surface, marked by concentric lines of changed color, as well as by the development of zeolites, and also of an amorphous nuvtter compared by Daubrdo to palagonite. In these fragments, the amount of combined water had increased from two or three hundredths in the centre, to eight hundredths in the outer inliltrated portion, in which the amount of matter soluble in nitric acid was equal to fourteen or fif- teen hundredths, including a notable proportion of potash, supposed by Daubrile to have been taken up from the waters. The silica, alumina, and lime of the new mineral species were derived from the cement and the biicks, the calcination of which had probably rendered them more susceptible to chemical cliange. As has been pointed out by Duubrde, the resemblance between these species and the similar ones found in many rocks extends even to minor details of crystalline form and association. The small geodes linad with crystals, in the bricks, as the writer can testify, cannot be distinguished by inspection from many similar cavities in certain amygdaloids. § 93. Similar phenomena have since been noticed in the ancient constructions around the thermal waters of Luxeil, Bourbonne, and otliers in France, and at Oran in Algeria. These localities have added little more to our knowledge of the production of silicates, though at some of them, and notably at Bourbonne, besides zeolites, have been found various crystalline metallic sulphides derived from the transformation of metallic objects enclosed in the concrete. The water of the last named locality, which, unlike that of Plombidres, rises from the muschel- kalk, has a temperature of about 60° C, and is a neutral ,/ f el- lal V.J THE CUKNITIC HYPOTHESIS. 168 saline contiiining seven or eight thousantltlis of mineral mutterH, cliielly 8ulphates and cliloricls of alkalies, and of linie and magnesia; while that of PlombiOres contains only about tlireo ten-thousandths, and is also said to be neutral. As renuirked by Daubr<;e, it is probable that the a'jtion of the water in the formation of these mineral silicates is, to a great extci't, independent of its composi- tion, since pure water, in aciing upon finely divided alka- liferous materials, soon becomes itself alkaline. As regards other silicated deposits from thermal waters, we may notice the case of the L-.ths of St. Honord (Nievre), the waters of which, having a temperature of 81° C, yield a finely laminated white translucent sub- stance in concentric layers, which appears from analysis to be a hydrous silicate of alumina, with a large excess of silica, but is probably a mixture. Mention is also made of a similar deposit from a mineral spring at Cauterets, which is talcose in aspect, and, according to qualitative analysis, is a silicate of alumina, with magnesia and alka- lies.* In this connection mention should be made of the occurrence at the thermal spring of Olette (Pyrenndes Orientales) of a crystalline silicate, having, according to Descloizeaux, the crystalline form of stilbite, of which it has also the composition.! § 94. As an example of a zeolite apparently in process of formation, may be mentioned the observations of R. Hermann, who found in the crevices of a columnar basalt at Stolpenau, in Saxony, an amorphous white plas- tic substance, which after some time changed into acicular crystals of scolecite.J More recently, Renevier has de- scribed the occurrence of a white subtranslucent matter, unctuous to the tou h, gelatinous at first, but becoming a * For a summary of the observations of Daubrde, the details of which are found in several papers, see his G^ologie Experimentale, 1879, pp. 179-207. t Cited by Dana, System of Mineralogy, 5th ed., p. 443. t Jour, f iir Prakt. Chemle, Ixxii. Cited by Dana, System of Mineral- ogy, su6 roce Scolecite. • ■ *. ■ n I m M 1" 1 if 1 , 'lis'''' ■f '£»-,, I II' 1 ' '• uk '^ i I" 154 THE ORIGIN OF CRYSTALLINE ROCKS. [V. plastic mass, and called by the quarrymen " mineral lard," found in constructing a tunnel in the molasse or tertiary sandstone near Lausanne, in Switzerland, in 1876. This substance, which formed layers of from one to three centi- metres on tlie walls of fissures, was said by observers to have, in some cases, taken on a crystalline form, a fact, however, which Renevier was not able to verify. When dried at 100° C, it was found to be a hydrated double aluminous silicate, giving; the oxygen-ratios of chabazite, 1:3:8:6; the bases being lime and potash, with 3.14 per cent of magnesia.* § 95. A remarkable fact in the history of zeolites is- that lately made known b}^ the researches of Murray and R^nard, that a decomposition of volcanic detrital material goes on at low temperatures in the depths of the ocean, transforming basic silicates, "represented by volcanic glasses such as hyalomelane and tachylite," into a crystal- line zeolite on the one hand, and the characteristic red clay of deep-sea deposits on the other. To quote the lan- guage of the authors, this process, " in spite of the temper- ature approximating to 0° C, gives rise, as an ultimate product, to clearly crystallized minerals, wliich may be considered the most remarkable products of i iie chemical action of the sea upon the volcanic matters undergoing decomposition. These microscopic crystals are zeolites, lying free in the deposit, and are met with in greatest abundance in the typical red-clay areas of the central Pacific. They are simple, twinned, or spheroidal groups, which scarcely exceed half a millimetre in diameter. The crystallographic and chemical study of them shows that they must be referred to christianite,"t which is but an- other name for phillipsite. We have here, as in the case of palagonite, and in ordinary zeolitic rocks, the breaking-up of a basic igneous silicate into an acidic crystalline alumi- nous silicate of lime and alkalies, and a more basic insolu- * Bull, de la Soc. Vaudoise des Sci. Naturelles, x t Lecture, in Nature, June 5, 1884, p. 133. disl . tioj eqi will pr< drii mil ha^ beij digl ±0. v.] THE CRENITIC HYPOTHESIS. 166 ble residue, rich in iron-oxyd ; a portion of which, as is well known, separates from these red clays in the form of concretions, often with oxyd of manganese. § 96. We have next to examine the conditions under which zeolites, fekbpars, and related silicates have been artificially produced in the chemist's laboratory. When, according to Berzelius, three parts of silica and two of alumina are fused with fifteen parts or more of potassic carbonate, and the cooled and pulverized mass is exhausted with water, there remains a double silicate, which has the composition of a potash-anorthite, with the ratios, 1:3:4, corresponding to potash 28.68, alumina 32.04, and silica 39.31 ; the excess of silica being dissolved as an alkaline silicate.* The analogous soda-compound may be pro- duced in like manner. A similar silicate, according to Ammon, is obtained when recently precipitated alumina is added to a moderately concentrated and boiling solution of caustic soda, mixed with silicate of soda. The alumina is at first completely dissolved, but a white pulverulent precipitate soon separates, which is a hydrous silicate of soda and alumina, having for the fixed bases the same ratio as before, 1:3:4; corresponding to anorthite and to thomsonite. f § 97. C. J. Way, in his studies on the absorption of bases by soils, prepared artificial aluminous silicates by dissolving alumina in soda-lye, and adding thereto a solu- tion of silicate of soda containing not more than one equivalent of silica to one of alkali (R : Si=l : 3), to which any convenient excess of soda might be added. A precipitate was thus obtained, which, when washed and dried at 100° C, was a white pulverulent silicate of alu- mina and soda, holding twelve hundredths of water, and having almost exactly the oxygen-ratios, 1:3:6:2; being a true soda-mesolite. This artificial silicate, when digested with lime-water, or with any neutral salt of lime, * Cited In Gmelin's Handbook, iil., 431. t Jaliresbericht der Chemie, 1862, p. 128. 'i! IE tm' II 156 THE ORIGIN OP CRYSTALLINE ROCKS. -B(t I exchanged its soda for lime. It was difficult thus to sepa- rate the whole of the soda, but in some cases the replace- ment was almost complete, and a scolecite was formed. Either of these compounds, when digested witli sulphate or nitrate of potassium, was converted into a potash-meso- lite. With a solution of a magnesian salt, these com- pounds gave a magnesian double silicate, which was not particularly examined.* Berzelius, by adding a solution of silica to one of alumina in potash, in proportions which are not indicated, found the mixture to solidify in a few minutes to an opaque jelly in consequent e of the separa- tion of a silicate of alumina and potash having tl" oxygen- ratios, 1:3:8, which are those of analcito.t Farther investigations are required to make Icnown the precisu conditions for the production of these different silicates, which give for their fixed ele^ients the ratios respc'ctively of thomsonite, meholite, and aiialcite. The most basic of these, according to Berzelius, is formed in the presence of an excess of a soda-silicate. § 98. Henri Ste.-Claire Deville, by mingling solutions of silicate of potash and aluniinate of soda in such propor- tions as gave for the oxygen-ratios, al : Si= 3 : G, obtained a gelatinous precipitate, Avliich in sealed tubes, at temper- atures of from 150° to 200° C, was gradually changed into hexagonal plates of a potash-soda zeolite with the oxygen- ratios, 1:3:6:2, having the physical characters of levy- nite. The residual liquid was nearly free from both silica and alumina. On repeating this experiment at a higher temper iture, a very diffe^'ent result was obtained. There was an abundant separation of silica in crystalline grains, with a little levynite, while an alkaline aluniinate remained in solution. This remarkable dissociation of the first- formed aluminous silicate into free silica and soluble alumina recallti the conditions of the separation of quartz, * Way, On th i Power of Soils to absorb Manures, Trans. Royal Soc. Agriculture, 1852, xili., 12.3-143. t Cited in Gmeliu's Handbook, Hi., 439. by J not the |: spai- a mij tion c an e: cryst, at a quenl albite with in thj from howei the e^ ing ir soda was * Til precedil des JVIijT t cJ n THE CRENITIC HYPOTHESIS. 157 already noticed in § 87. The crystalline silica produced in this reaction may be either quartz or tridymite, which latter form of silica, mingled with quartz, was obtained in 1879 by Friedel and Sarrasin by heating gelatinous silica with an alkaline solution to about 400° C. The dissocia* tion of alumina from silica, observed in this experiment, serves to throw light on tlie origin of corunaum and spinel. In other experiments with mixtures of solutions of silicate and aluminate of potash in sealed tubes at 200° C, Deville got a crystalline compound with the formula of phillipsite, 1:3:8:5. Subsequently, De Schulten, in similar experi- ments, at 180° C, with silicate and aluminate of soda, obtained crystals of analcite, with the ratios, 1:3:8:2.* § 99. More recent investigations in the same direction by Friedel and Sarrasin are very instructive, as showing not only the generation of feldspars in the wet way, but the production at will, under similar conditions, of a feld- si^ar or a zeolite. These chemists had already, by heating a mixture of silicate v 'I alumina (precipitated from a solu- tion of chloride of aluminium by silicate of potasli) with an excess of a solution of silicate of potash, obtained crystals of orthoclase, mingled with crystals 'o£ quartz or. at a more elevated temperature, of tridymite. In subse- quent experiments, undertaken for the production of albite, a similar hydrous silicate of alumina was mingled with a solution of silicate of soda (the silica and alumina in the proportions of the soda-feldspar), and heated to from 400° to 500° C. Instead of the anhydrous albite, however, were obtained crystals of analcite, 1:3:8:2; the excess of silica, with soda and some alumina, remain- ing in solution. When, however, an excess of silicate of soda was employed, the whole of the silicate of alumina was transformed into albite. f Thus analcite, which is * The results of Deville, Friedel and Sarrasin, and De Schulten in the preceding paragraphs are cited from Michel Levy and Fou'iu^, Synthase des Mine'raux et des Roches, Paris, 1882, pp. 87-134 and 161-164. t Compte Rendu de I'Acad. des Sciences, le 30 Juillet, 1883. 168 THE ORIGIN OF CRYSTALLINE ROCKS. [V. '■"111 H.S" ii '': formed by the action of thermal springs below 70° C, is equally produced at 180° C, as in the experiments of De Schulten, and at 400° C. and upwards. § 100. We h.^ve th\i8 far considered among aluminous double silicates those which present the oxygen-ratio of R : al = 1 : 3, and have only mentioned incidentally the epidote and meionite groups. The numerous experiments already detailed suffice to show that the double silicates of alumina and alkalies, formed under very varied condi- tions in the wet way, in the presence of an excess of alkali, always present this ratio, of 1 : 3. When, however, we pass to aluminous double silicates with other protoxyd- bases, we find many with the ratio, 1 : 2, as in the epidote and meionite groups ; with 1 : 1, as in the alumina-garnets, gehlenite, and biotite ; or even 2 : 1, as in melilite, phlogo- pite, and many hydrated alumino-magnesian species of the chlorite group. The genesis of these various calcareous and magnesian alumina-silicates, so conspicuous in the rocks, is an important and unsolved problem. Artificial zeolitic compounds, like the soda-mesolite formed by Way, with the ratio, R : al = 1 : 3, may, as we have seen, exchange their alkaline 'oase for lime or mag- nesia, but for the silicates in cjuesticii, v... »•. ii'.oh this ratio is 1 : 2, or 1:1, or 2 : 1, the coi-respLndiii':- silicates of alumina ani alkalies are as yet unknown to chemistry, being soluble, and probably unstable and uncrystallizable. Analogy, however, as well as the modes of occurrence of these calcareous and magnesian silicates, would lead us to expect the production of such alkaline double silicates, under certain conditions, in solution, and we are not without evidence of the occurrence of such compounds. The soluble alkaline extract from the decomposition of an al \imipous glass, in Daubr(!e's experiment (§ 87), holding in solution I oth silica and alumina, g|ive, if the data are eycyo.t the oxygen-ratio for R : al : Si = 3 : 1 : 4. We lavt) also, in Friedel an- Sarrasin's experiment (§ 99), t iv gep iri,tion <: f anaicite from a like solution, which re- § * t v.] THE CRENITIC HYPOTHESIS. 169 tained both silica and alumina in solution. Researches in this direction will probably make known to us the condi- tions under which such residual solutions may be pro- duced, containing alkalino-aluminous silicates with the ratios corresponding to epidote, garnet, biotite, phlogopite, and the chlorites. § 101. Magnesian silicates corresponding to the zeolitic and feldspar group are rare, and known to us only through the artificial compound of Way, the species iolite, esmar- kite, and fahlunite, and certain partially magnesian zeo" lites. Chabiizite, when finely pulverized, according to Eichliorn, exchanges a portion of its lime for potash when digested with a potassium salt, but is very slightly at- tacked by a solution of magnesian chlorid.* The more silicic of these zeolites are apparently indifferent to such substitutions and, as we have seen, phillipsite is formed in sea-water. We should, however, expect the more basic of the calcareo-aluminous silicates, Avith the ratios, R : al = 1 : 1 or 2 : 1, to be very susceptible to replacement by magnesia. Bunsen has shown that palagonite, a hydrous silicate of this class (§ 67, footnote), with a large propor- tion of calcareous base, decomposes even a solution of ferrous sulphate, which removes its lime, and it would doubtless decompose in a like manner magnesian salts. I have long since shown that an artificial hydrous silicate of lime readily decomposes a solution of magnesium- chlorid, with the production of calcium-chlorid and a magnesian silicate ; a result in accordance Avith the earlier observations of Bischof on the power of solutions of sili- cate of lime to decompose magnesian salts, f § 102. While on one side of what we may call the normal type of alumina-protoxyd silicates, with the ratio, R : al = 1 : 3, as seen in the group of the feldspars and the zeolites, we have those with an excess of protoxyds (including scapolites, epidote, garnet, idocrase, melilite, * Cited by S. W. Johnson, Amer. Jour. Sci., 1859, xxviii., 14. t Hunt, Chem. and Cicol. Essays, p. 122. i I i *i "'^^V'.u' m 160 THE ORIGIX OF CRYSTALLINE ROCKS. [V. I I-: 1^ .J gehlenite, biotite, phlogopite, and the chlorites), there is another series of aluminous silicates in which the propor- tion of protoxyds falls below this normal ratio, and still another series in which protoxyd-bases are absent. Of the latter we need only name the anhydrous species, andalusite, fibrolite, and cyanite, and the hydrous species, pyrophyl- lite, pholerite, and kaoliiiite, with the amorphous halloy- site, a more highly hydrated and colloidal form of the kaolin-silicate, and others. The aluminous protoxyd-sili- cates with a diminished proportion of alkali, constitute an important group, including most of the tourmalines and the pj'incipal non-maguesian micas, muscovite, margaro- dite, euphyllite, damourite or sericite, and paragonite, but excluding the rarer lepidolito of veinstones, which is more highly alkaliferous. In the following list, the formulas for the last four species named have been taken from Dana's "Systcn of Mineralogy," while the three given for different varieties of muscovite have ijcen devised so ■ as to facilitate comparison with the latter, and at the same time to represent, as near as may be, the variable composition of the anhydrous mica. NON-MAG]SrEST.\N OR MUSCOVITIC MICAS. Muscovite (a, Muscovite (b) >fiisco.tke (c^ Miin.-arodite . Damoo-ito . ParagonUs; ., R : r : Si : 11 9 9 9 9 9 12 12 § lOS, The free oient occurrence of muscovite in endo- genou;-. granitic veins with orthoclase and albite, shows that Lhis species, like the feldspars, may be crystallized from solutions. At the same time, their composition and groud like tjf theles T.J THE CRENTTIO HYPOTHESIS. 161 their geological relations suggest thot this and the related micas have more generally beea cleri >'ed, directly or indi- rectly, from the sub-aerial decay of the feldspar of granitic rocks. While these micas are rare, or altogether absent from the oldest granitoid gneisses, they become compara- tively abundant in the younger gneisses and their asso- ciated mica-schists, and, finally, in the forms of damourite, sericite, and paragonite-schists, characterize great masses of strata among the still younger Transition strata. We have called attention to the fact that decayed feldspars, already changed to the form of clay, and approaching to the kaolin-ratio, in which al : Si = 3 : 4, still retain, in many cases, a few hundredths of alkali (§ 63) ; while the three anhydrous silicates of alumina, — andalusite, fibro- lite, and cyanite, — which are frequently found crystallized in certain mica-schists, have each the ratio, 3:2. It will be readily seen that the separation of these highly alumi- nous silicates from clays still holding a little alkali would leave residues having essentially the composition of the micas given in the above table. There are, however, other mica-schists which are not accompanied by such anhydrous aluminous silicates, but on the contrary are associated with serpentines and chloritic minerals, indica- ting in the waters of the time a very different condition from that which we have first supposed, and pointing to the intervention of soluble silicates. That these, b}' their union with the kaolin from decayed feldspars, might yield muscovitic micas, will be evident, when we note that the elements of one equivalent of kaolinite united with one of thomsonite, or of natrolite, would give essentially the oxygen-ratio of muscovite or margarodite, and two of kaolinite with one of thomsonite that of damourite or paragonite. * § 103 A. [The tourmalines constitute an important group of double aluminous silicates, which, though very un- like the muscovitic micas in physical characters, are never- theless, as long since pointed out by Rammelsberg, closely 1^ i^ '""..ii . nriGiK or cbystALLOT bocks. tV. •ifto THE oraciJ* ^^ _ •^^ . • A nresent sinii- .eUtea to «,e™ in e— ^ »»po^^^^^^^^^^^^^ Ld the pro 1- va.^ing veUt,7 « ^^„,^ „, ^.e to— ^e/s tnxvd-bases. int e^ analyses ot spti.ii" tS chemtat in 1850, based on the a J ^ ,,tirfactory So^ thirty localities, gave nm he d ^,^^^^ ^^, ^^^^ dassu.eatio„ of «--»- /ly be added as a s.xtt. tho red tourmaline of ^°^" ^ ^„„,,„, a considerable In ot these cont^n as is ^v 1 ded though varying ™'°™* °* ^ portion oi silica. These by lUininelsberg »' "^P^^'^ ^.^a'^aliUe by the nature oi /ve divisions are d'^tu guisl^ ^^^^^^^ ^^^^^.^ ^ , „£ their inotoxyd-bases, ana uy ^^^^^^ yellow, or pvotoxyd, ses ^^ ^^^ ^,_^,^ „£ .o^o. coronite, has the ratios,! .8 . &, ^,„y ^^ called The black fervo-inagne-an^P - ^^^^^^^ species, aphr. sehorlite, gives 1; * • 6. the ^^^^^j^^^ i ., 12 15 7ite li6:8; in^ ^^^^J^t^r -^ — of the various tour- constitution and tne a t>i THE CRENITIC HYPOTHESIS. 163 malines will be found discussed at length in Essay VIII., §§ 85, 86.] § 104. There exists an important class of hydrous alkaline aluminous silicates, related to the muscovitic micas in composition, but differing widely from them in structure and physical characters. It includes what has been variously designated as pinite, gieseckite, agalmato- lite, and dysyntribite, which sometimes occur in crystal- line forms in other rocks, and at other times themselves constitute rcck-masses. Amorphous, and granular or compact in texture, its hardness and general aspect liave often led observers to compare it to serpentine. The many varieties of this substance, as D:.iia has remarked, agree closely in physical characters, as well as in composi- tion, and he has deduced from their analyses a formula corresponding to a hydrous silicate of potash and alumina, with the ratios, 1 : 8 : 12 : 3, which requires potash 12.0, alumina 35.1, silica 46.0, water 6.9 = 100; in which the po.tash may be partially replaced by soda, lime, or magne- sia. Dysyntribite, as first described by C. U. Shepard, forms rock-masses, associated with specular hematite, in St. Lawrence county. New York ; and similar deposits, often of considerable extent, occur in the crystalline schists of the Green Mountain range, both in Vermont and Quebec. In the latter province, a bed of it in Stan- stead, interstratified with chloritic schists, is one hundred and fifty feet wide, schistose, and often with an admixture of quartz. Layers of the pure pinite from this deposit, formerly described by the writer under the synonym of agalmatolite, have a banded structure, a ligneous aspect, and a satiny lustre. The mineral is translucent, soft, unctuous, and somewhat resembles steatite. A similar deposit occurs in argiliite, among the crystalline schists of St. Francis, Beauce, which is honey-yellow in color, and granular in texture. The pinites from these two locali- ties agree closely in composition. That of the latter con- tained silica 50.50, alumina 33.40, magnesia 1.00, potash I nr i;:!l (.ill; I 164 THE ORIGIN OF CRYSTALLINE ROCKS. tv. 8.10, soda 0.G3, water 5.36 (with traces of lime and iron- oxyd) = 98.99. These elements give almost exactly the oxygen-ratio of 1:8: 13| : 2^, closely agreeing with Dana's formula, except in an excess of silica, perhaps due to an admixture of quartz, which is apparent in the deposit at Stanstead.* I'he variety of pinite formerly described by the writer as parophite, from its resemblance to serpentine, occurs in uncrystalline Cambrian shales at St. Nicholas, near Quebec, f Related to pinite are the minerals which have been called onkosine and oosite. The name of cossaite has been given to a similar min- eral having the physical characters of pinite, from which it differs in containing soda instead of potash. The formula which has been deduced from its analysis, is identical with that of the soda-mica, paragonite. We cannot be certain, in the case of massive minerals like these, whether this same general formula is not as well adapted to pinite as that proposed above. In any case, it is evident that we have in the pinitic group a widely dis- tributed class of natural silicates, not less important than the muscovitic group, and probably similar in origin. § 105. The constancy in composition and the wide * See, for an account of these various forms of pinite, there described AS agalniatolite, the Geology of Canada, 1863, pp. 484, 485. t There are several other hydrous silicates of alumina, sometimes with alkali, which, like pinite, are sometimes found among uncrystalline strata, showing that the conditions of their deposition have been con- tinued down to comparatively recent times. Such is the bravaisite de- scribed by Mallard, a soft unctuous matter, with a fibrous texture, occur- ring in layers in shales of the coal-measures in France. It is a hydrous silicate of potash and alumina, with a little lime and magnesia, and according to its author, after deducting impurities, gives essentially the ratios, 1:3:9:4. The hygrophilite of Laspeyres is also a soft, unctuous, cryptocrystalline matter, found in sandstone, which somewhat resembles bravaisite, and is compared to pinite. It contains potash and some soda, and gives the ratios, 1:5:9:3. A somewhat similar substance, found re- placing coal-plants in the Tarentaise, has been also referred to pinite or to the so-called giimbellite. Genth, on the other hand, found pyrophyl- lite replacing the substance of coal-plants in Pennsylvania. (See Dana's System of Mineralogy, Supplements, I., 6; II. ,29, 63; and III., 18, 54). Also farther, Essay Y I. , respecting bamelite, glauconite, and related sili- cates. are ▼.J THE CUENITIO HYPOTHESIS. 165 distribution of pinite show it to be a compound readily formed and of great stability. Such being its character, it might be expected to occur as a frequent jjrodiict of the aqueous changes of other and les table silicates. It is met with in veinstones, in the shape of crystals of neplie- lite, iolite, scapolite, feldspars, and spoduniene, from each of which it is supposed to have been formed b}- epigene- sis. Its frequent occurrence as an epigenic product is one of the many examples to be met with in the mineral king- dom of the law of " the survival of the fittest." It is, however, difficult to assign such an origin to beds of this mineral like those which have been above described, which are probably the results of original depojition or of diagene^'o. It is (?. characteristic of our present unnatural system of mineralogy to banish to the category of doubtful species most of the substances which are sipposed to be of epigenic origin, and which do not ordinarily present a definite crystalline structure. Several mineral compounds are apparently indispcjsed to assume a crystalline condi- tion, and among these are pinite and serpentine. The latter is probably, like pinite, in certain cases, a product .of epigenesis; but few, we think, who have studied the mode of its occurrence and distribution in crystalline limestones, will ascribe to it, in such conditions, an epi- genic origin. § 106. Dana has compared serpentine and pinite on the ground of their phj'sical resemblances, and has said that pinite is "an alkali-alumina-serpentine, as pyrophyllite is an alumina-talc."* The relations between the minerals thus compared are, however, mimetic only and not genetic. A true system of mineralogical classification must not be based on analogies such as these, nor on assumptions regarding water as replacing fixed bases, or alumina as taking the place on the one hand of silica or on the other of protoxyd-bases. Some of the relations suggested by formulas constructed in accordance with ♦ Dana's System of Mineralogy, 5tli ed., p, 479. 'J 4 a n* %* ^. V^.'^'v?- IMAGE EVALUATION TEST TARGET (MT-3) // ^ /. &•> f/^ 1.0 I.I |2j8 |25 2.2 Hr 1^ IIIIIM - 6" 11.25 11.4 IIIIII.6 V] sms^ S ""^^'^ ' ,esian carbonate. In of the well-known terhy-l ate'l »»? trfturating .n a iL manner, the ^^'j '"rids ^f calcium and magnesium, mortar a solution of eUo^^s ot c ^^^^^^^ j , i„ equivalent P™Pf*'"* :'tr„f o^a- 's, at a tempc^ture solution of neutral carbonate oi ^^^^^^^^ j„to an S"rom 65' to 80° C. » -"^ed, a^ ^^^^^ ^^ ^ aggregate of translucent ovystaU^ hJ,odolomite of Von dfuble carbonate, vesembln g *be hy ^^ ^^, ^^ ^ ^^^^^ KobeU. At tempera uieot ^^^^^^ . magma changes slowly u to am ,^^ ^, i,es from pound. The process °« f »f;„ed "to consist m the Welvc to twenty-five days app^^^.^^^.^^ ^^^^^^^^ formation of nuclei «"» f '^^ voluminous, opaque, and until every particle «« ^^^^^^^ translucent, dense, and amorphous precipitate had b 'Jom ^^ ^^^^^^^ ^^^^^ crystalline." The product is made P ^^^ ^^^^. apparently oblique, grouped — ;„ dimeter. Zes forming spheres 6™ °"'^ ^j u^e and magnesia, ■ The hydrated double ^^^^^\^,,,, of carbonate of thus formed ^ presenc oJ a si g ^^^ ^^„, „( ,u soda, was found to »»"*» ^^^^ether this did not proceed latter, but it wa^ "»* f'^Ja^us double carbona e of from an admixture of the hya ,„„bination itself Ume and soda, Sf "t^;^'! Jeomposition of a gayluss'te was described as having «»»P ^^^ production in which magnesium leplaees v.] THE CRENITIC HYPOTHESIS. 171 of crystals of true gaylussite, as observed by Fiitzsche, by the slow crystallization of the gelatinous precipitate got when a strong solution of carbonate of soda in excess is mingled with one of calcium-chlorid, is another remark- able example of the phenomenon under consideration. Fritzsche moreover observed that it is not necessary that the lime-carbonate should be in its gelatinous form in order to produce this compounl, since the previously precipitated carbonate, when digested with a solution of carbonate of soda, slowly combines with it to form the crystalline hydrous double salt. J/Iore remarkable still is the observation of H. Ste.-Claive Deville, which I have repeatedly verified, that a paste of magnesia alba and bi- carbonate of soda, with water, is slowly changed, at a temperature of from 60° to 70° C, into a transparent crystalline anhydrous double carbonate of lime and soda, rhombohedral in form, and called by its discoverer a soda- dolomite.* § 113. In this connection, it should be said that we have here an explanation of the formation of the double carbonate of lime and magnesia which constitutes ordi- nary dolomite. The origin of this mineral ^ecies, which so often constitutes rock-masses, is still generally misun- derstood. The baseless notion of its production by a metasomatosis or partial replacement of the lime in ordi- nary limestone, imagined by the older geologists, is still repeated, and holds its place in the literature of the sci- ence despite the facts of geognosy and of chemistry. I have long since shown, bj'^ multiplied examples, that the ordinary mode of the occurrence of dolomite in nature is not in accordance with this hypothesis of its origin, since beds of dolomite, or more or less magnesian limestone, are found alternating, sometimes in thin and repeated layers, with beds of non-magnesian carbonate of lime. Moreover, beds of crystalline dolomite, conglomerate in * Hunt, Contributions to the History of Lime and Magnesia Salts; Part II., 1866. Amer. Jour. Science, vol. xlii., pp. 54-57. 172 THE ORIGIN OF CRYSTALLINE ROCKS. m lip' character, are found to enclose pebbles and fragments of pure non-magnesian carbonate of lime. I have also ex- plained at length the natural reactions by which precipi- tates consisting of a greater or less proportion of hydrous carbonate of magnesia, mixed with carbonate of lime, must, in past ages, have been laid down in the waters of lakes and inland seas, in some cases with, and in others without, the simultaneous formation of sulphate of lime. It was, moreover, found that the reaction at an elevated temperature in presence of water, between sulphate of magnesia and an excess of carbonate of lime, supposed by Haidinger and Von Morlot to explain the frequent asso- ciation of gypsum and dolomite, does not yield the double carbonate, since the carbonate of magnesia separates in an anhydrous form, and does not combine with the carbonate of lime. Finally, it was shown that mixtures of hydrous carbonate of magnesia and carbonate of lime, when heated ^. :ether in presence of water, unite to form the anhydrous double carbonate which constitutes dolomite. In my ex- periments, their combination, with the formation of dolo- mite, was effected rapidly, at 120" C, but many consid- erations lead' to the conclusion that its production in nature is effected slowly at much lower temperatures, and that the formation of the hydrous double carbonate already described is, perhaps, an intermediate stage in the pro- cess.* The existence of a soluble and active form of mag- nesian carbonate, as described in § 110, throws an addi- tional light upon the formation of dolomite. § 114. The reactions described in the preceding para- graphs between the elements of comparatively insoluble substances in the presence of water, resulting not only in the conversion of amorphous into crystalline bodies, but in the breaking-up of old combinations, as well as in the union of unlike matters mechanically mingled t form * Hunt, Contributions to the Ciiemistry of Lime and Magnesia, part i., 1859, Amer. Jour. Sci., xxviii., pp. 170, 365; and part ii., 1806, ibid,, vol. xii., p, 49; also in abstract in Cheiu. and Geol. Essays, pp. 80-92. . is thi grain^ chemj takes tallini v.] THE CRENITIC HYPOTHESIS. 178 new crystalline species, are instructive examples of what Giimbel has termed diagenesis. The changes in the masonry of the old Roman baths in contact with thermal waters, resulting in the hydration of the substance of the bricks, and its conversion into zeolitic minerals; the hydration of volcanic glasses with similar results, going on, even at low temperatures, in the deep sea ; the decom- position of common glass by heated water ; the conversion of basaltic rock into palagonite, and the production there- from of zeolites; the similar changes seen elsewhere in amygdaloids, and even in massive basic plutonic rocks, are also examples of this process of diagenesis, and serve to show its great geological significance. We have already suggested the intervention of similar reactions in past ages among the sediments from the sub-aerial decay of feldspathic rocks, in some cases with the concurrence of the secretions from the primary basic stratum, which, in accordance with the crenitic hypothesis, we suppose to have been the source of soluble mineral silicates. In the diagenesis of these early argillaceous sediments, aided by crenitic action, will, it is believed, be found the origin of many of the crystalline schists of the Transition rocks. § 115. An instructive phase in this diagenetic process is that of the gradual conversion of smaller crystalline grains or crystals into larger ones, which is familiar to chemists. This action is in fact nearly akin to that which takes place in the transformation of amorphous into crys- talline precipitates, since in both cases a partial solution precedes the crystallization. It is well known that, as a result of successive solution and redeposition, large crys- tals may be built up at the expense of smaller ones. To quote the author's language of fifteen years since, this process, as H. Deville has shown, " suffices, under the in- fluence of the changing temperature of the seasons, to convert many fine precipitates into crystalline aggregates, by the aid of liquids of slight solvent powers. A similar agency may be supposed to have effected the crystalliza- :^K "|, pB|| K^ ;. !' r-' 174 THE ORIGIN OF CRYSTALLINE ROCKS. tion of buried sediments, and changes in the solvent power of the permeating water might be due either to variations of temperature or of pressure. Simultaneously with tliis process, one of chemical union of heterogeneous elements may go on, and in this way, for example, we may suppose that the carbcriates of lime and magnesia become united to form dolomite or magnesian limestone." * § 116. The tendency of the dissolved material in this process to crystallize around nuclei of its own kind, rather than on foreign particles, is a familiar fact, and its geolog- . ical importance, to which I first called attention, as above, in 1869, was again pointed out by Sorby in 1880, when he showed that dissolved quartz might be deposited upon clastic grains of this mineral in perfect optical and crys- tallographic continuity, so that each broken fragment of quartz is changed into a definite crystal, as was seen in his microscopic studies of various sandstones.f Tliis fact has been confirmed by the observations of Young, Irving, and Wadsworth in the United States ; J and Bonney has suggested the possible extension of such a process to feld- spar, hornblende, and other minerals. § Vanhise has very recently announced that his micro- scopical examinations of certain sandstones of the Kewee- nian series, from Lake Superior, afford evidence of the secondary deposition of both orthoclase and plagioclase feldspar, in crystallographic continuity, upon broken feld- spathic grains, in one case uniting the two parts of p broken feldspar-crystal. The sandstones which have yielded these examples are made up in part of feldspathic fragments, and in part of fragments of " some altered basic rocks." They are, moreover, interstratified with and, in * Hunt, The Chemistry of the Earth, Report of Smithsonian Institu- tion, 1869; also Chem. and Geol. Essays, p. 306. t Sorby, Presidential Address, Quar. Jour. Geo. Soc. London, xxxvi., 83. t Young, Arner. Jour. Sci., xxiv., 47. Irving, ibid., xxv., 401. Wads- worth, Proc. Boston Soc. Natural History, Feb. 7, 1883. § Bonney, Quar. Jour. Geol. Soc, xxxix., 19. T.1 THE CRENITIC HYPOTHESIS. 175 some cases at least, immediately underlie the basic plu- tonic rocks of the same Keweenian series.* When we consider that orthoclase is a common secretion of these basic rocks, as is shown by its frequent occurrence in them with zeolites and epidote, it may perhaps be ques- tioned whether the secondary feldspar in the sandstone has been derived from the adjacent grains of this mineral, or has come into solution from the transformation of the basic rocks. The apparent stability and insolubility of orthoclase and oligoclase at high temperatures in the pres- ence of water, as observed by Daubr6e, wouhl seem to favor the latter view. In any case, it is a striking illus- tration of the tendency of mineral species to crystallize around nuclei of their own kind, which is so marked a factor in the development of the crystalline rocks. IV. ■CONCLUSIONS. § 117. We reviewed in the first part of this essay the history of the different hypotheses hitherto proposed to explain the origin of the crystalline rocks, and, in doing so, reached the conclusion that not one of them affords an adequate solution of the various problems presented by the chemical, mineralogical, and geognostical characters of the rocks in question ; at the same time, we endeavored to show succinctly what are the principal conditions to which a satisfactory hypothesis must conform. In the second part, we sketched the growth and development, during the last quarter of a century, of what we believe to be such a hypothesis. In the third part, we sought to bring together a great number of facts, both new and old, which serve to illustrate the new hypothesis ; according to which the crystalline stratiform rocks, as well as many erupted rocks, are supposed to have been derived by the action of waters from a primary superficial layer, regarded as the last portion of the globe solidified in cooling from a state of igneous fluidity. This, which we have described * Vanhise, Amer. Jo'ir. Sci., 1884, xxvii., 399. Ml 176 THE OUIOIN OF CIIYSTALLINE ROCKS. [V. .« as a basic, quartzless rock, is conceived to liave been fis- sured and rendered porous during crystallization and refrigeration, and thereby made permeable, to consider- able depths, to the waters subsec^uently precipitated upon it. Its surface being cooled by radiation while its base reposed upon a heated solid interior, upward and down- ward currents would establish a system of aqueous circu- lation in the mass, to which its porous but unstratified condition would be very favorable. The materials wiiich heated subterraneous waters would bring to the surface, there to be deposited, would be not unlike those which have been removed by infiltrating waters in various sub- sequent geological ages, from erupted masses of similar basic rock ; which, we have reason to believe, are but dis- placed portions of this same primary layer. The mineral species removed from these latter rocks, or segregated in their cavities, are, as is well known, chiefly silica in the form of quartz, silicates of lime and alkalies, and certain double silicates of these bases with alumina, including zeolites and feldspars, besides oxyds of iron and carbonate of lime ; the latter species being due to the intervention of atmospheric carbonic acid. The absence from these minerals of any considerable proportion of iron-silicate, and, save in rare and exceptional conditions, of magnesia, is a significant fact in the history of the secretions from basic rocks, the transformation of which, under the action of permeating waters, has resulted in the conversion of the dissolved portion of the material into quartz and vari- ous silicates of alumina, lime, and alkalies, while leaving behind a more basic and insoluble residue abounding in silicated compounds of magnesia and iron-oxyd with alumina. § 11 8. The peculiarities resulting from this comparative insolubility of magnesian silicates long ago attracted the attention of the writer. The addition to solutions like sea-water, of bicarbonate of magnesia, which is a product of the sub-aerial decay of basic rocks, would, it was magn( the atl renio) SOJid II therefi ^e suj magnej Would I *Ar P- 122. > TO THE CRENITIC HYPOTHESIS. 177 shown, effect a separation of dissolved lime-salts in tho form of carbonate, leaving the nmgiie.sia in solution as chlorid or as 8ul[)hate; while on the contrary tho reaction of such a natural water with certain silicates, whether solid or in solution, containing lime or alkalies, would effect a removal of the dissolved magnesia. At the same time it was shown that "by digestion at ordinary temper- atures with an excess of freshly precipitated silicate of lime, chlorid of magnesium is completely decomposed, an insoluble silicate of magnesia being formed, while nothing but chlorid of calcium remains in solution. It is clear that the greater insolubility of the magnesian silicate, as compared with silicate of lime, determines a reaction the very reverse of that produced by carbomites with solu- tions of the two earthy bases. In the one case, the lime is separjited as carbonate, the magnesia remaining in solu- tion, while in the other, by the action of silicate of soda, or of lime, the magnesia is removed and the lime remains. Hence carbonate of lime and silicate of magnesia are found abundantly in nature, while carbonate of magnesia and silicate of lime are produced only under local and exceptional circumstances. It is evident that the produc- tion from the waters of the early seas of beds of sepiolite, talc, serpentine, and other rocks in which a magnesian silicate abounds, must, in closed basins, have given rise to waters in which chlorid of calcium would predominate." * § 119. From this reaction it would follow that the magnesian salts formed when the first acid wa.jrs from the atmosphere fell upon the primary stratum, would be removed from solution, either by the direct action of the solid rock, or by that of the pectolitic secretions derived therefrom in the earliest ages. The primeval ocean, if, as we suppose, a universal one, would soon be deprived of magnesian salts, and henceforth the early-deposited rocks would be essentially granitic in composition, and non- * Amer. Jour. Sci., 1865, vol. zl., p. 49; also Chem. and Geol. Essays, p. 122. 178 THE ORIGIN OF CRYSTALLINE SOCKS. [V. \('i- ■• t, magnesian, until the introduction of magnesia into its waters from an exterior source. The pectolitic silicates themselves, which, in the cavities of exotic basic rocks, are deposited in crystalline forms, would, if set free in a sea deprived of magnesian salts, be readily decomposed by the carbonic acid every where present, with separation of free silica and carbonate of lime. From this would be formed the first deposits of limestone, which make their appearance in the old gneissic rocks and become mingled with magnesian carbonate and silicates from tiie introduction of magnesian salts into the waters. The comparative instability of the lime-silicate is seen when wolhistonite is compared with the corresponding silicates, pyroxene and enstatite. It is possible, notwith- standing the absence of magnesian species from zeolitic secretions, that, under certain conditions, small portions of magnesian silicate mny liave been included in the early crciiitic deposits, but the rarity of such magnesian silicates in these, and their abundance in parts of the later Lauren- tian and in younger deposits, point to a new source of the magnesian element, namely, the extrusion of portions of the underlying plutonic mass, and its sub-aerial decay. It would be instructive to consider in this relation the gradual removal of a large proi)ortion of silica from the primary plutonic stratum in the forms of orthoclase, albite, and quartz, and the consequent partial exhaustion of por- tions of this underlying mass, so that its succ eding secretions should consist 'chiefly of less silicic silicates, such as labradorite and an .'esite, without quartz, as in the Norian series. § 120. The conditions of this first exoplutonic action cannot be fully understood until Ave have settled the ques- tion of the permanence of continental and oceanic areas, and the exter' of the early crenitic rocks which constitute the fundamental granites and the granitoid gneisses. Whether these are spread, with their vast thickness, alike underneath the great areas of tne paleozoic series and our <♦ ▼.] THE CRENITIC HYPOTHESIS. 173 )ur modern oceanic basins, — in brief, whether or not they are universal, as supposed by Werner, is a question wliich cannot here be discussed. There is, however, nothing in- compatible with what we know of the chemistry of the early rocks and the early ocean in the supposition that they were universal, since there is apparently no evidence that the products of sub-aerial decay of exposed rocks in- tervened in their production. Such a condition of things was, however, necessarily self-limited; the progressive diminution in volume of the primary plutonic stratum from the constant removal of portions of it in a state of solution, and the weight of the superincumbent accumu- lated granitic and gneissic material, could not fail to result in widely spread and repeated corrugations and foldings of the overlying mass, the effects of which are seen in the universally wrinkled and frequently vertical attitude of the oldest gneissic rocks. Such a process, like the similar though less considerable movements in later times, would probably be attended v/ith outflows, in the form of fissure-eruptions, of the underlying basic stratum, which, in accordance wich our hypothesis, was permeated with water under conditions of temperature and pressure that must have given to it a partial liquidity. Such a process of collapse and corrugation of the crenitic deposits, attended with extravasation of the underlying plutonic stratum, wou^d doubtless be often repeated in these early periods, resulting in frequfint stratigraphical discordances, which are, however, in all cases to be looked upon as local accidents, and not as wide-spread catastrophes. Hence the appearance, from time to time, of oxoplutonic masses, with upliftings and depressions of the crenitic rocks, which caused the exposure of both alike to the action of the atmosphere. § 121. The consequent sub-aerial decay of these two types henceforth introduced new factors into the rock- forming processes of the time, anJ :^ade the beginning of what Werner called the Transition period. The decompo- \^ fS 'M t' :' ■ 'lii ' I M^iij^ maadi 180 I' \ ■ ■ - THE ORIGIN OF CRYSTALLINE ROCKS. [V. sition of these, under the influence of a moist atmosphere holding carbonic acid, resulted in the more or less com- plete removal of the alkali from the feldspars of crenitic rocks, and their conversion into kaolin, while the corre- sponding changes in the basic exoplutonic rocks were still more noteworthy. These rocks, while containing feld- spars, consisted in large part of silicates of lime and mag- nesia, presumably pyroxene and chrysolite, which, as we are aware, yield tc the action of the atmosphere the whole of their lime and magnesia. These, in the form of car- bonates, passed into soluiioD, together with a large propor- tion of silica, leaving beli.'ud the remaining portion, together with iron-oxyd and the kaolin from the feldspars. The carbonates of alkalies, of lime, and of magnesia, re- sulting from the sub-aerial decay of the exposed exoplu- tonic and the crenitic rocks alike, were carried to the sea, there to play an important part. Besides the direct influx of carbonate of lime into the waters of that time, it is evident that both the alkaline and the magnesian bicar- bonates would react upon the calcium-chlorid of the primeval sea, with the production of a farther amount of lime-carbonate, and the generation of alkaline and mag- nesian chlorids. In this way, the sea becoming magne- sian, a new order of things was established. Henceforth, the pectolitic matters brought up fruui the primary layer would at once react upon the dissolved magnesian salts, and the production of such compounds as chondrodii-e, chrysolite, serpentine, and talc would commence. No one who has studied the mode of occurrence of these silicp'eo in the upper part of the Laurontian series, where serpentine not only forms layers, but frequent concretions like flints, often around nuclei of white pyroxene, can fail to recognize the process which then came into play, result- ing later in the production of abundance of pyroxene, amphibole, and enstatite, and apparently reaching its cul- mination in the vast amount of magnesian silicates found in the deposits of the II:;ronian age. 1 v.] THE CRENITIC HYPOTHESIS. 181 § 122. The solutions of simple silicates' of alkalies, which by heat had deposited their excess of silica in the form of quartz, as in the case of the soluble matter from glass, probably gave rise by their reaction with magnesian solutions to the basic protoxyd-silicates, like chondrodite, chrysolite, serpentine, and pyroxene. That we have no anhydrous quadrisilicates corresponding to apophyllite and okenite is apparently due to the fact that such sili- cates, in contact with water at elevated temperatures, break up into anhydrous bisilicates and quartz ; as is seen in the artificial association of pyroxene and quartz in the experiments of Daubrde, and the frequent occurrence of admixtures of the two in beds ;among the ancient gneissic rocks. A noticeable fact in the history of the surbasic silicates of magnesia and related protoxyd-bases, men- tioned above, is their frequent association with non-sili- cated oxyds. Examples of this familiar to mineralogists are the occurrence of aggregates of chondrodite and magnetite ; of chromite, picotite, ilmenite, and corundum, with chrysolite and serpentine ; and of frankli.iite and zuicite with tephroite and willemite. These collocations are probably connected with the solvent power of solu- tions of alkaline silicates, already insisted upon (§ 89), and also with the dissociation of silicate of alumina in heated alkaline solutions, noticed by H. Deville (§ 98). The separation, by the alternate action of decaying organic matters and of atmospheric oxygen, of iron-oxyd, whica readily passes from a soluble ferrous to an insolu- ble ferric condition., and conversely, has probably played an important part in the formation of deposits of iron- Dxyds, wliich are much more general in their asso- ciations than corundum, or the compounds of chromic, titanic, aluminic, manganic, and zincic oxyds mentioned above, to which we have assigned a very different origin. It will remain for the mineralogist to determine what de- posits of magnetite and of hematite are to be ascribed to the one and what to the otlier origin. J * n 182 THE ORIGIN OP CRYSTALLINE BOCKS. [V. § 123. We have seen, among the secretions of basic rocks, lime-alumina silicates, like epidote and prehnite, in which the ratio of the protoxyd-bases to alumina, instead of being 1 : 3, as in the feldspars and the zeolites, is 1^ : 3 or even 2 : 3. Although, probably on account of their solubility and their instability, we do not know of any natural silicates with a still larger proportion of lime to the alumina, we have indirect evidence of their former existence in solution, in the frequent occurrence of double silicates of magnesia and alumina, in which the oxygen- ratio of R, al, instead of being 1 : 3, >s in the feldspars, or 2 : 3, as in prehnite, becomes 3 : 3 and even 6 : 3, as seen in the magnesian micas and in the chlorites. Such silicates, often with epidote, abound in the rocks of Huronian age. This process by which, through the intervention of sili- cated secretions from the substratum, the magnesian salts are removed from the sea-water, is, as we have shown, the reverse of that which takes place through the action of the carbonates from the sub-aerial decay of silicated rocks precipitating lime-salts and giving rise to magnesian wafers, if not over oceanic areas, at least in inland basins of greater or less extent. Alternations of this kir 1 must have been frequent in geological history, and we have evidence of a widespread phenomenon of this kind fol- lowing the Huronian age, when in seas from which mag- nesian salts were apparently for the most part excluded, were deposited the gneisses and mica-schists of the Mont- alban series. These, in very many places, are found resting directly, often in unconformable superposition, upon the older or Laurentian gneisses, but elsewhere upon the Huronian, showing the intervention of extensive movements of elevation and subsidence, and probably of denudation, subsequent to the Huronian time. § 124. The introduction on a limited scale, into the sea-basins of the Montalban time, of magnesian salts is evident from the occasional appearance of magnesian sili- mus^ marl cess a pi dimii * Amerl mf. of h v.] THE CllENITIC HYPOTHESIS. 183 cates in the Montalban rockf . The most noteworthy fact ill their history is, however, the appearance in this series, with gneisses which differ from those of older times in being finer grained and less granitoid, of deposits contain- ing aluminous silicates characterized by a diminished pro- portion of protoxyd-bases. Such as these are the beds of quartzose schists holding non-magnesian micas and the simple silicates, andalusite, fibrolite, and cyanite. It has already been mentioned that in the formation of these rocks the more or less completely decomposed feldspar from the sub-aerial decay of older crenitic rocks may have been brought into the areas of deposition. Either such clays, still retaining a portion of alkali from undecayed feldspar, or else admixtures of kaolin with the elements of a feldspar or a zeolite might, as has been suggested, yield, by diagenesis, muscovite and quartz, with one of the simple aluminous silicates just named. That a pro- cess of sub-aerial decay was in progress in the Montalban time is shown by the presence in the mica-schists of this series, at several localities in Saxony and elsewhere, as described by Sauer and subsequently noticed by the present writer, of "boulders of decay," having all the appearance of those formed during the atmospheric decay of the older gneisses.* The intervention in the deposits of that period of somewhat basic zeolitic minerals, is shown by the presence in the younger gneissic series of Germany of large masses of so-called dichroite-gneiss or iolite-gneiss, and the occasional occurrence of iolite in the younger or Montalban gneisses of New England. § 125. The predominance of micaci^ous schists of the muscovitic type in the upper portions of the Montalban, marks the growing change in the conditions of the pr- - cess which gave rise to the indigenous crystalline rocks ; a process continued with many modifications, and with diminished energy, through the subsequent period of the * Sauer, in 1879, Zeltschrift f. d. ges. Naturwiss, Band lii; also Hunt, Amer. Jour. Sci., 1883, vol. xxvi., p. 197, and Essay X., § 80. i ill I ! i J m hi,''\ i i m rv Jl'il 184 THE OlilGlN OF CRYSTALLINE ROCKS. rv. Taconian. This was marked by the deposit of quartzites, limestones, and argillites, and also by the intercalation of schistose beds characterized by an abundance of damour- ite or relat3d micaceous minerals, as well as by the pres- ence of matters apparently feldspathic, which seldom take upon themselves the characteis of well defined species, though found transformed by sub-aerial decay into a form of kaolin, and in some instances apparently assuming the state of an imperfect gneiss. These Taconian schists, which require careful chemical and microscopic study, also include serpentine, talc, pyroxene, epidote, and gar- net. The appearance in jjaleozoic argillites of crystals of rutile, of tourmaline, and of staurolite, indicates a later stage of that condition of things which marked the cre- nitic process in pre-paleozoic times, and made possible the formation of the vast series of Primitive and Transi- tion crystalline schists which we have sought to include under the names of Laurentian, Norian, Arvonian, Huro- nian, Montalban, and Taconian — designating in their order the upward succession of these great groups from the fundamental granitoid gneisses (here included in the Laurentian) to the dawn of paleozoic time. The Arvo- nian or petrosilex group intervenes between the Lauren- tian and the Huronian. The peculiar characters of the Norian, and its localization to some few limited areas in Europe and North America, make it difficult for us, as yet, to define its precise relations to the Arvonian. The Norian, however, like the Arvonian, probably occupies a horizon between Laurentian and Huronian. Much time may pass, and many stratigraphical studies must be made, before the precise relations of the Huronian and the suc- ceeding Montalban can be defined. It seems probable, in the present state of our knowledge, that the Montalban series, though of great thickness, was, in manj' cases, de- posited over areas where the Huronian had never been laid down. Notwithstanding the great geographical ex- tent and the importance of these two series, neither can doi call in «l ti ScleJ ^a-^' v.] THE chenitic hypothesis. 185 claim that universality which apparently belonged to the primitive granitic stratum; a universality soon interrupted by the uplifting of portions of dry land, an event which preceded Huronian time. § 126. That the production of large quantities of simi- lar pectolitic silicates, in regions remote from exotic rocks, was continued from the pre-Cambrian to far niore recent times is evident, from the presence of a considerable de- posit of serpentine among the horizontal Silurian dolo- mites of Syracuise, New York, of which the writer hag elsewhere recorded the history,* and also from the well known beds of sepiolite found with opal in the tertiary dolomites of the Paris basin.f The recent amorphous zeo- litic deposits in tertiary sandstone in Switzerland (§ 94), and the compounds referred to on page 164, should not be forgotten in this connection. . Whether the silicates brought from below by crenitic action were directly separated as feldspars, rs crystalline zeolites, or as gelatinous precipitates to be subsequently changed by diagenesis into crystalline hydrous or anhy- drous species, are questions for farther dij.cussion. The range of temperature through which we have noted uhe crystallization of chabazite, and the association of ortho- clase by contemporaneous or subsequent crystallization with hydrous species like zeolites and chlorite, lead us to conclude that for the hydrous and anhydrous aluminous double silicates alike, a considerable range of temperature is permissible. In any case, we lind notliing in the condi- tions of the formation of zeolitic minerals in the past, any more than in modern times, incompatible with the exist- ence of organic life. § 127. The phenomena of exoplutonic action, or so- called vulcanicity, though relegated to a secondary place in the crenitic hypothesis, are yet, as we have said, of • See farther, Essay X., §§ 27-34. t Hunt, on the Dulomites of the Paris Basm, 1860. Science, xxix., p. 284. Amer. Jour. hi 186 THE ORIGIN OF CRYSTALLINE ROCKS. tv. great importance and significance, and are by no means simple. They were, according to our hypothesis, confined in early times to fissure-eruptions of the underlying plu- tonic stratum. This, although in the course of ages it has suffered a gradual change from the ceaseless crenitio action, which has removed from it the elements of the va- rious series of crystalline rocks, including the primitive granitic and gneissic series, probably still retains in the lower portions somewhat of its original constitution. A second phase in the history of exopluton.ic rocks, al- ready foreseen by the Huttonians, here presents itself for our consideration. The more deeply buried portions of the primitive crenitic deposit must themselves have been brought within the influence of the central heat, and, per- meated as they were by water, would have suffered a softei ing which permitted them, as a result of subsequent movements of the crust, to appear again at the earth's surface as exoplutonic or exotic rocks of the trachytic or granitic type. We can hardly suppose the displacement, either of the plutonic stratum, or of the early granitic deposits, to have been attended with the evolution of permanent gases, such as attend modern volcanic eruptions and are to be ascribed to the action of subterranean heat on more re- cent deposits, including carbonates, suli^hates, chlorids, and organic matters. Such materials, when mingled with silicious and argillaceous sediments, and brought by local accumulation and depression within the heated zone, would give rise to the various gases which characterize the volcanic eruptions of recent periods, in which, how- ever, the materials of the underlying plutonic and crenitic layers also apparently intervene. V ' , By thus ascribing a threefold origin to the products of exoplutonic action, it becomes possible to classify and harmonize the apparently discordant phenomena of erup- tive rocks. While the typical basalts and related basic rocks would be derived from the primary plutonic or ig- !♦ /?\j, u v.] THE CKENITIO HYPOTHESIS. 187 neous stratum, and the trachytic and granitic rocks from the earlier crenitic deposits, the more fusible portions of the hater Transition and Secondary strata may liave fur- nished tlieir contingent, not only of gases and vapors, but of lavas and volcanic dust. § 128. The history of the origin of crystalline roclfs is the history of the origin of the mineral species which compose them. The crystalline masses are essentially made up of a few groups of species. Various feldspars, and occasional zeolites, some of which apparently occur as integral parts of roc) :s chiefly feldspathic, form a great central group. On one side of these are the aluminous double silicates, represented by basic species like garnet, epidote, magnesian micas and chlorites, all with an excess of protoxyd-bases ; while on the other hand are the alu- minous double silicates of the muscovitic and pinitic groups, in which the diminished proportion of the pro- toxyd-bases prepares the way to the associated simple aluminous silicates, pyrophyllite, andalusite, cyanite, etc. To these groups must be added the non-aluminous sili- cates, including amiDhibole, pyroxene, enstatite, and chrys- olite, and the hydrous magnesian species, serpentine and talc. Besides these are free silica, generally as quartz, free oxyds, including the spinel and corundum groups, which, together with the carbonates, make up the essen- tial parts of the crystalline rocks. § 129. Rock-masses, and the mineral species which compose them, present variations in time, as we find in tracing the history of the great successive groups of crys- talline strata ; and they moreover show local changes, as seen in different parts of their distribution in the same geologic.il group. As regards the causes of these varia- tions, very much remains to be discovered by the patient collection and recording of facts concerning the associa- tions of mineral species, their artificial production, and their transformations under the influences of fire and water, and of solutions of potassic, sodic, calcareous, and I ! 188 THE ORIGIN OF CRYSTALLINE HOCKS. [V. wm magnesian salts. The instability of silicated compounds of igneous origin in the presence of water and watery- solutions, so widely diffused through nature, is the war- rant for a general aqueous hypothesis; while, on the other hand, the derivation of stable nuneral species, under such influences, from matters of igneous origin justifies ua in assuming for these species an igneous starting-point. Igneous fusion destroys the mineral species of the crys- talline stratified rocks, and brings them back as nearly as possible to the primary undifferentiated material. P'ire is the great destroyer and disorganizer of mineral as well as of organic matter. Subterranean heat in our time, acting upon buried aqueous sediments, destroys carbo- nates, sulphates, and chlorids, with the evolution of acidic gases and the generation of basic silicates, and thus repeats in miniature the conditions of the ante-nep- tunian chaos, with its surrounding acidic atmosphere. On the other hand, each mass of cooling igneous rock in contact with water begins anew the formative process. The hydrated amorphous i)roduct, palagonite, is, if we may be allowed the expression, a soH of silicated proto- plasm, and by its differentiation yields to the solvent action of water the crystalline silicates which are the con- stituent elements of the crenitic rocks, leaving, at the same time, a more basic residuum, abounding in magnesia and iron-oxyd, and soluble, not by crenitic, but by sub- aerial action. Palagonite, or some amorphous matter re- sembling it, probably marks a stage in the sub-aqueous transformation of all igneous rocks, though only under special conditions does this unstable, hydrous substance form appreciable masses. In all cases, igneous matter, of primary or of secondary origin, serves as the point of de- parture. According to the proposed hypothesis, which derives rocks of the granitic type, composed essentially of quartz and feldspars, by aqueous secretion from a primary igne- ous and quartzless mass, it would follow that the highly v.] THE CUKNITIC IIVl'OTHESIS. 189 basic compound, assumed by Bunsen to represent the typi- cal pyroxenic or basaltic rock (§ 24), would be the above mentioned insoluble residuum; and that less basic varie- ties of similar rocks would correspond to portions of the same primary plutonic mass, less completely exhausted by lixiviation, or modified by partial separation through crystallization and eliquation, as will be explained in Essay VI., and consequently approacliing in composition to admixtures of the basaltic and granitic types, as main- tained on other grounds by Bunsen himself. § 130. The principles which have been enumerated in the preceding pages, will, it is believed, lead the way, not only to a natural system of mineralogy, but to a natu- ral system of classification of crystalline rocks, considered with regard alike to their chemical composition, their genesis, and their geological succession. A valid hypothe- sis for the crystalline rocks must seek to connect all the known facts of their history, by alleging a true and suf- ficient cause for the production of their various constitu- ent mineral species. Such a hypothesis will violate no established principles in chemistry or in physics, but will show itself to be in accord with them all, and will com- mend itself to the acceptance of those who take the pains to understand it. The crenitic hypothesis set forth in these pages is the result of many years of patient study applied to the elu- cidation of a great problem ; and as such is offered to chemists and mineralogists as a first attempt at a rational explanation of the fundamental questions presented by the history of the crystalline rocks of the earth's crust. I m: VI. THE GEXETIC HISTORY OP CRTSTALLINE ROCKS. Thii Kwiay wai prpsentod to tho Roynl Society of Canada at Iti meeting In Ottawa, May, 18H0, iind will appear In ItH TransactionH, vol. Iv., with the above title. In May, 188S, there wan read before the Banie iociety a paper In which the phenomena of (tratltlcatlon In endogenous velnntoMfH and In eruptud rocks were discussed In rela- tion to'tbecrenlttc process, and to the hypothesis of ellquatlon. Of this paper, which was published only In abstract In the Canadian Ilecord of Science, under the title of The Geognosy of Crystalline Uooks, the present essay It but an extensloii and a development. I. § 1. In a preceeding essay on The Origin of Crystalline Rocks, we have considered at length the different views hitherto maintained as to the mode of their production, and have set forth what we have called the crenitic hy- pothesis. It is proposed in the following pages to ex- amine still farther the new hypothesis in some of its aspects, to show how far the conception of a single con- solidated igneous mass under the combined action of water and heat may be made to explain satisfactorily the various facts in the history of the earth's crystalline crust, and thus to reconcile many of the contradictions which still divide the geological world as to the relations of stratified and massive crystalline rocks. Hence the title of the present essay. Of the great divisions adopted by the Wernerian school in geology, those of Primary and Secondary correspond respectively to Original and Derived rocks, and were sup- posed to represent earlier and later periods in geologic time ; the name of Transition being applied to the rocks of an intermediate period, believed to mark the passage from the conditions of the primary to those of the second- ary age. The name of Tertiary given to the rocks of a 190 n.l GENETIC ITrSTOTlY OF CRYflTALLINR ROCKS. 191 still lator ago and, marking? a subsequent period in the process of derivation, needs no explaiuition. Hy tlio geol- ogists of the Huttonian school tlio rocks called primary or original by the Werncrians were imagined to bo in many, if not in all cases, secondary or derivcil rocks, the materials of which, got from the disintegration of pre- existing masses, had been arranged by water, and subse- quently transformed by combined mechanical and chemi- cal agencies into their present crystalline condition; in accordance with which hypothesis they have been called Metamorphic rocks. By rejecting, as their master Mutton had done, all "incjuiry into the first origin of things," or "the commencement or termination of the present order," and by teaching that tha rocks called by Wernerians primary and transition, were for the most part, if not wholly, metamorphosed portions of derived rocks which, themselves, in their prolongation into other regions, could be recognized as secondary or as tertiary strata, the Huttonians have sought to destroy the chronological value of the Wernerian terminology. With the abandon- ment of the Huttonian or so-called metamorphic doctrine, now shown to be false, so far at least as regards the sec- ondary or tertiary age of crystalline stratified rocks, we are naturally led back to the nomenclature of Werner and his school, which should be equally acceptable to endoplutonists and to neptunists, whether the latter adopt the chaotic hypothesis of Werner, the modified or ther- mochaotic hypothesis set forth by De la Beche and Dau- brde, or the crenitic hypothesis more recently maintained by the present writer in the essay just cited. § 2. The term "crystalline rocks " is conventionally used in geology to designate those original aggregates of which crystalline silicates make an essential part. Such silicates may however be associated in these aggregates with quartz, or with oxyds like magnetite, with carbonates, as in lime- stones and dolomite, and even with phosphates, as apatite, or with sulphates, as karstenite and gypsu.n. By a certain 192 THE GENETIC HISTORY [VI. license the term may also be extended to masses of defi- nite hydrous silicates, such as serpentine and pinite, which are in great part amorphous and colloidal, and also to un- crystalline silicates, often hydrated, and of indefinite com- position, such as palagonite, tachylite, pitchstone, and obsidian. The silicates having tlie composition of serpen- tine and of pinite assume, in some cases, proper crystalline forms ; palagonite is by heat readily changed in large part into a crystalline zeolite ; while glassy silicates, such as obsidian, by devitrification are in like manner resolved more or less completely into crystalline species. Hence rock-masses, including or even made up of these various un crystalline materials, may all be regarded as inchoately crystalline, and for geognostical purposes may be conve- niently classed with the crystalline rocks into which they graduaie. § 3. When stratified masses of quartz, calcite, dolomite, and karstenite are found among contemporaneous crystal- line silicated rocks, they generally enclose indigenous crystalline silicates, which give them a title to be regai-ded as parts of the accompanying crystalline series. The mineral species just named have, however, in other cases be.?ome aggregated in crystalline rock-masses in times and under conditions which did not permit the genesis of such species as feldspars, micas, amphibole, and pyroxene, which are the most characteristic silicates of the crystalline rocks. Hence we find beds of crystalline quartz, limestone, dolo- mite, karstenite, and gypsum interstratified with uncrys- talline rocks of detrital origin, and of secondary or tertiary age. It is worthy of note, however, that the conditions for the production of certain mineral silicates have continued ii' later ages, as is shown by the frequent formation of zeolitic, pectolitic, and other crystalline sili- cates in younger and uncrystalline rocks, and even down to our own time, and, moreover, by the occurrence among uncrystalline sediments of later geological periods, of deposits of serpentine, sepiolite, and glauconite. The J (1 mi OF CRYSTALLINE ROCKS. 193 history of both zeolitic and pcctolitic silicates as formed by secretions in basic rocks, and as generated in deep-sea ooze, and in the channels of thermal waters, has been dis- cussed at some length in the preceding essay, but there are facts in relation to the other silicates just mentioned which are of such importance in connection with the origin of crystalline rocks as to merit consideration in this place. § 4, Two examples of crystalline silicates related to zeolites in composition, which are found injecting organic remains in paleozoic limestones, have been observed by Sir J. W. Dawson, and were farther described and analyzed by the present writer in 1871. The first of these is from a Silurian limestone which is found near Woodstock, in the province of New Brunswick, and consists almost wholly of comminuted organic remains, including frag- ments of tiilobites, gasteropods, brachiopods, and joints and plates of small encrinites, the whole cemented by cal- cite. The pores of the crinoidal remains are tilled by a peculiar silicate, seen in sections or on surfaces etched by an acid. Surfaces thus treated show a congeries of curved, branching, and anastomosing cylindrical rods of the inject- ing mineral, sometimes forming a complete network, and exhibiting under a microscope coralloidal forms, with a white, frost-like, crystalline aspect resembling the variety of aragonite known as jlos ferri. The same crystalline mineral, as observed by Dawson, occasionally fills the interstices between the larger fragments of organic forms in the limestone, and, as he observes, "was evidently deposited before the calcite which cements the whole mass." § 5. The limestone in question is nearly pure, contain- ing very little magnesia or iron-oxyd, and leaves, after the action of cold dilute chlorhydric acid, five or six hun- dredths of insoluble residue which is the mineral in ques- tion with about one fourth its weight of silicious sand. The silicate is of a pale grayish-green color when seen in i if 1 f ' "^ m' 194 THE GENETIC HISTOKY tvi. mass, and, losing water, becomes bright reddisli-bro-wn by calcination. It is partially decomposed by strong heated chlorhydric acid, and completely by hot sulphuric acid, which dissolves alumina, ferrous oxyd, magnesia, and small portions of alkalies, leaving flocculent silica, which is readil}' separated by a solution of carbonate of soda from the accompanying quartz-grains. Thus analyzed, the mineral, which under a lens appeared, wholly crystalline and homogeneous, save the accompanying quartz, yielded silica 38.93, alumina 28.88, ferrous oxyd 18.86, magnesia 4.25, potash 1.G9, soda 0.48, water 6.91. The atomic ratio of this for ])rotoxyds, alumina, silica, and water is very nearly 1:2:3:1, whi h, abstracting the water, is that of zoisite ; the hydrous silicate jollyte being 1:2:3:2. I have given to this crystalline silicate, which is of curious interest alike for its composition and the mode of its occurrence, the name of hamelite, for the Rev. Dr. Hamel, rector of La\al University, Quebec* S 6. The second silicate above referred to is not unlike hamelite in its characters and manner of occurrence, though differing somewhat in atomic ratios. It was found in a mass of fossiliferous limestone said to be from a locality in the island of Anglesey, and including, " be- sides a small coral-like body referred to the genus Verti- cillopora, joints and plates of crinoids, small spiral gasteropod shells, with fragments of brachiopods, and a sponge-like organism with square meshes." All of these organic forms are more or less penetrated with a greenish silicate, which fills the cavities of the gasteropods, the central canal of the crinoids, and the pores of the Verti- cillopora. It has also replaced, or filled, the spongy fibres, and injected the minute cells of some of the crinoidal frag- ments, though many of these are solid throughout, in which respect the specimen differs from that from New Bruns- * Amer. Jour. Science, 1871, i., 379; also J. W. Dawson. The Dawn of Life, pp. 120-123, with figure of a portion of infiltrated crinoid on p. 103. U VI.] OF CRYSTALLINE ROCKS. 195 wick described above, where the infiltration of the cri- noidal remains is much more complete and perfect. Sir J. W. Dawson, to whom we owe these observations, sup- poses that in both cases the infiltration took place while the remains were still recent. § 7. Decalcified surfaces of this limestone from An- glesea show similar appearances to those presented by the New Brunswick specimen, and the casts of the gasteropo- dous shells, two millimetres in length, are in some cases perfect. The limestone is nearly pure with the exception of a little fine yellow ochreous matter which is insoluble in the dilute chlorhydric acid, and remains suspended in the solution, but is easily separated by washing from the pale grayish-green silicate. This equals about three hun- dredths of the weight of the limestone. When ignited in the air it assumes a bright fawn color, and under a lens contrasts strongly with the colorless grains of quartz with which it is mixed. Its chemical characters were like those of hamelite, and analyzed in the same manner it gave, after deducting 21.0 per cent of insoluble sand, the following composition : Silica 35.72, alumina 22.26, ferrous oxyd 21.42, magnesia 6.98, potash 1.49, soda 0.67, water 11.46 = 100.00.* This gives for protoxyds alumina, silica, and water very nearly the atomic ratios 3:4:7:4; but we are not sure of its homogeneous character. A silicate very like this in aspect and mode of occurrence has been found in a band of fossiliferous limestone near the base of the coal-measures in southern Ohio, but has not yet been chemically examined. § 8. In connection with these minerals should be no- ticed a greenish fibrous asbestiform silicate, elsewhere described by the writer, which occurs in veins traversing the anthracite and the carbonaceous shales of the coal- measures at Portsmouth, Rhode Island, either without admixture, or mingled with pyrites, or penetrating white quartz, and also coating the fragments of the crumbling * Amer, Jour. Science, 1871, ii., 57. THE GEI^ETIC HISTORY lyi. ! ( H » « '", 1 ^^® ,. • „ hvdvou3 silicate of ata- litic minerals or e^yy^ fprrous oxyd, n^ve gi> , ^ u p— ,rt' o^«- ^2^^^ "'"TtotScatf lil^e -r"*'"l;5:'^:'t named occur. T^:x^*a* — s i>et^-r s;t« often forms beds w-.h but ^^^^^ ^hat glauc »-'^ ^T^ilXTt si«.S oi fora»in«e.a a.^^^« „ite is met witli W™8 eologioal times, a "^"■^ ."rim; — in recent foramm - „ v^ .^ °°r:: Tride o^ its occ-ence - t^„ ^^,^^,,„ ous sea&. , ^r +Vift aluminous douDie»^ compo- ^'"""f™* t *-t - '^^"ir' atd'biir -e»t-"y forms from '""";.„ variable ; and.wni .itiono£glaucomte>s^eiy ^^^ .^^^ a, '* -"^^^^i,. tlrl OF CRYSTALLINE ROCKS. 197 Indeed, a so-called green-sand from the calcaire grossier, according to Berthier, is rather a highly ferrous serpentine, containing, silica 40.0, ferrous oxyd 24.7, magnesia 16.6, lime 3.3, alumina 1.7, water 12.6 = 98.9.* § 11. These variations show that the material in ques- tion is a mixture, and render it difficult to fix its real constitution. According to the multiplied analyses of Haushofer, the iron present in glauconite is for tlie most part in the ferric condition, the ferrous oxyd in various examples ranging from three to seven hundredths. The formula proposed by him represents glauconite as contain- ing 6.3 of ferrous oxyd, 8.3 of potash, anJ 9.6 of water, with 22.7 of ferric oxyd and 3.6 of alumina, giving for the atomic ratios of protoxyds, sesquioxyds, silica, and water, 1 : 3 : 9 : 3.f The very variable quantity of alu- mina found in glaucjnites may, however, well be owing to a zeolitic admixture ; and, if we hazard the conjecture that the large proportion of ferric oxyd therein is due to a partial oxydation of what was originally a ferro-potassic silicate, we should have for its composition before peroxy- dation (deducting the alumina as a zeolite with the above atomic ratios, like faujasite) a silicate with the ratios for protoxyds, silica, and water, of 3 : 9 : 3 ; corresponding to sepiolite, and to an unknown pectolitic silicate inter- mediate between pectolite and apophyllite, which may be supposed to have given rise alike to talc, to sepiolite, and to glauconite. The variable amounts of magnesia in glauconite itself would thus be due to an admixture of sepiolite. The reaction of such a soluble pectolitic com- pound, having a lime-potash base like apophyllite, with the dissolved magnesian salts in sea-water would generate a magnesian silicate having the ratio of talc and sepiolite (which latter forms beds in tertiary sediments), and with ferrous solutions by a similar double decomposition might * Beudant, Traits de Mineralogie, ii., 178. See also Report Geol. Survey of Canada, 1866, p. 231. t Cited in Dana's System of Mineralogy, 5th ed., p. 462. 198 THE GENETIC HISTORY [VL *,'4,i < . 1 •I 'I yield a ferro-potassic silicate like glaucoiiite. It is well known that un^er proper conditions decaying organic matters acting upon sediments containing ferric oxyd reduce this and give rise to such solutions, in which fer- rous carbonate is often associated with a proportion of an organic acid. Such a solution and redeposition in the forms of sideriteand pyrite goes on in sedimentary depos- its through this agency (Essay VII., § 35), and this would permit the conditions necessary to produce glauconite with the pectolitic silicate, which in the absence of the iron-solution would generate sepiolite by reaction with magnesian salts. § 12. The variations in the composition of glauconite- like minerals, and the existence in silicates similar to it in their mode of occurrence of more or less alumina and magnesia, probably corresponding, as suggested above, to admixtures of zeolite and sepiolite, are farther illustrated by the following analyses by the writer. I. is a typical glauconite from the green-sand beds of the cretaceous series in New Jersey ; II. a glauconite, remarkable for its fine green color, which forms layers in the Cambrian (Potsdam) sandstone at Red Bird, Minnesota ; III. a simi- lar material found in a Cambrian sandstone on the island of Orleans, near Quebec. The results, after deducting silicious sand, are calculated for one hundred parts, and the whole of the iron is represented as ferrous.* t I. 11. III. SUica . . . 50.70 46.58 50.'i Ferrous oxyd . 22.50 20.61 8.6 Magnesia . . . 2.16 1.27 3.7 Lime . . . . 1.11 2.49 — Alumina . . , 8.03 11.45 19.8 Potash . . . . 5.80 6.96 8.2 Soda . . . .76 .98 .5 Water 8.95 9.66 8.5 100.00 100.00 100.00 r * Geology of Canada in 1863, p. 486; also Rep. Geol. Survey of Can- ada 1863-^9, p. 232. VI.] OF CRYSTALLINE ROCKS. 199 § 13. The crenitic hypothfpis advanced by the present writer in the preceding essay to exphiin the aqueous ori- gin of the mineral species which make up alike the gran- ites and the crystalline stratified rocks, supposes that from an early period watery solutions analogous to those which in later times have given rise to zeolitic and pectolitic minerals, played an important part in the chemistry of the earth. The double silicates of alumina and lime or alkalies then dissolved, are conceived to have been the source not only of the feldspars and the zeolites, but of epidote, garnet, muscovitic micas, and tourmalines, and, b}'^ their reactions with magnesian and ferrous solutions, of the chlorites and the highly protobasic micas. At the same time the dissolved protoxyd-silicates not onl}^ gave rise to species like pectolite and apophyllite, but, by similar reactions, to pyroxene, amphibole, chrysolite, serpentine, talc, sepiolite, and glauconite, arid, by decomposition through carbonic dioxyd, to carbonate of lime. In both cases the solutions, like those in later zeolite-bearing rocks, carried free silica and iron-oxyd, which were deposited as quartz and magnetite and hematite. These silicated solu- tions, according to this hypothesis, resulted primarily from the action of permeating waters at high temperatures, ■ under pressure, upon the universal stratum of basic plu- tonic rock; and secondarily from their action upon the dis- placed portions of this stratum, which, in a more or less modified form, have appeared in all geological periods as erupted basic rocks. These, in their secreted minerals, show us in later times, and on a smaller scale, the process which in previous ages built up great masses of indige- nous and endogenous crystalline rocks. To what extent these deposits, more or less concretionary in their origin and their arrangement, were laid down horizontally, and to what extent in inclined or vertical layers, as in many veinstones, is a question which will be discussed farther on in this essay. § 14. Having thus briefly restated the crenitic hypoth- 200 THE GENETIC HISTOET CVI. .li ! esis so far as it is related to the classes of rocks already noticed, we have to consider in the next place the ques- tion of exoplutonic or eruptive rocks. It will be remem- bered that the existence of such rocks, having an igneous origin, was not admitted by the "NVernerians, who conceived not only all endogenous rocks, but also all exotic masses, except modern lavas, to be of aqueous origin. By the earlier Huttonians, wlio understood better the geological importance of the eruptive rocks, these were looked upon as results of the fusion of deeply buried detrital materials, themselves derived from similar rocks of higher antiquity. The hypothesis of great chemical changes to explain the genesis of many crystalline rocks from such material by what was comprehensively designated as "metamorphism," and generally involved a supposed metasomatic process, was devised at a later day by the disciples of Hutton. Haidinger and Bischof may be looked upon as the origina- tors of that view of metasomatic changes in rock-masses by aqueous action which, from its supposed analogy with the phenomena giving rise to what are called pseudomor- phous shapes or pseudo-crystals, has been infelicitously de- scribed as " jjseudomorphism on a broad scale." * § 15. The stratiform arrangement, which extends to the intimate structure of crystalline masses such as gneisses and mica-schists, is by endoplutonists supposed to be due to movements in an imperfectly homogeneous semi-fluid material dependent on unequal cooling and the rotation of the globe, and to be analogous to the banded structure apparent in lavas and furnace-slags. In the exoplutonic hypothesis, on the contrary, it is maintained that the in- ternal movements in such material, when forced outwards and upwards chrough the earth's superficial crust, have given to the masses that laminated structure and that arrangement of the constituent elements which, alike by Wernerians and Huttonians, are regarded as evidences of deposition from water. This latter or exoplutonic view * Ante, page 100. VI.] OP CRYSTALLINE ROCKS. 201 was clearly expressed by Poulett Scrope, sixty years since, in his "New Theory of the Earth," published in 1825 (^ante, page 81), wherein he imagines the granite to have formed the original surface of the globe, and supposes that movements in extruded portions of the mass com- pressed beneath overlying sediments gave to it the gneissic structure. He insists upon the friction of its elements "as they were urged forward in the direction of their plane surfaces towards the orifice of protrusion, along the eximnding granite beneath, the laminae being elon- gated and the crystals forced to arrange themselves in the direction of the movement." This view was adopted, though without acknowledgment, by J. D. Dana in 1843, when he argued that the schistose structure of gneiss and mica-schist is not a satisfactory evidence of sedimentary origin, since erupted rocks may assume a laminated arrange- ment.* § 16. The same notion has continued to find favor among geologists of the plutonist school up to the present time. Poulett Scrope himself, in rewriting his famous treatise on Volcanoes, after a lapse of thirty-seven years, restates his argument with great precision. He therein supposes that the primitive material of the globe, so far as known, was an aggregate consisting essentially of feldspar, quartz, and mica, in a crystalline or granular condition. This material, which was impregnated with water and highly heated, possessed a certain plasticity, and when extruded by pressure took upon itself a stratiform struc- ture, being "bodily forced up the axial fissure of disloca- tion in crumpled zigzag folds or upright walls of vertical laminated rock." To show to what extent this view had met the approval of other geologists, Scrope farther ob- served, " The late Mr. Sharpe and Mr. D: vwin, as is well * Scrope, Considerations on Volcanoes, etc., 1825, p. 22. See also J. D. i)ana. On the Analogies Between Modem Igneous Rocks and the so-called Primary Formations, 1843; Amer. Jour. Science, 1843, xlv., 104-129 ; and ante, pages 89, 90. I 202 THE GENETIC HISTORY tvi. f'^ ■I : known, concurred in the opinion here given, that at least as respects the oldest or fundamental gneiss, its foliated structure is duo not to original sedimentary deposition, but to the movement of the particles under great pressure while the nuiss was in a condition of imperfect igneous fluidity. Prof. Naumann has still more recently advo- cated the same view, which is, however, resisted by Lyell, Murchison, Geikie, and others." * § 17. The same view has very recently been brought forward by Joh. Lehmann, who maintains, with Scrope, that the schistose structure in crystalline rocks is no evi- dence of aqueous deposition, but is imposed upon them by the process of extrusion. The Saxon granulites, according to Lehmann, were intrusive masses which con- solidated among sedimentary strata far below the surface, and, being afterwards forced up by great pressure, took upon themselves a banded schistose arrangement, the adja- cent strata, more or less impregnated by the granulitic ma- terial, appearing as micaceous gneisses and mica-schists.f This whole g-anulitic series of Saxony may be described as made up of fine-grained binary gneisses (granulites), passing into micaceous gneisses and mica-schists, and has been by the present writer elsewhere referred to the younger gneissic or Montalban series of crystalline rocks.J § 18. An example of the resuscitation of the views of Poulett Scrope in North America is found in a recent note by Prof. H. Carvill Lewis on the crystalline schists of east- ern Pennsylvania. A belt of these which crosses the Schuylkill near Philadelphia, long ago described by H. D. Kogers, and since by the present writer,§ includes a band of granitoid gneiss succeeded by micaceous gneisses and * Scrope, on Volcanoes, 2d ed., 1862, as revised in 1872, pp. 300-305. t Joh. Lehmann; Untersuchungen iiber die Enstehung dov Altkrys- tallinen Schiefergesteine, 1884. Not having boon able to consult this work, I am indebted for a notice of its argument to a review in the Amer. Jour, Science, xxviii., p. 39. t See Essay X., §§ 71), 80. § See Himt, Azoic liocks, pp. 10-15 and 200; also Essay X., § 18, t» VI.] OF CRYSTALLINE ROCKS. 203 micaceous schists, often giirnetiferous, comprising a layer of serpentine witli steatite and dioritic rocks, the whole rep- resenting both the okler and the younger gneissic series so well known in eastern North America as Laurentian and Montalban. The rocks in tiiis belt, notwithstanding their stratiform character, are, in the opinion of Lewis, "of purely eruptive origin, consisting of syenites, acid gabbros, trap-granulites, and other igneous rocks, often highly metamorphosed. It is the outer peripheral portions of this zone to which attention is here directed. While the rocks are massive in the centre, this outer portion has been enormously compressed, folded, and faulted, with the result of producing a tough banded porphyritic fluxion- gneiss." Lewis supposes "a recrystallization of the old material under the influence of pressure-fluxion," by which he conceives the feldspar to have been recrystallized. " In similar manner the biotite has been made out of the old hornblende, garnets have been developed, and the quartz has been granulated and optically distorted by the pres- sure." In another example mentioned by him, a belt of sphene-bearing amphibolite schist, described as included unconformably in the mica-schists of Philadelphia, is sup- posed by Lewis to be " a highly metamorphosed intrusive dike of Lower Silurian age. The original augite or diallage has been completely converted into fibrous hornblende, and the influence of pressure is shown in the perfectly laminated character of the schist, in the close foldings produced, and in the minute structure of the rock." " The chemical changes and interchanges of elements which might result from a loosening of molecular combinations under extreme pressure," and their subsequent re-arrange- ment to form new compounds, suggest to Lewis great possibilities in the so-called " mechanical metamorphisra " now advocated by some to replace the discredited dogma of chemical metamorphism, which has hitherto played such an important part among a school of geologists.* * H. C. Lewis, Proc. British Association, in Nature, Oct. 8, 1885, p. 560. V:: > ';! « f 204 THE GENETIC HISTOIIY tvi. § 10. Tims, wliile the ancient Wernerians maintained the direct dopoHition of granite from aqueous solutions in a chaotic ocean, the plutonists, from Ponlett Scrope in 1825 to Darwin, Naunuuin, Lclnnann, and I^ewis, assert the igneous origin not only of granites but of gneisses and micaceous and amphibolic schists, and the followers of the Iluttonian or metamorphic school hold an untenable and an illogical position between the two, — deriving the materials of both of these rocks from a primary granitic mass, whose origin is unaccounted for, and whose sup- posed transforniiitions chemistry cannot explain. § 20. It remains to notice, in connection with the nep- tunian, the plutonic, and the metamorphic hypotheses, regarding the sources and the geognostic relations of the crystalline rocks, a view which has been proposed to ex- phain the attitude of certain apparently exotic masses: wliich is that their present position is due neither to de- position from solution nor to intrusion in a fluid or plastic condition, but to local movements which have permitted portions of rigid rock to displace and even penetrate softer and more yielding materials in their vicinity. Examples of this are described by Stapff as seen in the St. G(.Miard tunnel in the Alps, where great masses of serpentine 'nve been caused to traverse adjacent schistose strata ; the solid condition of the intruding rock being made evident by the accompanying breccia, consisting of its fragments.* There is reason to believe that such instances are not un- common, and that in many cases the phenomenon of in- trusion is due to the superior hardness of the intruding rock, broken beds or masses of which are forced through softer strata; the conditions being the reverse of those which attend plutonic or volcanic injections. The notion that rocks when in a solid condition may be intruded aniong others, is found in the pages of more than one writer on geological questions, but, so far as the writer is aware, is for the first time clearly and satisfactorily de- * See farther, Essay X., §§ 128-130, wliere details and references are given. VJ.] OF CUYSTALLLNE HOCKS. 205 fined in tlio description of Rtapff, whioli ia an important concoptiun gu';ied fur the student of gongnosy. § 21. The endoplutonists. us wo have .seen, have sought to explain tlio laminated structure of certain crystalline rocks, not, like the exoplutonists, by the pressure attend- ant on extrusion, but by movements in an imperfectly fluid material in which, during refrigeration, a separation of solid matters and a process of elii^uation were going on. The possible production in this manner alike of unstrati- fied and stratiform crystalline rocks from an igneous mass is ingeniously set forth by Thomas Macfarlane in his studies of the geology of Lake Superior.* He notes first the occurrence of fragments of denser and more basic hornblendic aggregates enclosed in lighter and less basic granitoid masses, and, from these facts, and the composi- tion and specific gravity of granitic veins penetrating tlio masses, conjectures that these various products represent different stages in crystallization from a primitive magma, the first separated ])ortions from whicli were more basic and the later more silicious. If this took place when the mass was undisturbed, a granitoid rock would be formed ; but if while it was in motion, "hornblendic and micaceous schists and gneisses were most probably the results of this process, and the strike of these would indicate the direction of the current at the time of their formation." The material thus sepa- rated, notwithstanding its greater specific gravity, is sup- posed to have formed at the surface of the molten mass, as a result of cooling; but in Macfarlane's view "there arrived a time when, from some cause or other, these first rocks were rent or broken up and the crevices or interstices became filled with the still fluid and more silicious material which existed beneath them. This gradually solidified in the cracks, or in the spaces surrounding the fragments, and the whole became again a consolidated crust above a * Geological Features of Lake Superior, Canadian Naturalist, May, 1867. fill *, li t}'t "' m nr 0^ .}■?. 206 THE GENETIC HISTOIIY [VI. fluid mass of still more silicioiis material," which by sub- sequent movements would again be intruded in the form ci veins in the broken crust. This restatement of the hypothesis of the solidification of a molten globe from above downwards, already taught by Naumann,* serves to show how the endoplutonist school explains the origin alike of massive and of stratiform crystalline rocks, and may be compared with the detailed statement of the exo- plutonist view as set forth by Poulett Scrope. § 22. The broad distinction sometimes drawn between stratified crystalline rocks as of indigenous and aqueous origin, and unstratified rocks as intruded or exotic masses of igneous origin, thus finds no place in the hypotheses of the plutonic schools, according to both of which these two classes of rocks have come directly from a primitive fused mass, which was either simple or had become complex through differentiation. The Huttonian school also, which teaches that eruptive rocks, in many if not in all cases, were originally sediments which, as a result of pro- found alteration, have lost their bedded structure, arrives by a different route to a conclusion not unlike chat of the plutonists ; namely, that the differences between stratified and unstratified rocks are due solely to superinduced structure and geognostic relations. Those who, for the most part unfamiliar with any other view, acquiesce in the metamorphic hypothesis of Hutton and his followers, now so popular with a school of writers on geology, are scarcely prepared, without farther study, to criticise intel- ligently either the plutonic or the crenitic hypothesis of l^e origin of crystalline rocks. The latter, as set forth in the preceding essay, and concisely resumed on page 199 of the present, supposes that the source of all crystalline rocks is to be sought in a previously solidified primary plu- tonic material. The elements of these rocks have been de- rived in part indirectly, by aqueous solution, and in part directly from this original mass, more or less profoundly * See ante, page 85, VI.] OF CRYSTALLINE KOCKS. 207 II altered alike by previous aqueous action and by differen- tiation through ciystallization and eliquation. By this hypothesis, as we have elsewhere attempted to show, we may hope to lay the foundation of a rational geogeny and geognosy. § 23. We have already, in the preceding essay, consid- ered at some length the views of those who, noting the existence of predominant types of crystalline rocks, have sought to explain their origin by supposing the presence beneath the earth's solid crust of two distinct layers of molten rock : an upper, lighter, and more viscous silicious or so-called acidic stratum, the material of trachj^tes, gran- ites, and gneiss ; and a lower, heavier, and more fluid basic layer, the source of doleritic and basaltic rocks, — a view which was put forth by John Phillips, defended by Bun- sen, and elaborated and more definitely formulated by Durocher. To this are opposed the modified view of Von Waltershausen, of a gradual passage downward in a liquid mass from a more acidic to a more basic portion, and the entirely distinct view held and defended by the present writer as the basis of the crenitic hypothesis. According to this the plutonic underworld, so far as it intervenes directly in geologic phenomena, is an essentially homoge- neous basic rock, not in a state of simple and original igneous fusion, but solidified and subsequently impreg- nated with water, which communicates a certain plasticity to the highly heated mass, and, moreover, dissolves and removes therefrom the materials of the trachytic and gran- itic rock, — which are thus primarily of aqueous origin. § 24. This process implies secular changes in the com- position of the plutonic stratum, which are mr reover local, since the conditions of solution and upward perco- lation will vary in different areas, and during different periods in the same area. It involves also a coxT'^spond- ing change in the nature of the materials dissolved, so that differences greater or less are to be looked for in the composition alike of eruptive plutonic and of crenitic 1*1 f i 1 t- ♦ ■ii '1 1 208 THE GENETIC HISTORY [VI. are compf^^'e^^' i^e evi composition or xi independent of aqueous Uon i ^^^^^^ ,^,evyation of Plutonic mass did «.«\^47^^i,,,,,,ed by him m lus le- ?)uroclier, and was m 1857^^^ petrology.* To this I 1 ni.iA pssav on Compaiative x ryurocher's view nS tXn in 1858, ^''''"If *t" imagined by hin. S^ t o st.au of -«»»;::, Itr; a partial .y^^ c^occasionally move o less. „ ^^^^ „,e to the nation ami ^l""*'","' -L and basic cvystalline locl-s.t principal vavisties of acid.c ana ^,^^„„,, bj ^\ 25. This view was sU^i ™' J, .vhich have pvo- Duvooher, who declared ' Jhe »• g ^ ^ t,^u„ duced the igneous rocks ate to ^j j„^,o„, t hs, which, holding many »* ^ ;,u„y„ according to separate in solid.tymg jto j^tt ^^.^^^ ,. _ t, en-cmn- the circumstances of their son ^^^^^ than of an rtan es being ^'^r^'^ fcoVing » 1^-° »"* nterior order." S^lj'^^^^^^'^th a tJachytic porphyry, hiffiily aluminous phonol.te wit ^j^^t an Se'silicious and '-'^^^'""^^.tns would give the admixture of these »\,<=S,^e,^nd expresses the opm- composition of a normal ttacliyte, j^^^, u the two oTthat the rocks thus conipa ed a^^^. P ^^^^ ^^^^^ .„ ,, opposite products of an el.'l"^'^" ^^^^aon of two oppo- Sst of the liquid mass, as in the .^ ^^ ^^^^,_,^ .^^^ ^ site alloys, into which a ™<=*f'f elicuation he conceived epa^te" Tl«'e rhenome^i of ^e^u ^^_^^ ^^^^^ ^,k, r: b^-^rSf-^^e^ibe eart^mi in its caverns td crev^^t as -U a-* the surface^ ^^ ''"l 26. The P--o'^^^''fy f. t lenoniena due to the ehUts .dio l-y^^ -'^tuing and solidifying points erystalluation and differ _^^^^^_, „,„„,,„«.. ., . Ann»l« des Mine., ^L, 21. j,^^^„_,^ „, 1859. VI.] OF CRYSTALLINE ROCKS. 209 of metallic alloys, as, for example, the separation of lead from its silver-bearing alloy in the Pattinson process, and tlip eliquation of this metal from its alloy with copper. It was adopted by Macfarlane in 1867, in explanation of the relations of more or less basic hornblendic and gran- itic rocks, already cited in § 21, and finds a striking illus- tration in the late experiments of Fouqu^ and Michel Ldvy on the artificial production of crystalline mineral species from fused vitreous mixtures. From such a mixture, con- taining the elements of six parts of chrysolite, two of pyroxene, and six of labradorite, kept at a heat near whiteness for forty-eight hours, there separated crystals of chrysolite, 0.5 millimetre in diameter, together with mag- netite and spinel (picotite) ; a vitreous magma still re- maining, from which crystallized, at a lower temperature, macled crystals of labradorite, with pyroxene, magnetite, and spinel, as before. It is apparent that with, a greater lapse of time, and the formation of larger crystals of chrysolite, which has a specific gravity of about 3.4, these would, under ':he influence of gravity, subside, together with magnetitt and spinel, from a fused glass holding the elements of pyroxene and feldspar, the r.iore so as the den- sity of fused doleritic and basaltic material is less than 2.8. From such a slowly cooling mixture the process of eliquation would, under favorable conditions, give rise to a highly chrjoolitic aggregate on the one hand and to a dolerite with little or no chrysolite on the other. More- over, if, as is probable, there are conditions under which pyroxene may be separated in a similar manner from the feldspathic element, we should have a farther differentia- tion, giving rise to heavier and highly pyroxenic portions on the one hand and to lighter and more feldspathic por- tions on the other. § 27. The careful student of crj^stalline rocks will have noticed many examples in nature of variations in different portions of eruptive masses, which find a ready explanation in a process of partial solidification and eli- 210 THE GENETIC HISTORY m. ■■: :■■% (jUiition, as suggested by Duroclier and illustrated by the experiments of Fouqu^ and Michel L6vy. This is well displayed in certain rocks intruded among the Ordovician strata of the St. Lawrence valley, near Montreal, and form- ing the hills known as llougeniont, Moutarville, and Mount Royal. These, as I have long since described then), are essentially doleritic, but present very great dif- ferences in the proportions of their mineralogical elements in contiguous parts. Thus in some portions of these masses we have a pyroxene and labradorite rock in which these two elements are pretty equally distributed, while in other portions the rock is almost wholly a black, coarsely crystalline pyroxene, with but an insignificant piuportion of the feldspathic element. Elsewhere the arrangement of these two species gives rise to a stratiforn \ structure. § 28. As described by me in 1863,* for Mount Royal, " mixtures of augite and feldspar are met with, constitut- ing a granitoid dolerite, in parts of which the feldspar pre- dominates, giving rise to a light grayish rock. Portions of this chaa'acter are sometimes found limited on either side by bands of nearly pure black pyroxenite, giving at first sight the aspect of stratification. The bands of these two varieties are found curiously contorted, and . . . seem to have resulted from movements in a heteroge}i80us pasty mass, which have effected a partial blending of an augitic magma with one more feldspathic in nature." In the doleritic mass of Montarville the alternation of a coarse- grained variety of dolerite, porphyritic from the presence of large crystals of pyroxene, with a finer-grained and whiter variety is noticed, the two "being arranged in bands whose varying thickness and curving lines suggest the notion that they have been produced by the flow and the partial commingling of two fluid masses." Of this stratiform structure it was then said, it seems to be due to " the arrangement of crystals during the movement of the * Geology of Canada, 1863, pp. 605. 867, and Amer. Jour. Science, 1S04, xxxviii., 17.J-178. (• VI.] OF CRYSTALLINE HOCKS. 211 half-liquid crystalline mass, but it may in some instances arise from the subsequent formation of crystals arranged in parallel j^lanes." * § 29. The feldspars mentioned, as shown by the pub- lished analyses by the writer, are near in composition to labradorite. The composite rocks described also contain, besides pyroxene, more or less magnetite and raenacanite, with chrysolite. This last specie? is for the most part distributed sparsely through these rocks, but occasionally, like the pyroxenic element, occurs in predominant quan- tity. An example of this is seen in a coarsely granitoid chrysolitic aggregate, exposed witli the same characters, over an area of many hundred square feet, on Montarville. The chrysolite in this rock is in irregular crystalline masses from five to ten millimetres in diameter, and was separately analyzed, as was the black pyroxene, in still larger and well defined crystals from the mass, and also the feldspathic element, selected as carefully as possible. For an analysis of the rock as a whole, it was attacked in fine powder successively by dilute sulphuric acid and by a weak solution of soda, the portions thus dissolved being analyzed separately, as well as the insoluble residue. The relative proportions of these being 55.0 per cent of the former and 45.0 of the latter, it became possible to calcu- late the composition of the rock as a whole. * Farther Illustrations of this are given by the author In a communi- cation to the Boston Society of Natural History, January 7, 1874: "Among these was a specimen shown from Groton, Connecticut, in which a large angular fragment of strongly banded micaceous gneiss is enclosed in a fine-grained eruptive granite, the mica plates in which are so arranged as to show a beautiful and even stratification in contact with the broken edges of the gneiss, but at right angles to the strata of the latter. Another example is afforded by the eruptive diabase from the mesozoic sandstone of Lambertville, New Jersey, which is conspicuously marked by light and dark bands, due to the alternate predominance of one or the other of the constituent minerals ; and still another is a fine- grained dark micaceous dolerite dike from the Trenton limestone at Mon- treal, in which the abundant laminsB of mica (probably biotite) are arranged parallel to the walls of the dike." Chem. and Geol. Essays, p. 18C. I I'.i m >■ :/ ! •MHIti 212 THE GSNETIC HISTORY [VI. § 30. In the following table, I. is the composition of the feldspar ; II the pyroxene ; and III. the chrysolite ; IV. the soluble portion (55.0 per cent), chiefly chrysolite ; V. the insoluble portion (45.0 per cent) ; VI. the rock as a whole, including an undetermined amount of titanic oxyd with the iron-oxyd. For the purposes of comparison we give under VII. the composition of the supposed basic magma of the earth's interior, as deduced by Bunsen from the mean of several analyses of basic eruptive rocks, and under VIII. the composition of the same as calculated by Durocher, who, however, admits a range in proportions through geologic time which includes the figures adopted by Bunsen. The last five analyses are :\ecessarily calcu- lated for one hundred parts, and the whole of the iron is represented as ferrous oxyd, although an unknown pro- portion exists in a higher state of oxydation. i ■■ M' I. II. III. IV. Silica . . . 53.10 49.40 37.17 37.30 Alumina . . 26.80 6.70 — 3.00 Lime . . . 11.48 21.88 — — Magnesia . . .72 13.06 39.68 33.50 Ferrous oxyd . 1.35 7.88 22.54 26.20 Soda .... 4.24 .74 — — Potash . . . .71 — — — Volatile . . .60 .50 — — 99.00 100.11 99.39 100.00 V. VI. VII. VIIL Silica . . . . 49.35 42.70 48.47 51.5 Alumina . . . 18.92 10.16 14.78 16.0 Lime . . . . 18.36 8.27 11.87 8.0 Magnesia 6.36 21.29 6.89 6.0 Ferrous oxyd , 4.51 16.45 15.38 13.0 Alkalies . . 2.50 1.13 2.61 4.0 100.00 100.00 100.00 § 31. The process which has thus given rise in parts of a mountain mass of dolerite to considerable areas of a rock containing over 21.0 of magnesia, and more than one half HI VI.] OP CRYSTALLINE ROCKS. 213 its weight of chrysolite, find in other parts of the same mass to an aggregate of pyroxene and labradorite almost, and in some cases wholly, destitute of chiysolite, is readily explained if we admit a separation from a still fluid mass of the previously crystallized and heavier chrysolite by a process like that imagined by Durocher. It will be noticed that the insoluble and non-chrysolitic portion separated from the Montarville rock, V., is near in compo- sition to an ordinary dolerite, or to the normal basic types of Bunsen and Durocher. We may conjecture that iloler- ites of average composition are, perhaps, themselves pro- ducts separated by eliquation from a more chrysolitic aggregate. § 32. The segregation of groups of crystals, which takes place in the devitrification of glasses, shows, within narrow limits, the process of differentiation through crys- tallization in a homogeneous mass. The operation of this process on a larger scale, giving rise to remarkable miner- alogical differences, is well shown in the careful studie? by Fouqud, in 1873, on the recent eruptive rocks from Santorin. The ordinary type of these lavas examined by him was a vitreous mass enclosing crystals of feldspars, with pyroxene, chrysolite, and magnetite. The feldspar was chiefly labradorite, but its association with crystals of albite, and with some anorthite, was established. Druses in this same rock were, however, filled with anorthite, associated Avith a pyroxene and a chrysolite, both differing from those contained in the paste in being less dense and in containing less ferrous oxyd. In an obsidian-like rock from the same region were rounded masses, sometimes a metre in diameter, gray in color, and made up of crystal- line anorthite, with pyroxene, chrysolite, sphene, and magnetite, Avith very little paste. The small portions of alumina found in the analyses of these pyroxenes were apparently, according to Fouqu^, derived from adherent anorthite, but another variety of pyroxene, seemingly very pure, and freed from anorthite, contained 12.4 per cent of :i^:' ,':?-i; ;':!• ■ t' THE GENETIC HISTORY IVI. 214 „,„,, _a trao aluminous ry«>f "^ ..,„,.,' dvic acia.wluoh veaaUy attacks the <;°;'^S^f J '„v,,aaorite, an.l anorth.te, alike the vitreom pas e atote^ ,,„,y,„ute, winch, but leaving ^'=1""^*' \ Suy^taekea by the acid ,• or, Uke aun^luV-oli^. ''';'' ^"'fS its action. , . , like stauvolite and .jrcon, '««'^' „{ ^^e hypothesis o£ S 88. Dniochev, in h s sta einen ^^^^^^^ ^^ ^^^^.^j^ ^j^,, eliquation as applied to J'"l« ,, j^ ,,„t an illustration, Jeess of ^eS''-g:'»'°^r he nuestion of differentiation, ■aises, in connection ™ *7f„„„„l„de8 from his com- auothev not less impo.tant- "" „,e of the ages parative studies that, " m the ^^\ ^,,Ms fi-om which divide the piimaiy ami u composi- th othe.-," theve have heen chan -^^ ^^^ ^^ tion of the fluid ™:'^' ^'"™ "„f the acidic layer-the and, nioreovev, tha in the cas ^^^_.the.e was a source of the gram ic '''^^^ *3redths in the proportion diminution of eight or nn^ » „^^,,,, while the propoi- o£ silica, and of one-fitth >" «'« J ^i„„,t doubled, and ti„„s of lime and """-"^^X rf»"S'=«' -''""''"^ '" that of the soda tripled. Sunda ^ b^ ,.,j„,,ented by l,i™,have taken place "*%;^ the comparative study dokultes, te^'l'^' ™f P f't'l'i^ the ferro-ealeiferous layer of which he concludes that application ,nd 94 Pe^-^^^^^^^^g Jiies, lime, magnesia, iron ^^^«' j,,^^, ;^[rBerhtrS?'Acad. Wissenschaft, U, 1885. Nov. 6, 1885.) Mf^ ▼14 OP CRYSTALLINiS ROCKS. Cl-i from the primary to the tertiary period . . . there was a sensible diminution of silica and potash, and a notable augmentation of soda and lime." Of these changes " the diminution of silica and potash in the modern rocks, both of the acidic and basic groups," was by Durocher exi)lained by supposing that while these imaginary igneous layers remain distinct from each other, there is, nevertheless, in each a partial separation of these elements by gravity, resulting in an accumulation of silica and potash in their upper portions, and of lime in their lower portions. The augmentation in the proportion of soda was by him re- ferred to a special and independent cause, the supposed "intervention of sea-water in the formation of igneous products during the later geological periods," which, as he writes, would explain "the considerable increase of soda in the more modern of the igneous rocks, whether they be derived from the acidic or the basic layer." § 34. While Durocher included in the category of eruptive rocks certain masses, such as those of magnetite, serpentine, and various amphibolic rocks, for which an igneous origin is not admissible (so that some of his data may be questioned), the correctness of his important generalizations, which suggest a vast geogenic problem, cannot be contested. As regards his proposed explana- tion, it is easy to conceive that a separation by specific gravity might possibly cause such variations, alike in the acidic and the basic layer, that the ejections in the course of ages from successively lower portions of each of these would show the gradual diminution observed in the pro- portions of silica and potash, as well as the augmentatio.i of lime. To this ingenious explanation, however, it is to be objected that it is based upon the unproved and, in the opinion of many modern philosophers, the untenable hypothesis of a molten substratum, and, moreover, one divided into two distinct zones. The whole of the phe- nomena in question, moreover, admit of a simpler and, it is believed, a more probable explanation by the crenitic 216 THE GENETIC HISTOltY [VI. M'? hypothesis. This, as we have seen, supposes a constant and progressive differentiation of an original basic plu- tonic mass through the action of water, which removes therefrom in tlie elements of orthoclase and quartz, — the chief constituents of granitic rocks, — preponderant pro- portions of iiilica and potash; an action which would result at last in the partial exhaustion of the lixiviated portion of the basic rock, which, with the diminution of the amount of available silica and potash, would finally yield to the solvent action of the waters only the elements of the more basic feldspars. As a result of this continued process, the crenitic products themselves will naturally show a diminution in the proportions of silica and potash, by reason of the progressive exhaustion of the source of these, while this residual portion of basic rock will not only exhibit a reduction in the proportions of silica and pota'^h, but a relative increase in the proportion of lime. Moreover, the sodium and magnesium-chlorids which, from the results of sub-aerial decay, find their way into the sur- face-waters, which subsequently pass downwards in the process of lixiviation, may by double exchange effect the displacement of potash and the fixation of soda and mag- nesia in the basic mass, as explained farther on. § 35. This hypothesis thus explains at the same time the origin of the highly silicic and potassic rocks, repre- sented by the granites, and the conversion of the original plutonic stratum into a more and more basic material, pro- gressively richer in alumina, soda, lime, and magnesia. It moreover requires that the long-continued lixiviation of a given area of plutonic rock should at length reach a point at which water could no longer remove from it the elements of orthoclase and quartz. With the disappear- ance of the latter would come the elements of the more basic feldspars, such as andesite and labradorite, as well as protoxyd-silicates, which together predominate in the norites and the diorites, characteristic crenitic rocks of the later crystalline series, as the Norian and Huronian, n.1 OF CRYSTALLINE ROCKS. 217 i''^ which succeed the granitea anil the granitoid gneisses of the earlier [)eriods. Tlie crenitio hypothesis, as we have elsewhere seen, in- volves the conception that all trachytic and granitic rocks are primarily of crenitic origin, and that penetrating gran- itic masses, when not, as is the case with most granitic veins, directly crenitic or endogenous masses, are displaced portions of older crenitic deposits. The first-formed granitic layer itself, it is held, may become softened under the combined influences of wate" and internal heat, and, being then displaced, may appear in an eruptive form. § 3G. The question here arises as to the respective parts which crenitic action, on the one hand, and crystalli- zation and eliquation, on the other, may play in the genesis of various types of crystalline rocks. It is apparent, from the illustrations which we have given, that by the latter process aggregates could, in paleozoic times, be formed in which chrysolite makes more than one half the weight of the mass, and others in which either pyroxene or labra- dorite may largely predominate. The texture and the gen- eral facies of these different mineral aggregates, not less than their geognostic relations, however, suffice to distin- guish them from crenitic deposits of somewhat similar composition. It was from a failure to recognize these differences that the original Wernerians denied or. min- imized the significance of igneous rocks, on ^he one hand, and that the later plutonists of both schools, on the other hand, have argued the igneous origin of rocks of manifestly crenitic origin. The Wernerir is, from the stratiform structure of gneiss, which they ascribed to its aqueous origin, argued for a similar origin for the granite into which it appears to graduate, while the plutonists from an analogous structure in undoubtedly igneous rocks conclude to the igneous origin of gneiss. We have already noticed this laminated or stratiform character in Plutonic rocks, the true significance of which as evidences of igneous flow should not be lost sight of (page 210). t«; 218 THE GENETIC HISTORY [vr. / ■M fV ,'.i § 37. It must be kept in mind that the cronitic p'":cesa, unlike eliquation, modifies the primary mass not only by abstraction bnt by addition, since the surface-watev, which, by the hypothesis, is the dissolving agent, will bring with it in solution, in varying propoitions, salts of calcium and magnesium, of potassium and of sodium, the action of all which upon the heated plntonic mass will effect certain interchanges, resulting in the fixation of bases like mag- nesia, whose 3ilicated compounds are comparatively insol- uble in the circulating waters, and perhaps in a substitu- tion of soda for lime. It is not improbable that jiotassic solutions from some hical source * could thus l)e introduced, and give rise by their action upon a doleritic mass, either integral or partially differentiated by eliquation, to a ma- terial so rich in potash as to furnish the elements of leu- cite, — \Vhich has the oxygen-ratios of andesite. § 38. The genesis of rocks like phonolite, which are essentially made up of a feldspar having the orthoclase- ratios, with an admixture with a more basic silicate, as nephelite or a zeolite, can, however, hardly be explained save as an educt of crenitic action, like trachyte and gran- ite. It represents, however, a period in the history of the plutonic mass when, from a diminution of silica, the pro- duction of quartz ceases, and more basic feklspathi') or zeolitic compounds begin to replace the orthoclase. When from compounds like these, in which the proportion of protoxyds to alumina falls below the normal oxygen-ratio of 1 : 3, we pass to others, like the muscovitic micas, most tourmalines, and the pinite-like minerals, with a dimin- ished proportion of protoxyds, we have probably in all * While In ordinary spring-waters the proportion of potassium to so- dium salts is small, seldom exceeding two or three hundredths of these bases, calculated as chlorids, I have shown that in an alkaline spring- water from paleozoic shales at St. Ours, Quebec, containing in a litre about 0.3 gramme of alkalies, chiefly as carbonates and chlorids, the potas- sium thus calculated equalled 25 per cent. In the case of the water of the St. Lawrence River it equals 10 per cent, and of the Ottawa River 32 per cent. See for a discussion of the question of potassium in natural waters the writer's Chem. and Geol. Essays, pp. 135-137. m OP CUYSTALMNE HOCKS. 219 cases to do either with crenitic products or with tlio direct results of sub-aerial (h>cay. § 89. Fouc^ud and Michel L(ivy, in tlioir recent experi- ments, have shown us how to form artiliciully, from mix- tures in igneous fusion, in which the pro[)ortions of ele- ments, were pre-arranged, crystalline aggregates eontaini'ig leucite with lahradorite, pyroxene, magnetite, and spinel, and others holding chrysolite in similar associations. The problem which lies behind their discovery is to deter- mine how the materials are so grouped in nature's labora- tory as to yield the mixtures necessary, in the ( 'ic case, for the production of a leucitophyro and, in the other, for a chrysolitic dolorito. The research of the natural processes by which these combinations are reached has been the object of the preceding ini^uiry hito the results of elitiua- tion, on the one hand, and of the solvent and replacing action of percolating waters, on tlu; other. § 40. It is farther to be noted that the experiments of Fouqud and jNIichel L6vy were made by the slow cooling of mixtures from simple igneous fusion, and the question must here be raised how far these reactions would be affected by the intervention of water; in other words, whether, as maintained by Poulett Scrope, Scheerer, Elie de Beaumont, and many others, water is not always pres- ent in the mass of igneous rocks. So far as experiments go, the process of cooling from simple igneous fusion would seem to be inadequate to account for the origin of manj"- of the minerals of eruptive rocks. Fouqud and Michel Ldvy inform us that they " have vainly sought to produce by igneous fusion rocks with quartz, orthoclase, albite, white or black mica, or amphibole," * although the occasional accidental production of orthoclase as a furnace- product has been noticed. The presence of albite in the recent lavas o^ Santorin in association with lahradorite, pyroxene, and chrysolite has been shown by Fouqud (§ 32), and its probable occurrence in a diabase has been * S3rnth&se des Mineraux ct des lloches, p. 75. I mmmmmmm \' ?: i , li .. U 'i i m- m'l 220 THE GENETIC HISTORY [VI. pointed out by Hawes.* Both orthoclase and albite have, however, been formed in the wet way, at elevated temperatures, under pressure (awie, page 157) ; and pyrox- ene, while readily generated from the products of igneous fusion, was got by Daubr^e by the action of superheated water on glass at the same time with crystallized quartz and magnetite or spinel (ihid.^ V^g^ 148). The frequent occurrence of pyroxene in veinstones, in intimate associa- tion with orthoclase, qvni,rtz, apatite, and calcite, suffices to show its aqueous origin, in common with all of these species. In like manner, magnetite, which is readily formed in fused basic mixtures, is found crystallized with orthoclase and quartz, with apatite and pyrite, in granitic veinstones. Moreover, the fact of its association with garnet, and with zeolitic minerals, in the secretions of basic rocks suffices to prove that magnetite, as well as hematite, may be formed by aqueous action. Chrysolite, also, is produced by igneous fusion, but its presence in crystalline limestone in the form of forsterite, and in massive magnetite as hortonolite, shows that, like the related and similarly associated species, chondrodite, it may be formed in the presence of water (Essay X., § 122-124). § 41. The evidences of the intervention of water in eruptive rocks have since the time of Scropo been too often pointed out to need repetition here. Its elements may even be retained in fused compounds at the temperature of ignition, under the ordinary atmospheric pressure, as seen not only in the hydrate and the acid-sulphate of potas- sium, but in certain vitreous borates of sodium and potas- sium, long since described by Laurent, which at a red heat and in tranquil fusion hold an amount of hydrogen equal to 1.2 and 1.3 hundredths of water, and are, under these conditions, slowly decomposed by metallic iron, with abundant disengagement of hydrogen gas, which burns with a green flame from the presence of combined * See Essay VIII., § 75. VI.] OF CRYSTALLINE EOCKS. 221 boron.* That, under greater pressure, water may be held by other compounds, such as silicates, is undoubted. Hy- drous glasses like pitchstone and perlite are examples of these, and differ from obsidian in containing three or four hundredths of water. § 42. The late researches of Tilden and Shenstone on The Solubility of Salts in Water at High Temperatures throw much light on the geological relations of water. While the solvent power of this liquid rapidly increases, when under pressure, at temperatures above 100° C, they have shown that "the increase of solubility follows the order of tlie fusing-point of the solid." Thus, of potas- sium-iodid, which melts at 634°, 100 parts of water at 180° dissolve 327 parts, while of barium -chlorate, melting at 400°, 100 parts of water at 180° dissolve 526, parts. Of potassium-nitrate, melting at 339°, 100 parts of water at 120° dissolve 495 parts, or nearly five times its weight ; while of silver-nitrate, whose fusing-point is 217°, 100 parts of water at 125° dissolve 1622.5 parts, and at 133° 1941.4 parts, or nearly twenty times its own weight. Of certain substances it can be said that they are infinitely soluble at certain temperatures. This is true of the deca- hydrated sodium-sulphate, which melts at 34°, and nearly true for benzoic acid. This substance, which melts at 120°, requires for its solution 600 parts of water at 0° and 25 parts at 100° ; but when heated in a sealed tube to a few degrees above its fusing-point it is miscible with water in all proportions. These heated solutions, in the case at least of barium-chlorate and potassium-nitrate, are described as notablj' viscous, a condition which indicates that they are perhaps colloidal.f § 43. From these results it is easy to conceive what might be expected at elevated temperatures with mate- * The potassium-borate in question, apart from combined water, con- tained boric oxyd .58.0, potash 16.3, giving the oxygen-ratio 72:5, and tlie sodiuni-borate liati tlie same atomic ratios. Aug. Laurent, Compte Rendu des Travaux de Chimie, 1850, pp. .36-42. t Philos. Trans., 1884, part 1, pp. 23-36. lilt ^'ii i mil 222 THE GENETIC HISTORY [VI. M! ! I :;; I rials as insoluble at ordinary temperatures as quartz or the natural silicates. .A few hundredths of water at several hundred degrees Centigrade would probably convert these into a viscid fluid, from which, as from an anhydrous magma, by rest or by partial cooling, definite compounds might successively crystallize; — the mixture becoming, to use the simile of Poulett Scrope in speaking of lavas, like a syrup holding grains of sugar. From such mixtures par- tially cooled, or from a heterogeneous plutonic mass impregnated with water and not yet raised to the full temperature of solution, or what has been aptly termed "igneo-aqueous fusion," the more soluble portions, re- moved by percolation or by diffusion, we conceive to have constituted the liquids which in earlier times produced the various creuitic rocks. The fact that, as shown b}'- Sorby,* pressure augments the solvent power of water, irrespective of temperature, should not be lost sight of in this connec- tion. The remarkable observations of Tilden and Shen- stone serve to explain and to justify the view of the intervention of water in giving liquidity to various erup- tive rocks, originally put forward by Poulett Scrope, and afterwards ably maintained, anong others, by Scheerer and Elie de Beaumont.f § 44. We have already noticed the banded structure (p. 210) which often results from movement in the extru- sion of more or less differentiated masses of eruptive rocks, simulating that produced by the separation from water either of mechanical sediments or of crystalline deposits. It is important in this connection to distinguish between the latter two processes, and to insist upon the more or less concretionary character of the matters separated from solution, often shown in the lenticular shape of beds of this character, and well displayed in the crystalline schists. * Proc. Roy. Soc. London, xii., 538. t Scrope, Jour. Geol. Soc. London, xii., 326. Scheerer, Bull. Soc. Geol. de Franre, 1845, iv., 468, and filie de Beaumont, ibid., 1249 et seq. See farther the author's Chera. and Geol. Essays, 188-101, and also 5, 6, for farther references to the literature of the subject. N VfcJ OF CRYSTALLINE FOCKS. 223 The conditions under which these were laid down from water were less like those of ordinary sediments than of the accumulations of crystalline matter in geodes and in veins. Many facts with regard to the banded character of mineral veins are familiar to geologists, and the strati- form character of such deposits has often been remarked in smaller vein-like masses. I have elsewhere called attention to the fact that crystalline masses having the relations of veinstones may assume great proportions, and that much granitic rock often regarded as eruptive is really of concretionary and endogenous origin, and discussed the question at some length in 1871.* Veins of this kind were then described, sixty feet in breadth, traversing the gneisses and mica-schists of the younger gneissic or Mont- alban series in New England, often coarsely crystalline and banded, and evidently concretionary, but sometimes so finely granular and homogeneous in portions as to be quarried for architectural purposes, like the indigenous gneisses of the series, which they often closely resemble. Remarkable examples of the same phenomenon are to be met with in the older gneissic or Laurentian series, some of which are concpicuous in the sections of these rocks visible in the caflon of the Arkansas Klver and elsewhere in Colorado. Still more striking examples are met with in the similar gneisses in parts of Canada, and are well displayed in Ottawa County, in the province of Quebec, where, in the township of Buckingham, veins eighty feet in breadth, and made up almost wholly of orthoclase and crystalline cleavable magnetite, traverse for considerable distances the stratified gneiss of the region.f § 45. In the same County, and near the Riviere aux Li^vres, are the great veins which have lately been exten- sively mined for apatite in what is known as the Lidvres district. Very similar veins also occur a short distance * Granites and Granitic Veinstones, Amer. Jour. Science, 1871. Granites, Chem. and Geol. Essays, pp. 191-202. t Geol. Report of Canada, 1863-66, pp. 20, 215. .' ' 'i THE GENETIC HISTORY I 1 \^ . vn. THE GENETIC Iii»xv.x.^ 224 ^ . ^g to the southwest, along ^^-f^^tt Utou^Sct. of Ontario, in what may be «"«'' described by the The veins in «* '^^raXC-ntly in 1868, in 1866, writer as eariy as 1848, and subse| J ^^ Ind in 1884.- The .« ttn ide'.ed together, w.l the two districts which may be ,^^ ^^^^ ^,^^^ j eerve to iUnstrate '"'"/J^pa' associates of the apatite evystalline rocks Tlie punc^pa _ ortlioclase, in tliese districts are Py'^^^^^'^l.^ia^ot the localities in quarts, calcite, and pyae. It w .lamination m ?1K Rideau J«t"f • "l2 i'p^ it occurs in - fl--\;» .e shows that the a p ^^^^^ „ ^^j^j^ a ::eh ...e shows that ''^^ '^^ JwalW while "a the stratification, "»'!,'';'• '^„eial contents is often yery banded -angeme", "U riouTminerals „,,a ,o.netimes well marked; -*"^J,__ „£ which the calcite, often occurring in alternate layers, <>' ^^^^^ „j a with i»<=l"'l«'l ,r tmS to<= '<'"^" Farther exam- coarsely crystathne 1^»^^ ' ^^^^ hilateral symmetry of pies were then given »™^™8 ; ^l presence in them of Lnv of the veins, and *« occas o i' „ ^-^^^ of drusV cavities. Moreover, althou ^ ^^^ ,.,^^^. apatae were observed ni what w re eg ^^^ ^^^^^ ^^^^^ stone beds of the enclosing /^^^'^^ ^^^^ , ^ ^,_^ ^^.^tions, workable deposits of »?'''»;«• J„ guch were the conclu- ai-e confined to the veins on ■ J ^ ^^^^ ^^^^ sions announced by the w er ^^ ^^.^ j^^^. .luently, in ^'f'^^^J^'tt^M^^o.e^^'^" ^rT:l district, he was led to w"te Hi ^^^^. j,^^ strata. apatite are in S^'^^t pait '« '"^^ fagmcnts of the wall- a^nd sometimes including anguta g ^^ ^^ g,,,; of ,o„k,-«Meh IS *e clu^actens ^^^^^_^ ^^ Uiterstratified the region,- they aie ' _ 132, ,„d for 1863-66, , Geol Survey of Cauada Kepo't forl8«, P^ ^ ,^,_ ^„, ^ „/2?^2»°»'»««*Syo"" "^ju heln to an understand- reLn (wWch are not mn ed) w.U ne ^^^ ^^ ^^.^^ ^ :|rf the nature and -'';»," k mine, these lesser fo? apatite. As seen at the High .^ ^-^^i, and vein, are fron. a few '"**' *'! ^^'grnatite, often includ- ^e chiefly of a hinary granite o^J" ^^^^ „ear their ■^g portions of *e ^^'^tl^r of two or three feet, a boriers presenting, *"",^ t^'J^ents of gneiss fro™ o;« veritable hveccia "t '"ff '^"^^^ ^^.„^^„„, ,„« „de8 to six inches in ^l'''">'=\'', ; J^ ^te and the other reddish two feldspars, one ™'="g J„ieavable masses. A little the latter forming cons deaWec ._^ ^e vems, „Wte mica is ^'f .^""f^^'ion, hold portions of green whicii, in parts of ^^'^J.t!"!., slender strings runmng cleavaUe vy^'^^T'^tZMs filling the greater part with tlie strike, hut » ."""^ . ,?" ',,,„s of white feldspar, If the vein, and "f »'!»'? ""^X^ fine and large crys- ^ul small masses of greenish apatite^ „„,eover, occa- ™ of which, and others of B™ ^"Ji " ,„itio veinstone. Inally found directly imbedded in ttie g . _^ ^^^^^^^ ' Td Veins of vitreous quartz afoot or „^^„„,, J met with in the i™— ™*of feldspar, hy an ' enclose crystals "^ =>?*"'' "'^^ i„to the binary granite admixture of which they S'^'f ""'^ ^"^t ^ transition from ■ Z pegmatite. There is *»^ J'^ J7„„e essentially pyr- pje quarts '" ^ 8™"Srbt ring' pa«e, ««* ofitse^ oxenic, each occasionally bea""8 J^ ^,, associated m the also forms rock-masses, .f \°"^ti bands or irregular larger veins, sometimes ^^ »^^<=™ Vckness, but at other lenticular masses a few 'n*^ " j^^j each. A frequent times attaining toea**» «*. "^e vri"^'""^^ °°"'"'vf intermediate type of '-'* m toe _ ^^^^^^ VI.] OF CKYSTALLINE EC IS. 227 li ill color, but occasionally bluish, and with cleavage- planes an inch in breadth. The quartz and feldspar in this aggregate sometimes predominate, offering a transi- tion into the granitic rock already noticed, which fre- quently includes crystals of pyroxene, apple-green or grass- green in color, and then sometimes holds clove-brown titanite, brown tourmaline, and, more rarely, zircon. § 49. These rocks, essentially made up of feldspar, quartz, and pyroxene, were long since noticed by the writer as occurring among the Laurentian gneisses in the Rideau district, and at various points in the province of Quebec, and were described in 1866 as generally " grani- toid or gneissoid in structure, sometimes fine-grained, and at other times made up of crystalline elements from two tenths to five tenths of an inch in diameter. . . . They are often interstratified with beds of granitoid orthoclase gneiss, into which the quartzo-feldspathic pyroxenites pass by a gradual disappearance of the pyroxene." The occasional presence in them not only of titanite, but of mica, amphibole, epidote, magnetite, and graphite, was then noticed, and attention was called to the fact that these mineral species are common to the pyroxenite rocks and to associated crystalline limestones. The feldspar of these intermediate rocks was described as having gener- ally the characters of orthoclase, as was shown by the anal- ysis of a specimen from Chatham, Quebec, but as in some cases triclinic and resembling oligoclase.* Dr. Harring- ton has since found for one of these the composition of albite. As will appear from the language just cited, these aggregates were then regarded as portions of the country- rock. The pyroxenite seen in North Burgess, in the Rideau district, was described as sometimes granitoid and at other times micaceous and schistose, interstratified Avith what was then called a binary granitoid gneiss, and also * Geology of Canada, 1803, p. 475; also Report Geol. Survey of Can- ada, 1S63-66, pp. 185 and 224-228. ' 'l r . *. ! H li! ;i THE GENETIO HISTOllY [VI. 228 , ^gj^^ holdirg one case ^o^^^^^TSl tUe strike, and was m p.its hundred and fifty leet w , , •„ two feet in thicV.ness. ^ ^^ i^ crystals, m §50 While apatite f ^%f ^^^ Ji^ the caleareous and scapolite, Py^^"^ . of these secondaiy '^*'l. i^-) ^yas *" venous oha. c^er o^ ^,^ ^ 3 tot it also intersect the rea ^^^^.^ ^ i^t^r pe"0 ^ became "Wf " .....ifom mass«« m . jlo-iting these *° *^ Ca ^ ftt that the P-«-\t:io Ss. and were enclosed , in repeated m *«»« , ^1,^0 the mineral »pec>es had bee l^.^.^ ^^^^^_ „„t fess tha that the Py'«r .,to" e masses, were portions g toterstratified l"»;*"^i„aes.« . .,,, found -trrtXHtrgeoiti^M-y^^^^^^^ who, m 184^' \^iune rocks m northern i • ^j^^^es, origin, a view which ^vas P^^^^^^ ^iX and Ltio relations ot tne v^^ Leonhard, bavi, ^i gnostic reia ^^j yon i^« of certain ^^^rLC::t tau^;- ;^-trr^.V Bana, « For an ai^alysis 01 ^nd CUem. and (.eoi. Jour. Science, 187^ !"•» VI.] OF CRYSTALLINE ROCKS. 229 n. d re 3d )m ,se, lUe lich was ,t it ided ^■euis :bese and the great [\ ir found inions, ers oi )rk, i-e- )U-ores, utonio nt geo- was in ivi, and £ certain ). Dana, see Ainer. who supposed that some of the so-called primary lime- stones " were of Igneous origin, like granite." The aque- ous origin of similar calcareous masses in Scandinavia had, however, been recognized by Scheerer and by Dau- br^e, and in Germany by Bischof, while the vein-like character of certain aggregates of this kind in which various silicates and other mineral species arc associated with carbonate of lime, in the ancient gneisses of North America, had been noticed by C. U. Shepard, H. D. Rogers, and W. P. Blake, among others, as was shown by the writer in some detail, in 1866. In a paper then read before the American Association for the Advancement of Science, it was said that deposits of carbonate of lime, sometimes of great dimensions, and holding the charac- teristic minerals of the crystalline limestones, are found filling fissures and veins in the Laurentian gneisses. These were then designated endogenous rocks, regarded as of aqueous origin, and to be carefully distinguished from intrusive or exotic rocks.* The subject was discussed in the same year in an ac- count of the mineralogy of the Laurentian rocks, when it was said, in commenting upon clie view of Emmons that such masses, and in fact all of the crystalline lime- stones of the series, are eruptive : — " The greater part of the calcareous rocks in the Laurentian system in North America are stratified, and the so-called eru^jtive lime- stones are really calcareous veinstones or endogenous rocks, generally including foreign minerals, such as pyrox- ene, scapolite, orthoclase, quartz, etG."t I had not at that time as yet discovered that these same endogenous masses may include, besides calcareous bands, others essentially quartzose, pyroxenic, and feldspathic, resembling more or less the strata of the enclosing gneissic series, nor con- * Proc. Amer. Assoc. Adv, Science, 1806, p. 54; also Can, Naturalisi (IL), iii., 123. t Report Geol. Survey of Canada, 18(5.3-66, p. 194. See also the facts resumed in Chem. and Geol. Essays, p. 218. il i qm i i :!; !fe- H I ' II 1 THE GENETIC HISTORY IVI. ^^^ *** iv.nds mii'l^t sometimes , ^ T d Burbank a L u rentian type, to whie - 1 ^^^ '^^,„i„„a tke enJogenous „t iutevvals for twenty-iivo mi , j ^^^;^^i the Box'o,-c„gU, and f '™;,''; t 'luestion of tUe»e vem- attention of Mr. Buvbank o _^^1 p„t,Ucatio„3 0U866. like masses. l^^^^«^^^^ „tee»atio„s, UaO, - ^f ; ^ He, as a result ot laiia« Umestonea of the le^i "tded himself "-' f "^J ,g'' „„ks, not erupt.ve, b« were newer than the <^f ^",f "f „i„g fissures in the ■ :7 a vein-like ctaacter -""l^y^/uong, in oer ani g,e-,ss, of wluoh charac ^J\„e tended str«ct«e rases, Ki-.e evidence.! ""."^.rious enclosed mmeials, vi b ein the arrangement of the vano ^^^^,^^ ,^^,,to.o. iLa described them 111 IBH""' , ;on under Inisiu Canada, and enumera ed » ^^«, J(,„,terite ov so-called boWomte), pWog P .^ .^^,^g,,,„ ^nds 01 Wsides serpentine m g'*'"\" .^^ „£ chrysotile. I0 titanite. t B Perry at the same time and itocej iottel""-lf"o!aui.to-^^^^^^^ ^ii! i iL VI.] OF CRYSTALLINK ROCKS. 231 New York, though possessing " the form of dikes," " have a vein-like structure, and sliould be regarded us true vein stones." He farther says of these deposits: "The foliated structure, with its accompanying series of mineral sub- stances, each occurring in a determinate order, evinces tliat the process of deposition was gradual and probably long continued." Thus these observers, in 1871, had, although without acknowledgment, confirmed my obser- vations, and adopted my conclusions of 1866, as to these endogenous calcareous masses of the ancient gneissic series. J. W. Dawson had in 1869 recognized Eozolin Canaihnse in a serpentinic limestone from Chelmsford, and both Burbank and Perry maintained thai all of the limestone nuisses of the region were veinstones, as an argument against the organic nature of Eozoon. § 54. The mineralogy of these endogenous, more or less calcareous masses, has been the subject of much study. While sometimes having the aspect of a coarsely crystalline limestone, and nearly pure, they may include apatite, fluorite, chondrodite, wollastonite, amphibole, pyroxene, danburite, serpentine, phlogopite, gieseckite, orthoclase, scapolites, brown tourmaline, idocrase, epi- dote, allanite, garnet (sometimes ehromiferous), titanite, zircon, rutile, spinel, volcknerite, corundum, menaccanite, magnetite, hematite, pyrite, and, more rarely, pyrrhotite, chalcopyrite, sphalerite, molybdenite, and galenite. To these must be added prehnite, stilbite, chabazite, and ba- rite. All of these species have been met with in the deposits studied in Canada and New York, while in the similar calcareous masses in eastern Massachusetts chrys- olite and petalite occur. Exceptionally, as in Frank- lin and Stirling, New Jersey, there are found in this connection zinciferous and manganiferous minerals, as willemite, tephroite, spartalite and franklinite.* * For farther and more detailed accounts of the occurrence of the mineral species already mentioned, and many others which are found with the calcareous masses of the Laurentian rocks, see Report Geol. 232 THE GENETIC HISTORY [VI. id; ■I- f ■; }( |i) The various associations of apatite in those aggregates are wortliy of notice. Crystals of this species have been observed by the writer directly inii)e(lded in the <[uartzo- feldspathic veinstone, in vitreous (|uartz, in calcite and dolomite, in pyroxene, in crystals of phlogoi)ite, in pyrite, in magnetite, in H[)inel, and in foliated gra[)hite, as well as in a massive granular apatite, which sometimes surrounds large and well delined crystals of the same species. Dr. Harrington has farther noted its inclusion in ami)hibole, in orthoclase, in sca[)olite, in steatite, and in lluorito. On the other hand, apatite crystals have been found to enclose quartz, calcite, fluorite, phlogopite, i)3'roxcne, zircon, titanite, and pyrite. The apatite of these dei)()sits, so far as known, is essentially a fluor-ai)atite, containing in one case, by the writer's analysis, 0.5 hundredths of chlorine. From these facts it is evident that the succession t)f species in these veins is by no means invariable. Mention should here be made of the apatite occurr.iig in disseminated grains in the great deposit of magnetite so extensively mined at Mount Moriah, New York. TiiC banded arrange- ment of the crystalline apatite, generally reddish in color and, in thin layers, occasionally predominating, gives a stratified aspect to the iron ore. A similar aggregate is found in the llideau district, in Ontario. § 55. The stratiform character of these endogenous deposits, as seen alike in the individual portions, and in the arrangement of these as constituent parts of a vein, is well shown at the Union mine, in the Lievres district. Here the great mass or lode is seen to be bounded on tlie west by a dark-colored amphibolic gneiss, nearly vertical in attitude, and with a northwest strike. Within the vein, and near its western border, is enclosed a fragment of the Survey of Ca .ada, 1863-66, pp. 181-229, which was reprinted, with the exception of tlie last six pages, in the Report of the Regents of the Uni- versity of New Yorlc, for 1807, Appendix E. See also, in abstract, Chem. and Geol, Essays, pp. 208-217, and farther the reports of Dr. Harrington and Mr. J. Fraser Torrance,, cited on page 224. I- I VI.] OF CRYaXALLINE ROCKS. 283 gneiss, about twenty feet in width, wliich is traced some yards along the strike of tlio vein, to a cliff, where it is lost from sight, its breadth being previously nmch diniin- ished. It is a sliari)ly broken mass of gray banded gneiss, with a re-entering angle, and its close contact with the surnnuiding and adherent coarsely granular pyroxenio veinstone is very distinct. Smaller masses of the same gneiss are also seen in the vein, which was observed for a breadth of about 150 feet across its strike, — nearly coincident with that of the adjacent gneiss, — and be- yond was limited to the northeast by a considerable breadth of the same country-rock. § 5G. In one oi)ening on this lode there are seen, in a section of forty feet of the banded veinstone, repeated layers of apatite, pyroxenite, and a granitoid quartzo- feldspathic rock, including portiojis of dark brown foli- ated pynjxene, all three of these being unlike anything in the enclosing gneiss, but so distinctly banded as to be readily taken for country-rock by those not apprised of the venous character of the mass. A fracture, with a latcal displacement of two or three feet, is occupied by a gran- itic vein twelve inches wide, made up of quartz with two feldspars and black amphibole, which themselves present a distinctly banded arrangement. This same granitic vein is traced for fifty feet, cutting obliquely across both the pyroxenite and the older granitoid rock, and at length spreads out, ond is confounded with a granitic mass interbedded in the greater vein. It is thus posterior alike to the older quartzo-feldspathic rock, the pyroxenite, and the apatite, — as are uiso many smaller quartzo-feld- spathic veins, which, both here and in other localities in this region, intersect at various angles the apatite, the pyroxenite, and the granitoid rock into which the latter graduates. We have thus included in these great apatite- bearing lodes, quartzo-feldspathic rocks of at least two ages, both younger than the enclosing gneiss. A small vertical vein of fine-grained black diabase-like rock inter- B .< ''ISM M-r - ;:«., iji 234 THE GENETIC HISTORY pn. sects the whole. No one looking for the first time at this section of forty feet, as exposed in the quarry, with its distinctly banded and alternating layers of pyroxenite and granitoid quartzo-feldspathic rock, including two larger and several smaller layers of crystalline apatite, would question the stratiform character of the mass, whose venous and endogenous nature is, nevertheless, dis- tinctly apparent on farther study. In other portions of the same great vein, which has been quarried at many points, this regularity of arrange- ment is lc33 evident. Occasionally masses are met with presenting a concretionar}"- structure, and consisting of rounded or oval aggregates of orthoclase and quartz, with small crystals of pyroxene around and between them; the arrangement of the elements presenting a radiated and zone-like structure, and recalling the orbicular diorite of Corsica. The diameter of these granitic concretions varies from half an inch to one and two inches, and they have been seen in several localities in the veins of this region, over areas of many square feet. § 57. In the Emerald mine the stratiform arrange- ment in the vein is remarkably displayed. Here, in the midst of a great breadth of apatite, were seen two parallel bands (since removed in mining) of pyroxenic rock, sev- eral yards in length, running with the strike of the vein, and in their broadest parts three and eight feet wide re- spectively, but becoming attenuated at either end, and dis- appearing, one after the other, in length, as they did also in deptl). These included vertical layers, evidently of contemporaneous origin with the enclosing apatite, were themselves banded with green and white from alternations of pyroxene and of feldspar with quartz. Accompanying the apatite in this mine are also bands and irregular masses of ilesh-red calcite, sometimes two or three feet in breadth, including crystals of apatite, and others of dark green {imphibole. Elsewhere, as at the High Rock mine, tremolite is met with. In portions of the vein at the u ar no OF CRYSTALLINE ROCKS. 235 Emer i.ld mine pyrite is found in considerable quantity, and occasionally forms layers many inches in thickness. Several large parallel bands of apatite occur here, with intervening layers of pyroxenic and feldspathic rock, across a breadth of at least 250 feet of veinstone, besides numerous small, irregular, lenticular masses of apatite. TliG pyroxenite in this lode, as elsewhere, includes in places large crystals of phlogopite, and also presents in drusy cavities crystals of a scapolite, and occasionally small, brilliant crystals of colorless chabazite, which are implanted on quartz. At the Little Rapids mine, not far from the last, where well defined bands or layers of apatite, often eight or tei. feet wide, have been followed for considerable distances along the strike, and in one place to about 200 feet in depth, these are, nevertheless, seen to be subordinate to one great vein, similar in composition to those just de- scribed, and including bands of granular quartz. In some portions of this lode the alternations of granular pyrox- enite, quartzite, and a quartzo-feldspathic rock, with little lenticular masses of apatite, are repeated two or three times in a breadth of twelve inches. § 58. The whole of the observations thus set forth in detail above serve to show the existence in the midst of a more ancient gneissic series, of great deposits, strati- form in character, complex and varied in composition, and, though distinct therefrom, lithologically somewhat similar to the enclosing gneiss. Their relation to the latter, however, as shown by the outlines at the surfaces of contact, by the included masses of the wall-rock, the alternations of unlike mineral aggregates, the evidences of successive and alternate deposition of mineral species, and the occasional unfilled cavities lined with crystals, forbid us to entertain the notion, that they have been filled by igneous injection, as conceived by plutonists, and lead to the conclusion that they have been gradually deposited from aqueous solutions. This conclusion is 236 THE GENETIC HISTORY fFh » ■ ^ \i 5) made more apparent when we compare these immense banded lodes with the many small veins from a foot in breadth upwards, also banded, and lithologically similar to the great lodes, which intersect not only these but the ancient gneisses, as already described at the High Rock mine, and also in many other localities, especiuliy in parts of the Rideau district. It may here be noticed that the very similar banded and vein-like deposits now largely mined for apatite in Norway, are regarded by Brogger and Reusch, who have lately studied them, as igneous masses erupted in a liquid condition, and slowly cooled from fusion, a hypothesis by which they e.ideavor to explain many of the phe- nomena of these deposits. For an analysis of their argu- ment and a forcible siateraent of the objections thereto, the reader may consult Dr. Harrington's report on the apatite region of the Lidvres.* § 59. These various endogenous deposits are instruc- tive illustrations of the creuitic process. The alterna- tions of stratiform layers of quartz, of calcite, and of feld- spathic and pyroxenic aggregates, with included layers of apatite, pyrite, etc., show that a process closely analogous to that which formed the older gneissic series was in ope- ration and gave rise to these banded mineral masses in Cue midst of rifted and broken strata of the Ider rocks after these had assumed their present attitude. . The lithologi- cal resemblances between the older and the younger de- posits are not less remarkable than their differences, and suffice to show the great similarity between the conditions which produced the veinstones and their enclosing rocks. The latter, however, appear, in the present state of our knowledge, to have been deposited not only on a vaster scale, but apparently in a horizontal or nearly horizontal attitude. § 60. What are regarded as examples of calcareous de- posits of the two ages were described by the writer, in * Brogger and Reusch ; Zeitschrift d. deutsch. Geol. Gesell. Heft III., pp. 64G-702. Report Geol. Survey Canada, 1877-78, G., pp. 11-12. VI.] OF CRYSTALLINE ROOKS. 237 1878, as occurring at Port Henry, on Lake Champlain, in the State of New York. Near the town is a quarry whence limestone has been got for the blast-furnaces of the local- ity. " Here elongated, irregular fragments of dark horn- blendic gneiss, from two inches to a foot in thickness, were found completely enveloped in crystalline carbonate of lime. In 1877, five such masses of gneiss were exposed in an area of a few square yards. One of these, a thin plate of the gneiss, having been broken in two, the enclos- ing calcareous matter had filled the little crevice, keeping the fragments very nearly in their place. The carbonate of lime, which is coarsely granular, and contains some graphite and pyrite, is banded with lighter and darker shades of color, and one of its layers was marked by the presence of crystals of green pyroxene and of brown sphene. The contact of this mass with the surrounding gneiss, which is near by, is concealed. No serpentine was found in this limestone, though it abounds in a limestone quarried in the vicinity. About half a mile to the north is still another quarry, opened in a great and unknown breadth of more finely granular and somewhat graphitic limestone, which near its border presents three beds of two or three feet each, interstratified with the enclosing gneiss." Of this it was said that "it presents alterna- tions of lighter felds].athic and darker hornblendic beds with others highly quartzose, and includes layers of a sulphurous magnetite, which are, however, insignificant when compared with the great deposit of this ore mined at Mount Moriah, in the vicinity." § 61. While the great breadth of limestone interstrati- fied with the gneiss was regarded as belonging to the ancient series, it was said of the limestone of the first- described quarry that it " seems clearly to be a brecciated calcareous vein enclosing fragments of the gneiss wall- rock." * Reference was then made to similar observations * Azoic Rocks, etc., pp. 166-167; also The Geology of Port Henry, Canadian Naturalist, X. , No. 7. iNIJ! i: I'* Ij.' I 238 THE GENETIC HISa?ORY [VI. in this vicinity, described by Prof. James Ifal. n\ 1876, who, from this breccia of gneiss-fragments in an exposure to crystalline limestone, rightly inferred thf; posterior de- position of the latter, and was led to co^^-jecture that it might belong to a newer geological series. The only evidence of this, however, was the enclosed fragments of the gneiss, which, in similar cases, had led Emmons and Mather to infer the eruptive character of these same limestones, regarded by the writer as endogenous masses or veinstones. The great thickness of the interstratified limestone-masses which form, according to Logan, integral parts of the vast Laurentian series, and their geographi- cal extent, were described in detail in the publications of the geological survey of Canada, in 1863, and farther in 1866. A summary of these results will be found in the writer's volum'i on Azoic Rocks,* and farther on in Essay IX. of this volume. § 62. As regards the genesis, according to the crenitic hypothesis, of the various^ mineral species found in this vast crystalline series, alike in the more ancient strata and in their included endogenous masses, we have already considered the formation of the double silicates of alu- mina with alkalies and lime, represented by the various feldspars, and more rarely by the scapoliles, epidote, gar- net, and the muscovitic or non-magnesian micas. These latter, though abundant, with garnet and black tourmaline, in some granitic veins in this geological series, are rare in those portions in which the protoxyd-silicates abound, — while the silicates of alumina without protoxyd-bases, such as are andalusite, fibrolite, cyanite, topaz, and pyro- phyllite, are unknown. On the other hand, aluminous double silicates with magnesia are abundantly represented by phlogopite, and protoxyd-silicates with magnesia, such as chondrodite, pyroxene, and amphibole, are abundant; the simple calcareous silicate, wollastonite, being more . rarely met with. The genesis of all these we have sup- * Azoic Rocks, p. 154. vi.] OF CEYSTALLINE ROCKS. 239 posed to be by the reaction of soluble calcareous silicates with magnesiau and ferrous solutions. The magnesia required may be found either in salts like those of sefv-water, or in solutions of magnesian bicarbonate from sub-aerial decay of plutonic rocks, which solutions, by reaction with lime-silicates, would give rise to insoluble magnesian compounds and soluble lime-carbonate. A similar reaction, with liberation of silica, would result from the direct operation of carbonic dioxyd upon the lime-silicate. The intervention of ferrous solutions in similar reactions has already been discussed, in consider- ing the origin of glauconite, on page 197. § 63, As regards the presence in these, and similar crystalline rocks, of basic oxyds uncombined with silica or with carbonic acid, such as alumina and magnesia in corundum, spinel, and some chromites, chromic oxyd in the latter and in some spinels, glucina and magnesia in chrysoberyl and periclase, together with zinc, manganese and iron-oxyds in spartalite, franklinite, magnetite, and hematite, not to mention titanic oxyd in rutile and ''n menaccanite and other titanates, it should be noticed tb .'; tliese various compounds, for the most part, occur in sncu intimate association with certain silicates as to suggest their contempoi'aneous production Thus corundum and spinel are found crystallized with certain micas, with chlo- rites or with feldspars, pyroxene or chrysolite, in which latter, or in serpentine, chromite is generally met with. Spartalite and franklinite are associated with silicates of zinc and manganese, and magnetite with quartz, with or- thoclase, with pyroxene, with chondrodite, or with chryso- lite, while rutile and menaccanite are found in like man- ner with feldspars, with phlogopite, or with serpentine. The intimate association of magnetite with calcite, with apatite, with pyrite, and with graphite, in these deposits, may also be noticed. We must conclude that all these simple and compound oxyds have been in solution, and have crystallized in the presence of the various silicates, H I "I I K' i 1, Ai I < in ,il "i ' 240 THE GENETIC HISTORY [VI. etc., and in many cases with quartz. It is evident that the partial reduction and solution of ferrous oxyd by the intervention of the products of organic decay, and its subsequent precipitation, which in later times, has played so large a part in the genesis of iron-oxyds and carbonate, is not the sole agency. A ^u'ocess which separates not only iron-oxyd, but chrome-oxyd, alumina, glucina, mag- nesia, and zinc and manganese-oxyds, from their silicated combinations, and has permitted them to crystallize side by side with silicates, and even with free silica, has inter- vened in the genesis of these ancient crenitic deposits. The solvent action exerted by solutions of alkaline sili- cates on oxyds of iron, manganese, zinc, magnesium, and calcium, as well as upon those of tin, antimony, copper, and mercury, throws, as elsewhere pointed out, an impor- tant light on this problem (pages 150, 181). To this wo must add the dissociation of silicate of alu- mina at elevated temperatures, under pressure, in presence of alkaline solutions, with separation of silica in the form of quartz, as observed both by Daubrde and Henri Sainte- Claire Deville (pages 148, 156). These experimenters obtained at the same time zeolites, and one of them pyr- oxene, apparently with magnetite, while Friedel and Sar- rasin, under similar conditions, got orthoclase and albite, quartz and analcite. We are as yet ignorant under what circumstances the liberated alumina might be separated from these solutions as corundum or diaspore. The con- ditions of temperature, and the presence of alkaline solu- tions in these experiments, approach very closely to those which we have supposed to concur in the formation of mineral species by the crenitic process. § 64. The geognostic and genetic history of the great endogenous crystalline masses which we have now dis- cussed in some detail is important for several reasons: 1. It brings before us the views of the plutonists, who see in great bodies of crystalline limestone, and of magnetite, as well as in granitic veins and in metalliferous quartz- VI.] OF CRYSTALLINE ROCKS. 241 lodes, the evidences of igneous eruption. 2. It shows the differences, alike mineralogic and geognostic, between true exotic rocks (which, with small differences in composition, have been erupted through widely separated geologic ages up to the present) and those endogenous deposits which are found only in eozoic rocks, and were formed in eozoic time, since their fragments are met with in the oldest overlying paleozoic sediments. 3. It makes evi- dent the close mineralogic resemblances between these endogenous crystalline masses and the more ancient en- closing rocks, and thus helps us to a clearer conception of the conditions under which these ancient gneissic strata, and the pre-gneissic granite itself, were generated. § 65. The crenitic hypothesis, as we have seen, sup- poses that the granite, and the succeeding crystalline schists, have been built up by matters dissolved from a primary plutonic substratum, upon which, as upon a floor, through successive ages, was laid down the enormoas thickness of crenitic rocks which, with small exceptions, make up the pre-Cambrian terranes. The bearing of this hypothesis upon the great problem presented bj- the cor- rugated condition of the older crystalline schists has already bepu noticed on page 179. The contraction of a cooling globe, which is often cited in explanation of this phenomenon, is clearly inadequate to account for this great and general corrugation of the strata, and the present writer in 1860 * suggested, as a farther element in explanation thereof, the condensation during crystalliza- tion of the mechanical sediments from which, in accord- ance with the Huttonian hypothesis, the crystalline schists were supposed to be derived. This explanation, based on an untenable hypothesis, must, however, be rejected. The endoplutonist must appeal to contraction in the igne- ous mass of the globe as the only explanation of the corru- gations of its outer envelope, while the exoplutonist adds * Amer. Jour. Science, xxx., 138, and Chem. and Geol. Essays, pp. 56, 71. I:- M: ^ liilil^ 242 THE GENETIC HISTORY LVI. thereto the diminution of the I'quid interior as the result of successive transfers of portions of its mass by ejections of igneous material from beneath a first-formed crust. Against this latter c ^lanati- i it is to be urged that, as we have endeavored ';;> sji. vv, the successive groups of stratiform crystalline i'n:kA v 'ch have been laid down on the pre-gneissic granite, id tt^en this primeval granite itself, are not igneous but aqueo .. in origin, so that the exoplutonic hypothesis itself is untenable. The amount of plutonic extravasation in pre-Cambrian times was ap- parentl}'' small. § QQ. The crenitic hypothesis, however, admits a trans- fer of matters from below upwards, in a state of solution, and the building-up from them, upon the solid floor of igneous rock, of the granite and all the succeeding crystal- line schists, as in the scheme of the exoplutonists. This new aqueous hypothesis thus offers, it is believed, for the first time, a reasonable and tenable explanation of the uni- versal corrugation of the oldest crystalline strata. The earth, according to this hypothesis, although intensely heated, had not, even at the early time when the waters were first condensed on its surface, a liquid interior, but was solid ; and its crust is supposed to have presented no variations in composition, except such as might result from crystallization and eliquation in a purely igneous con- gealing mass. The superficial quartzo-feldspathic or gran- itic layer, which is believed to overlie everywhere the quartzless basic doleritic rock, did not then exist, but has since been derived by crenitic action from the primary plutonic layer. This granitic stratum is, however, itself still subject, like the basic stratum beneath, to softening under the combined influences of water and heat, and to extrusion in the forms of eruptive granite and trachyte ; although it is less fusible, and, consequently, less suscep- tible of differentiation by eliquation. It is, moreover, at the same time, less liable to alteration by lixiviation, from the fact that it is not a mass cooled from igneous fusion. '1 ii VL] OF CRYSTALLINE ROCKS. 243 dno from con- Tran- the At bas iTiary itself eiiing and to chyte; ,u3cep- jr, at n, from fusion, but one deposited from water at comparatively low tem- peratures, and thus lacks the porosity which belongs to the original pluLonic stratum. § 07. The upward transference of the vast and un- known quantity of material constituting the ancient granitic and gneissic rocks, which are at least many miles in thickness, and the contraction of the plutonic substra- tum, diminished by the removal of this great mass, would necessarily result in great movements of subsidence, with plications and fractures of the gneissic strata. We are, of course, ignorant whether these processes went on to a uniform degree over the whole surface of the earth, and whether similar conditions of thickness, and similar corru- gations, exist in those great portions of t'le eozoic crust which are concealed beneath the ocean's waters, and be- neath accumulations of newer strata. It may well be that the plication of the ancient granitic crust was, as in the case of younger stratified rocks, limited to certain areas. It can only be affirmed, in the present state of our knowl- edge, that in the relatively very small areas of the oldest gneissic rocks known to us, this plication is great and ai)parently universal, diminishing, however, materially in degree, in the younger gneissic series. § 68. Within the fractures and rifts of the ancient gneissic strata resulting from these great movements, the products of the uninterrupted crenitic process would henceforth be deposited, filling them with masses closely resembling those of the enclosing strata. Repetitions on a smaller scale of these movements would give rise to newer fissures intersecting alike these strata and the first- deposited veinstones, in the manner shown in our studies of the Laurentian rocks, where the process which pro- duced the original quartzose, feldspathic, and calcareous deposits of the series was repeated at least twice, giving rise to primary and to secondary veinstones mineralogi- cally very similar to the first-formed or country-rock, and therebv showiujir the survival of the original chemical con- ■i. ^ ■ m mm I M ' '■{■ ■■( i-m ^^ Jill ! ■li i 244 THE GENETIC HISTORY [VI. ditions of solution and deposition after one, and even after two movemeiits of displacement and disruption in the region. § 69. We have thus endeavored in the present essay to bring together, in the first place, a number of facts wliich serve to throw light upon the generation of mineral sili- cates by aqueous processes, especially in later times, sub- sequent to the formation of the great series of crystalline schists, and thereby help to a better understanding of the crenitic hypothesis. We have next C(Misidered the two plutonic hypotheses as to the origin of crystalline rocks, and have discussed the question of stratiform structure in rocks whose eruptive character is undisputed. This has led us to consider the process of differentiation in such masses through partial crystallization and eliquation, and, farther, to a discussion of the possible relations of waior to the process. The secular changes which may be wrought in igneous masses by aqueous percolation are next discussed, with reference at the same time to the crenitic process. From this we are led to a discussion of the stratiform structure seen in vein-like masses for which an igneous origin is inadmissible, and which, it is maintained, are endogenous deposits of crenitic origin. An account of these, as they have been observed in the ancient gneissic rocks of North America, leads to a farther consideration of the crenitic hypothesis, alike in relation to the genesis of the silicates, carbonates, riud non-silicated oxides of the crystalline rocks, and also to the general plication of the ancient crystalline strata. § 70. The conclusions from this extended study are, briefly, as follows. The quartzless basic material which is supposed to have constituted the primary plutonic mass, and is the direct source of basaltic and doleritic rocks, has been subject to modifications from three agencies : — 1. The solvent action of permeating and circulating waters, which, from parts of it, have removed alumini, with preponderating proportions of silica and potash, — the ele- y^ OF CRYSTALLINE HOCKS. 245 I of the tlie ating with lie ele- ments of granitic, trachytic, and gneissic rocks, — and also silicates of ahunina and otlier protoxyds, 'wliicli have been more or less directly the sourco of the other silicated species, of the oxyds, and in part also of the carbonates of the crystalline schists and veinstones : — 2. The farther action of the same circulating waters in carrying down from the surface, alike in the condition of carbonates, formed by sub-aerial action, and of sulphates and chlorids, large ])ortions of calcium, magnesium, sodium, and potassium, — all of which, by interchange and re- placement, have variously modified the composition of the plutonic material : — 3. The process of differentiation in portijns of the plutonic mass by partial crystallization and eliquation, thereby giving rise to more chrysolitic and more pyroxenic aggregates on the one hand, and to more feldspathic aggre- gates on the other, — a process in which it is conceived water may intervene, giving to the material an igneo- aqueons fluidity. All of these agencies, it is believed, have, from the earlier ages, been at work on the plutonic substratum, causing secular changes alike in the crenitic products derived therefrom, and in the residual portion, from which have come, and are still derived, the basic eruptive rocks. ^ Appendix. The genesis of dense crystalline species in less dense colloidal fused inagraas, wliether hydrous or anhydrous (pp. 209, 222), not only involves the disengagement of heat, but, as Becker has shown ( Amer. Jour. Science, xxxi., 120), its disengagement at the maximum rate, tiius maintaining the liquidity of the crystallizing magma. The passage of certain dense species, when fused per se {post, pp. 299, .'^00), into vitreous or crystalline forms of less specific gravity is no exception to the law of condensation, since the chemical and physical conditions are unlike those of the more complex magma. When such a magma, holding combined a portion of water, is changed into anhydrous species, this will be liberated, as appears in the often observed disengagement from solidifying lavas of aqueous vapor, sometimes with boric oxyd, fluorhydric and chlorhydric acids, and various chlorids. Hence silicates like epidote, tourmalines, and micas, which contain such volatile elements, will only be generated imder con- ditions which prevent their liberation. VII. ■J ! ^ ,■1 1 1 j Jimr mm- II i THE DECAY OF CRYSTALLINE ROCKS. This essay, presentcil and read In abstract to the National Academy of Sciences at AVashingtun, April 17, 18M3, was publiHhed under the title of " The Decay of Uouks Geologically Considered " In September of the same year, In the Anierioau Journal Of Science, [III.], Jtxvl., 190-213. § 1. The subject of the decay of rocks has not yet received from geologists all the attention which it merits, and there still appear to be misconceptions with regard to it which warrant us in reviewing some points in its his- tory. F. II. Storer,* in a recent notice of a suggestion of Nordenskiold as to the liberation of gems thi-ough the decay of the feldspathic rocks in which they are often con- tained, cites with approval the opinion of Professor Stubbs of Alabama that "the decomposition of these rocks in southern latitudes has proceeded much faster than with the same rocks in higher latitudes," a " condition which can be accounted for, to a large extent, by climatic influ- ences." The cold and frost now prevailing in northern regions are supposed by him to retard the action of atmos- pheric waters, regarded as the chemical agent of this process of decay.f These views, implying that the pro- cess is one belonging to the present time, are accepted by Storer, who writes of the "more active and thorough- going disintegration which occurs" in these southern regions. § 2. That the presence in the northern hemisphere of a mantle of softened material, from the decay in situ of • Science, for Feb. 16, 1883, p. 29. t Bemay's Hand-book of Alabama, 1878, p. 199. 246 of no t0| vie as sii eryl foul prel t. Vll.] THE DECAY OF CUYSTALLINE ROCKS. 247 of a \tu of crystalline rocks, is more common at the outcroi)8 of these ill low than ill high latitudes, where it is often entirely absent, is a familiar fact; but it will, I think, be made evident that present climatic ditterences have nothing lo do with the fact that similar rocks are in one area covered with a thick layer of the products of decay, and in an- other are wholly destitute of it. § 3. The decay in question is well known to be duo to a chemical change of which the predu inant mineral sili- cates of t'"e rock, chielly feldspars and amphibole, are the subjects, and which results in the removal by solution of the protoxyd-bases, together with a great proportion of the combined silica, leaving a residue essentially of clay, mingled with quartz, garnet, magnetite, and such other mineral species as resist the process of decom[)osition. § 4. A memoir, by Fournet, published in 1834,* gives many facts regarding the early observations on rock-decay. Its author there describes the wide-spread decomposition of the granites near Pont-Gibaud in Auvergne, a change which Ueribier de Cheissac had already shown to be an- terior to the deposition of the tertiary rocks. Fournet, moreover, noticed the similar decay of basalts, phonolites, trachytes, and even obsidians, and described the process of exfoliation, by which rounded masses of undecayed rock are left. He cites in this connection the observations of Pallas, who, in his travels in Siberia (17G8-1774), noticed hills " that seemed composed of masses heajDcd together, as it were rounded by decomposition." The view of Werner, that the rounded form of masses such as these was due to a-iginal concentric structure, was rejected by Fournet. § 5. In 1818, Messrs. J. F. and S. L. Dana dascribed a similar phenomenon in the decaying greenstones at Sora- erville, near Boston, Massachusetts, where the rock was found to be converted by deca}' in situ into nodular masses presenting exfoliatincr concentric laj'ers of differing degrees * Ann. de Ch. et de Phys., [2], v., 225-256. k ' ' THE DECAY OP CRYSTALLINE ROCKS. [VII. of decomposition. These masses rest upon each other, the decayed material filling the interstices.* In 1825, J. W. Webster noticed the same example, and explained the formation of boulders by the exfoliation of the decayed greenstone.! Again, in 1858, W. P. Blake described the production of rounded masses both of sandstone and of granite through disintegration. He explains how angular blocks, separated by joints admitting water to all sides, would be " attacked most i apidly on the angles, thus pro- ducing a succession of curved faces gradually approaching a sphere," and illustrates the process by figures. He described, moreover, the boulder-like masses of granite in Placer County, California, lying on an uneven surface of the same rock, " as due to the manner in which the rock decomposes, and not to abrasion." Like Fournet, he rejects the notion of an original concentric structure in the rock.J 6. Hartt, in 1870, discussed the well known exam- ples of rock-decay found in Brazil, and called such rounded masses of rock as we have just described " boul- ders of decomposition." He moreover noted that the process of decay was there anterior to the supposi 1 glacial action, which had worked over the material of the previ- ously decomposed rocks. § Lyell already, in 1849, had pointed out that the tertiary clays and sands of the southern United States have been derived from the waste of the previously decayed crystalline rocks of the region ; || and, as we have seen, the ante-tertiary dge of the decay in Auvergne had long before been recognized. § 7. The account given by Charles Upham Shepard, in 1837, of the origin and mode of occurrence of the porce- lain-clays of western Connecticut is remarkable for its exactness and perspicuity. That at New Milford is de- • Mem, Amer. Acad. Sciences, 1st Series, iv., 201. t Boston Jour. Philos. and Arts, ii., 285. t Geol. Recon. of California, pp. 146, 286. § Scientific Results of a .Journey in Brazil, pp. 28, 573. II ijyell, A Second Visit to the United States, ii., 28. ' of til « t t »♦ Ill in L'ce- its de- vil.] THE DECAY OF CRYSTALLINE ROCKS. 249 scribed as occurring " upon the western slope of an ele- vated range of granitic gneiss. ... In many places the decomposition of the parent rock is so complete as to present the aspect of n secondary deposit ; but the pre- vailing appearance is that of the rock altered in place through the decay of the feldspar and mica. Indeed, the same relative arrangement of the quartz and the altered feldspar is observed in the bed as is presented by these materials in the undecomposed rock. Veins and seams of a perfectly impalpable white clay traverse the rock in various directions, analogous to the veins of feldspar in the granite of tlie neighborhood." Of a pure white clay in the town of Kent, our author says, " It forms a vein many feet in width, cutting through quartz rock. It owes its origin to a graphic granite, which must have been free from mica." A similar vein of clay is described as occur- ring in the town of Cornwall, and as including frequent crystals of black tourmaline ; the feldspar also being incompletely decomposed.* § 8. As showing that the process of sub-aerial decay is not confined to silicated rocks, it may be noted that J. D. Whitney described, in 1862, the existence, in the lead- region of Wisconsin, of a layer of red clay and sand, mixed with chert, sometimes thirty feet in thickness, which lie showed to be a reriduum from the secular decay of several hundred feet of the impure paleozoic lime- stones of the region. f A like occurrence was afterwards, in 1873, described by Pumpelly, in southern Missouri, where such residuary deposits sometimes attain a thick- ness of 120 feet.J This process is evidently due to a simple solution of the carbonates of lime and magnesia in meteoric waters. § 9. A similar decay is conspicuous along the outcrop of the Taconian limestones and their associated schists in * Shepard : Geological Survey of Connecticut ( 1837), pp. 73-75. t Geology of Wisconsin, i., 121. t Geological Survey of Missouri : Iron Ores and Coal Fields, p. 8. WU^mmt mmm:<^'^::v V'r '. : ■' i : n ■ I '' II Ml 250 THE DECAx' OF CRYSTALLINE ROCKS. [vn. the Appalachian valley, as will be noticed farther on, in § 25, and may also be seen at several points in the Tren- ton limestone and the Utica shale of the St. Lawrence valley. One of these localities, described by J. W. Daw- son, is at Les Eboulemens, on the north shore of the St. Lawrence, below Quebec. Here, at the southwest base of the liigh Laurentide hills, the post-pliocene clays, enclos- ing marine shells and large gneiss boulders, are found resting upon a mass of Utica shale, deprived of its calcare- ous matter, and so soft as to be readily mistaken for the newer clays of the region, but for its stratification and its organic remains. This, according to Dawson, had been changed to a great depth by sub-aerial action previous to the period of submcigence, during wbich it was covered with the boulder-clay.* Some facts connected with the decav of the Trenton limcsi iie near Montreal will be mentioned in § 40. § 10. It may be said that, with the exception of Dar- win, who had observed the decay of rocks in Brazil, and conjectured that the process might have been submarine, all observers have correctly regarded it as sub-aerial. The chemistry of the process was discussed, among others, by Fournet, in the paper already cited, and later by Delesse, in 1858 ; t also very fully by Ebelmen, who considered the question of rock-decay in its relations to the atmos- phere, in two memoirs, in 1845 and 18474 The same sub- ject was further considered at some length by the present writer, in 1880.§ § 11. Having thus briefly indicated some of the points in its h tory during the past century, we are prepared to notice in more detail the contributions made to the sub- ject, regarded in its geological bearings, during the last ten years. Previous to this, as we have seen, it had been * Dawson: Post-riiocene Geology of Canada; Can. Natm-alist, vi., 1872. t Bull. Soc. Geol. de France, x., 256. t Annales des I lines [4], vil. and xiii. § Ante, pp. 30-34: also Chem. and Geol. Essays, p. 100. aeriar great I and c| soil, ordins said, nencyl showii ciesm * Thd t Pr ScifnceJ and Hi red sent lints dto sub- last VII.] THE DECAY OF CRYSTALLINE ROCKS. 251 already recognized that the process of rock-decay was in operation not only in pre-glacial but in pre-tertiary times, and that the resulting material was the source of the tertiary clays and sands, and even, in certain cases, of glacial drift and boulders. § 12. In a review of Hartt's volume on Brazil, in 1 870, the present writer said : " The great wasting and wearing away of crystalline rocks in former geological periods, of which we have abundant evidence, is less difficult to understand when we learn that rocks as hard as those of our New York Higlilands become [are] even in our own time, under certain conditions, so softened as to offer little more resistance to the eroding action of a torrent than an ordinary gravel-bed."* Subsequently, in an account of some observations made in North Carolina, among the rocks of the Blue Ridge, and presented to the Boston Society of Natural Histoiy, October 15, 1873, he expressed the belief tliat the decay of crystalline rocks was a process of great antiquity; that it had been universal; that the cov- ering of decayed material now seen in the south, at one time extended to the rocks of northern regions, from which it had been removed by erosion during successive agt , cul- minating in the glacial period at the close of tiie pliocene, since which time the chemical decomposition of the surface has been insignificant. From the products of this sub- aerial decay, it was then maintained, has been derived a great part of the sediments alike of paleozoic, inesozoic, and cenozoic times. The permeable nature of tlie surface- soil, formed of highly inclined strata of decayed rocks, af- ording a natural subterranean drainage, explains, it was said, both the absence of lakes, and the comparative perma- nency of the surface to be remarked in uneroded regions ; showing that something more than ordinary aqueous agen- cies must have effected the removal of the decayed material. f * The Nation, New York, Dec. 1, 1870. t Proc. Boston Soc. Nat. History, Oct. 15, 1873, and Amer. Jour. Science, vii., 60; also Proc. Amer. Assoc. Adv. Science for 1874, p. 39; and Hunt, Cheni. and Geol. Essays, pp. 10, 250. i 252 THE DECAY OF CRYSTALLINE ROCKS. (VU. § 13. This communication of mine was speedily fol- lowed by a paper published in the Proceedings of the Boston Society of Natural History for November 19, 1873, by the late Mr. L. S. Burbank, repeating and insisting upon the same conclusions, and, moreover, dwelling espe- cially upon the proceHt? of decay (which he also had studied in North Carolina) as a preliminary to the forma- tion of boulders and glacial drift. In accordance with the views thus expressed in 1870 and 1873, it was conceived that the power of the usual eroding agents, ice and water, would be inadequate to the removal of great areas of rock unless this had been previously softened by decay, and in a review of the sub- ject by the present writer, in 1873, the conclusion was reached that the decomposition of rocks has been "« necessary preliminary to glacial and erosive action^ wJiick removed already softened materials.'''' * Such erosion and denudation would, in accordance with this view, consist in the removal of previously decayed rocks, and the forms and outlines of the sculptured surface thereby exposed would be determined by the varying depths to which the process of sub-ierial decomposition had already penetrated the once firm and solid rock. The bash; ji;:-^ depressions and the hillocks of the erodtJ surface, rt.t ;ess than the detached rounded masses or boulders, were thus, as the writer has ever since taught, the results of the previous process of rock-decay. § 14. I had long before this time been led to insist upon the evidences of a widely spread decomposition of crystalline rocks in very early periods of geological history. In an essay, entitled Some Points in Chemical Geology, published in 1859,t and another, on the Chem- istry of Metamorphic Rocks, in 1863,J both reprinted in iiiyAiiame of Chemical and Geological Essays, I have * TT^rper's Annual Record of Science, etc., for 1873, p. xlvli. t <'t»'0l. Jour., London, x-. , 488-406. t Ceo!. Soc. .our., Dublin, x., 85-95. proijd varioif the dl like A "uisccj ophylf tJiat "P ferent and ti study tion a Insist m of igical iuical tbem- >d in \iave vn.] THE DECAY OF CRYSTALLINE EOCKS. 253 pointed out the important pa.'t played by the protoxyd- bases liberated by the sub-aerial decay of feldspathic and hornblendic rocks. Starting from the conception of a primitive terrestrial crust consistii^g wholly of crystalline silicated rocks, we are forced to find in such a process of decay the source of all limestones and dolomites. These are derived from the carbonates of lime and magnesia generated either directly, during the process, from the bases previously existing in the state of silicates, or indi- rectly, by reactions between magnesian and alkaline car- bonates formed during the decay, and the calcic salts of the early ocean. The chemical genesis of the lime-car- bonate must evidently precede its : .-.similation by organ- isms. It was, in fact, thus shown, as the result of a great number of observations, that fossil sea-waters (mineral waters), representing the ocean of paleozoic and even of meso"oic times, contained large proportions of calcic chloride, such as are required by this theory.* The r^ui- tions of these reactions when " this decay of alkaliferous silicates is sub-aerial," as set forth in 1859 and 1863, will be found discussed at length in the volume above named, on pages 23-31, and page 108. [See, for a certain exten- sion and modification of this view of the source of lime- carbonate, ante, pp. 178, 239.] § 15. I farther proceeded at that time to consider the proportions l)etween the alkalies and the alumina in the various characteristic minerals of crystalline rocks, noting the decrease in the former which is seen when silicates like orthoclase and albite are compared with micas like muscovite, and with silicates like andalusite, cj'anite, i)yr- ophyllite, and staurolite. The conclusion was then reached that "the chemical and mineralogical constitution of dif- ferent systems of rocks must vary with their antiquity," and that "it now remains to find in their compartative study a guide to their respective ages " ; in which connec- tion a comparison was then attempted between the older * Hunt, Chein. and Gecl. Essays, pp. 41, 108, 117-121. i 'in. SiCtf ; found abundantly in the very ancient pre-Cam.irian (Keweenian) conglomerates on Lake Supe- rior, as 1 liav3 olsewhere described.^ Not less striking examples of roi,ii ^d ui'.sses of older gneisses occur in the Huronian se ies ;.; many localities, particularly on Lake * S(,.;jnd Geologic, Snrvey of Penn.; Azoic Rocks, Rep. E., p. 210. t iS'p, for these, Irvuig on the Mineral Resources of AVisconsin, Proc. Ame •, Inst. Mining Engineers, vol. viii., p. 305. For otlier analyses, see Geo. H. Cook, Geol. Survey of N^tv Jersey, Report on Ckys, 1878. } Hunt, Azoic Rocks, pp. 78, 230. )• t t FranJ § us to more ocess mate- eozoio latter as are 1 latav tnasses, r pre- |anci?nt S^ipe- V m the n Lal^e , p. 210. Iisin, rroc. lalyses, see |1878. VII.] THE DECAY OP CRYSTALLINE EOCKS. 255 Temiscaming, where are great beds oi conglomerate made up chiefly of gneiss boulders.* i ha\e elsewhere noticed a specimen in my possession which shows a perfectly well defined and rounded pebble of finely granular white lime- stone, measuring an inch in its greatest diameter, enclosed in a laminated hornblendic gneiss, from Grafton County, New Hampshire. Slices cut from the specimen for the microscope show a strong adhesion between the limestone and the quartz and feldspar of the matrix, without, how- ever, any evidence of chemical change at the contact.f § 17. The rounded masses and pebbles of gneiss found abundantly in several localities imbedded in the pre-Cam- brian micaceous schists of the Saxon Erzgebirge are not less remarkable examples of the same kind. I had in 1881 the opportunity of examining with Dr. Credner at Leipsic a large collection of these, which consist chiefly of type ; of various kinds of gneiss resembling those of the Lau- rentian series as seen in North America and in the Alps. These Saxon mica-schists, with their associated gneisses passing into granulites or leptynites, have all the charac- teristics of the Montalban or newer gneissic series of North America and of the Alps, to which I have elsewhere compared tliem in two communications,^ wherein are noticed the above-mentioned conglomerates, which had been previously studied in much detail by Sauer,§ in 1879. No one who sees these accumulations of rounded masses of gneiss and other crystalline rocks entering into con- glomerates at the various horizons above named, can fail to be struck with their close resemblance to those which are to be found either in the glacial or other modern deposits, or lying in situ as undecayed rounded masses in the midst of decomposed rocks. It is difficult to resist the conclusion that these rounded masses of the eozoic ages * Geology of Canada, 1863, p. 50. t Bull. Soc. Geol. de France, [3], x., 27. t Geol. Magazine, January, 1882, p. 39, and Bull. Soc. G6ol. de France, x., 26. § Zeitschrift f. d. ges. Naturwlss, Band lii. \' nT'T I— ■!! 256 THE DECAY OP CRYSTALLINE ROCKS. tyiL I ' N' j • i must have been formed under conditions not unlike those which gave rise to their more modern representatives. § 18. The various considerations above presented thus lerl the writer, in 1873, to assign to the beginning of the process of rock-decay an antiquity compared with which the tv^^e that has elapsed since the drift-period is to be regarded as of short duration. It was, however, then suggested by him that a climate and atmospheric condi- tions unlike those of modern times might have favored the process in the earlier ages. Further evidence was soon forthcoming both of the former spread of this decay over northern regions, and of its great antiquity. In 1874 I was called to examine the condition of the great tunnel then recently opened through the Hoosac Mountain in western Massachusetts, my report on which was published by the General Court of the State ; * while a note on the observations therein made which have a bearing on the present inquiry, was presented to the American Institute of Mining Engineers in October, 1874.t § 19. As ti.ore explained, the gneissic rock of Hoosac Mountain, at the west end of the tunnel, 700 feet above the sea, is completely decayed, the feldspar being con- verted into kaolin for a distance of se\ -,-al hundred feet eastward, along the line of the tunnel. The gneiss on the crest of the mountain, 2000 feet above the sea, and on the eastern slope, on the contrary, wherever • xposed, pre- sents the rounded surfaces common throughout the region, often marked by glacial striae, and without any appearance of decay. The softening and decomposition of the highly inclined strata of gneiss in the tunnel were described as complete for a distance of 600 feet from the west portal, where the floor of the tunnel is 200 feet from the surface, and were partial at 1000 feet, where it is 230 feet below ; while farther in, at 1200 feet, an included bed of ♦ House Document No. 9, 1875, t Trans. Amer. Inst. Mining Engineers, iii., 187. m:r't. vn.] THE DECAY OF CRYSTALLINE ROCKS. 257 I'monite, doubtless of epigenic origin, showed that the solvent and oxydizing action of atmospheric waters had penetrated to a depth of more than 300 feet from the present surface. At the western entrance to the tunnel the gneiss is immediately succeeded by the crystalline limestone and quartzite of the Taconian (Lower Taconic) series, the decayed rocks apparently coming from beneath the limestone. It was evident that this great mass of decayed gneiss at the western base of Hoosac Mountain is but a portion of a once widely spread mantle of similar materials, which has escaped the action that denuded and striated the surface of the other parts of the mountain. § 20. Numerous examples of similar remaining por- tions of decayed feldspathic rock have been observed far- ther southward, as in northwestern Connecticut (described in § 7), and among the Laurentian rocks of the South Mountain, in Pennsylvania, north of the SchuyEvdl. One of these decayed portions, at Siesholtzville, was seen in 1875, where a bed of magnetite, at that time mined, was found to overlie at a high angle a mass of granitoid gneiss completely kaolinized, but apparently protected from erosion by the incumbent iron-ore. In another example in the same region, about two miles south of Allentown, the Primal or Taconian sandstone was found resting for a little distance on the Laurentian gneiss, here much decayed. Wl ere this had been exposed in a recent cutting (in 1875) the reddish feldspathic rock, still retaining its color and its gneissic structure, though kaolinized, contained numerous "boulders of decomposi- tion," from three to twelve inches in diameter, consisting of undecayed gneiss, the laminated structure of which was clearly continuous with that seen in the enclosing decayed mass. These boulders, still in situ, spheroidal in form, and often with pitted surfaces, are identical with those found in the drift near by, on the southeast slope of the hill, and are very different in outline from the half- angular forms of adjacent sandstone blocks. This gneiss ! 1 i H 1 m ''1 H i HI "MfeM 258 THE DECAY OP CRYSTALLINE ROCKS. [VII. Pri.j li'ii ;•! •',.* ^^^^'"■#!iHiiii iii' rock, lying decayed in place, woidd, unless examined in fresh cuttings, wliicli show its liighly inclined foliation, be readily mistaken for the drift of the vicinity, which has evidently been derived from it. § 21. In my earlier notices of the decayed Montalban rocks of the Blue Ridge in North Carolina, I had de- scribed a mantle of from fifty to one hundred feet or more of decayed material, but this, according to the late William B. Rogers, sometimes exceeds two hundred feet, a thick- ness approaching to that observed at the western base of Ploosac Mountain. I have since noticed the decay of the Montalban rocks near Atlanta, in Georgia, where, with local exceptions of undecayed areas (as in Stone Moun- tain), the decomposition is more or less complete, in many places, to a depth of fifty feet. Here, as elsewhere, the more massive rocks include nuclear masses of undecayed material. The decayed highly hornblendic gneiss of At- lanta, though still retaining considerable coherence, has lost about two thirds of its weight, the specific gravity of unchanged portions being 2.^7-3.08, while that of the decayed material is reduced to 1.20, and even, for some specimens, to less than 1.0.* The decomposed gneiss in this region is, in some cases, sufficiently coherent to fur- nish blocks for certain purposes of construction, such as the walls of rude chimneys, but at the surface it readily disintegrates, yielding a strong red soil, often used as a brick-clay. The decayed mica-schists of the Montalban series, which still retain their micaceous asi)ect, have been called hydro-mica schists, though distinct from those of the Taconian, with which they have been con- founded. § 22. The relations to the general process of decay, of the large deposits of cupriferous iron-pyrites found in the rocks of the Blue Ridge, were discussed by the writer in 1873,* after a study of the copper-mines opened in Carroll County, Virginia, in Ashe County, North Carolina, and in * Azoic Rocks, p. 250. •1' VII.] THE DECAY OF CRYSTALLINE HOCKS. 259 W^ Polk County, Tennessee.* These ore-deposits were de- scribed as in each case in rocks of the Montalban group — the newer gneisses and mica-schists — and as constituting veins or lenticular masses of posterior origin, consisting es- sentially of pyrite, pyrrhotite, and dialer ^.y rite. The agent which kaolinized tiie enclosing rocks also oxydized the sul- phurets, removing the sulphur and the copper, and convert- ing the residue into limonite, which, in a vertical lode in Ashe County, was found to extend to depths of from forty to seventy feet. Beneath the oxydized portion is found in all cases the unchanged pyritous mass, seldom carrying more than four or five hundredths of copper. The limo- nites thus generated were for some years smelted for iron, both in Virginia and in Tennessee, before they were discov- ered to be th'' oxydized outcrops of cupriferous pyrites- lodes. Between the unchanged pyrites and the limonite there is often found, in favorable conditions, an accumula- tion known as black ore, consisting of imperfectly crystal- line sulphurets, rich in copper, and sometimes approaching to bornite in composition, occasionally with red oxyd and native copper ; the whole, doubtless, reduced from the oxydized and dissolved copper brought from above. § 23. The crystalline eozoic rocks of various ages, in the more northern parts of the continent, contain, as is "Well known, many deposits of cupriferous pyritous ores, both in veins and beds which, like the enclosing strata, are undecayed, showing that the process of oxydation, like that of kaolinization, has been a very gradual one, going back to remote ages. We have seen, from the observations in the southern United States, that the oxydation of the sulphids, their conversion into limonite, and the removal tlierefrora of the copper by solution, went on pari jmssu with the decay of the including rocks, and hence preceded their erosion. The copper thus dissolved was, as I have suggested, again deposited in rocks at the time in process * Proc. Amer. Inst. Mining Engineers, ii., 123, and Amer. Jour. Science, vi., 305; see also Chem. and Geol. Essays, pp. 217, 250. 3 IMAGE EVALUATION TEST TARGET (MT-3) // A /A IL U.. 1.0 !!:■- IIM I.I 12.2 Hf li£ 12.0 12 11.25 i 1.4 6" 1118 1^ 'W. ^ n % /: ^/»"^ '» 7 /A PhotDgraphic Sciences Corporation 23 WBT MAIN STREET WEBSTER, N.Y. 14580 (716)872-4503 y/£ A^ u. ^ a m . 260 THE DECAY OF CRYSTALLINE ROCKS. [VII. ^M !'i ' m of formation. The chief part of the iron being left behind, a veritable concentration of the copper would thereby be effected, and we should expect to find it separated on reduction as a rich sulphide, or as native copper. In ac- cordance with this view, it was said, in an essay on The Geognostical Relations of the Metals, in February, 1873,* that certain deposits of such copper-sulphids, found chiefly in limestones, probably of Cambrian age, which, in the province of Quebec, as at Acton and Durham, lie along the northwest border of the- crystalline Huronian belt, might be formed from "the results of oxydation of the cupriferous beds which abound in the crystalline schists of these mountains, from which the dissolved metal accu- mulated in basins at their foot," as suggested by Murchison with regard to the cupriferous Permian strata near the crystalline schists of the Ural Mountains. "To a like process," it was said, " we may perhaps ascribe the rich deposits of native copper in the Keweenaw amygdaloids and conglomerates, which rest upon the ancient Huronian schists." § 24. The farther extension of this view to the meso- zoic sandstones of Connecticut, New Jersey, and Pennsyl- vania, well known to be very often impregnated with copper disseminated in the form of sulphids, sometimes associated with organic remains, is obvious. It is to be noticed that the strata in question are generally deposited directly upon eozoic rocks, from the ruins of which they were formed, and that these, in our hypothesis, furnished the dissolved copper from which the disseminated ores were derived. If this view be admitted, we have farther and independent evidence that the decay of the eozoic rocks, with that of their contained cupriferous sulphurets, was going on in that pre-Cambrian period in which the Keweenian series was accumulated, and was still active in mesozoic time. § 25. Not less striking examples of rock-decay are seen • Proc. Amer. Inat. Mining Engineers, i., 341. LVII. ind, y be [on i ac- The ^73,* liefly a the along belt, 3£ the ichists 1 accu- •cbison ;ar the a VikQ he ricb ;daloids uronian e meso- >ennsyl- ed with metimea lis to be eposited lich they lurnished ,ted orea e farther ,e eozoio ilphurets, rhich the active in are seen vn.] THE DECAY OP CRYSTALLINE ROCKS. 261 in the great Appalachian valley, of which the Hoosac Mountain, the South Mountain, and the Blue Ridge form parts of the eastern rim. Therein, as is well known, large quantities of limonite are mined, from New England to Alabama. This ore, as well as its accompanying man- ganese-oxyd, is clearly of epigenic origin, and is, in most cases, still imbedded in ancient and highly inclined clayey strata derived from the sub-aerial decay in situ of the schists which accompany the dolomites and quartzites of the Primal and Auroral (Taconian) series. These oxy- dized ores have been formed hy the transformation of included masses of pyrites and of carbonates of iron and manganese.* The evidences of the pyritic origin of many of these limonites is similar to that for those of the Blue Ridge (§ 22), namely, their association with unchanged pyr- ites. An example of this is seen in the so-called Copperas mine at Breinigsville, near Trexlertown, Pennsylvania, long ago described by H. D. Rogers,! where large quanti- ties of pyrites have been mined from the same openings which yield limonite. Some of this I found still retain- ing the imitative forms of the adjacent pyrites (from which Rogers had inferred a conversion of limonite into pyrites), while the waters of the mine, like those of others in the region, were charged with sulphuric acid and with iron-sulphate. Another remarkable locality, where pyr- ites replaces the limonite in depth, was visible in 1875 at Seitzinger's mine, near Reading, Pennsylvania, and other similar cases are reported in the vicinity ; while at Salona, in the Nittany valley, the association of pyrites with lim- onite at this same geological horizon has also been noticed. § 26. The association of siderite or iron-carbonate with the limonites of the Appalachian valley is well known east of the Hudson, in New York and Massachusetts. This mineral is often manganesian, and passes into nearly pure rhodocrosite. Examples of the association of siderite • Azoic Rocks, pp. 201-203. t Geology of Pennsylvania, i., 265. i I .. ii.i NI1il!ll{{|ti{llfi 262 THE DECAY OF CRYSTALLINE ROCKS. [VII. with liuionite are also seen, among other localities, near Hackettstown, New Jersey, and near Hanover, York County, Pennsylvania. These carbonates, or at least the limonite and manganese-oxyj derived from them, are found in close association with pyritous deposits, as we have seen near Trexlertown. In like manner, pyrites and siderite, as is well known, often occur side by side in the coal-measures. § 27. I have elsewhere considered the change in sider- ite under the action of oxydizing atmospheric waters, which proceeds like that in feldspathic rocks, from with- out inwards, and is necessarily accompanied with consid- erable diminution of volume, which, in the conversion of a siderite of specific gravity 3.6 into a limonite of the same density, would equal 19.5 per cent. " The evidences of this contraction may be seen in the structure of the limonite derived from siderite, which often forms a porous or spongy mass. In the case, however, of nodules or blocks of solid ore, the conversion beginning at the outside of the mass, an external" layer of compact limonite is formed, and then another within this, and still another, till the change is complete. The void space resulting from contraction is then found between the ■layers, which are arranged like the coats of an onion, or sometimes wholly at the centre, where a cavity will be formed, holding in many cases more or less clay or sand, the impurities of the carbonate, which have been sepai-ated in the process of conversion into limonite. In this way are formed the hollow masses sometimes known as bomb- shell ore, which occasionally include nuclei of unchanged sideiite. Their structure will generally serve to dis- tinguish the sideritic from the pyritic limonites." * In the paper just quoted I have also considered the change of volume which should accompany the conver- sion of pyrites into limonite, a process generally com- * The Genesis of Certain Iron Ores; read before the Amer. Assoc. Adv. Science, 1880: Canadian Naturalist for December, 1880, ix., 434. IVIL J, near York ist the m, are , as we tes and 3 in the n sider- waters, ,m with- 1 consid- ersion of e of the en in the lich often iwever, of rinning at ; compact ,, and still oid space ;ween the onion, or ty will be ^y or sand, separated . this way i\ as bomb- unchanged Ive to dis- iidered the •he conver- ■rally cora- J Atner. Assoc, bo, Ix.. 434. VU.] THE DECAY OF CRYSTALLINE ROCKS. 263 plicated by the loss of a part of the iron as a soluble sulphr.te. § 28. Portions of the contorted and often highly in- clined schistose strata enclosing the limonite ores in the Appalachian valley, are still found but partially decayed, and while some are converted, to depths of 100 feet or more, into white or variously colored clays, others retain more or less of their original texture. From the presence in some of these of considerable quantities of a hydrous micaceous mineral, having the composition of damourite, they have been called damourite-slates. There are many reasons for believing that these ancient rocks were thus folded, and were decomposed, before the deposition of the Trenton and Chazy limestones, which rest upon them in the outlying or western valleys of the Appalachian region, alike in Pennsylvania and in Alabama.* § 29. Professor Lesley, in discussing the history of the limonites of the Appalachian valley, has fallen into an error with regard to my view of their origin. Referring, in 1876, to the opinions expressed in my paper of 1873 (already noticed in § 13) touching the decayed crystalline rocks of the Blue Ridge, that " the iron-oxyd from these has been in great part dissolved out by subsequent pro- cesses, and was the source of the immense deposits of hydrous iron-ores " in question, he supposes me to " con- jecture that the ores lying along the eastern edge of the Shenandoah valley had been washed into it from or across the Blue Ridge." This Lesley properly qualifies as an * The fact of the existence at various points in the Appalachian valley of beds of limonite interstratified in tertiary clays with lignite, as at Brandon, Vermont, must not be overlooked. First recognized by Edward Hitchcock, and subsequently noticed by liCsley, in 1864, the later observa- tions of Prime, Lewis, and others, show the presence of these ores and clays, with lignites, at various points in Pennsylvania and in Alabama, as well as in Vermont. These are but fragments of what were probably once extended deposits, and although of geological interest as resulting from re-solution and re-arrangement, in tertiary time, of a portion of the ancient decayed strata of the valley, are of comparatively little economic Importance. (H. C. Lewis, Proc. Acad. Nat. Sci. Phila., Oct. 27, 1879.) i 1 I J 264 THE DECAY OF CRYSTALLINE ROCKS. [VII. ' 'f ; kvaki III "absurd conclusion," since it does not explain the origin of the limonites found in the back or central valleys, a hundred miles or more to the west of the Blue Ilidge ; and declares that had R. S. M. Jackson continued his geological studies, "he would have published a satisfac- tory refutation of this surface-drainage theory of the brown hematites." * § 30. Those who have read what I had written on the" subject previous to 1876, and especially my discussion of the origin of these ores in 1874,f are aware that I have never advocated any such theory. I have, it is true, endeavored to find in the insoluble products of decay of these ancient crystalline rocks, the source not only of the clays and sands of the succeeding sediments, but of their contained iron, whether diffused, or accumulated in ore- masses. I have, however, at the same time, always main- tained that the ores associated with the so-called Primal and Auroral rocks of the Appalachian basin, like those of the higher horizons, up to the coal-measures inclusive, were deposits contemporaneous with the strata in which the valleys were subsequently excavated; and that, save in some cases where, as mentioned below, it was appar- ently deposited as peroxyd, the iron was accumulated in the form of carbonate, and more rarely of sulphid ; from the alteration of both of which, in sitUy the limonites have been formed. This view, which, as I then showed, was that advocated by Charles Uph^m Shepard, in 1837, for the limonites of western New England, was the same as that put forward, in 1838, by R. S. M. Jackson himself, who maintained, as stated in the language of Professor Lesley, "that the ore belonged to the stratified limestone beds themselves, and had been set free from them by chemical and mechanical decomposition." This history was clear to Dr. Persifor Frazer, who, having remarked that " the theory of alteration in situ of various iron-minerals result- * Second Geol. Survey of Penn., Report A, p. 83. t Trans. Amer. Institute Mining Engineers, iii., pp. 418-421. rocJ ore-j gin\ fissi witi beds tiin( the f 1 1 LI. in ,a je; his Ug- )wn the n of have true, ay of i the their a ore- main- Primal iiose oi ; }lus\ve, which t, save aijpar- ated in from es have ed, ^vas 837, for same as ,elf, wlio Lesley, ' ^e beds liemical as clear at "the [Is result- 18-421. vm THE DECAY OF CRYSTALLINE ROCKS. 265 ing in the formation of many of these limonites, advanced by C. U. Shepard, and ably discussed and adopted by Dr. T. Sterry Hunt, cannot be disregarded in seeking the cause which produced these limonites," adds, "In 1838, and independently of Prof. Shepard's observations, Dr. R. S. M. Jackson reported to Prof. H. D. Rogers sub- stantially the same conclusion, from the study of the limo- nites of Centre and Huntingdon Counties." ■ § 31. This same view was in fact well stated by Lesley himself in 1864, when he said, " The brown-hematite ore- deposits of Mount Alto follow the edge of the slates and sandy limestones," and are "but the residues of these beds after decomposition and dissolution, the honey- combed and altered edges" of the slates and limestones themselves, "after the lime has been washed out of them, and their carbonated and sulphuretted ii'on has been hydrated and peroxydized ; the slates having formed the red and white clays." He farther described at one locality of the region in question " an outcrop of almost unchanged blue carbonate of iron and iime, several feet thick. . . . and evidently in part changing into honeycombed brown hematite ore." f § 82. In 1867, Mr. Benjamin Smith Lyman expressed similar views in his account of the limonites of Smyth County, Virginia, found lying below the limestones of No. 11. (Taconian), where many localities " show the ore unmistakably in regular beds conformable to the other rocks." He at the same time supposed that some of these ore-deposits are, like one noticed in Wythe County, Vir- ginia, due to "the weathering of the upper part of a fissure-vein of iron-pyrites,'' but maintains that the ores, with such exceptions as this, were " deposited in regular beds, of greater or less extent and thickness, at the same time with the other rocks," and from the presence in the limonite of occasional masses of carbonate of iron, • Second Geological Survey of Penn., Report C, p. 143.' ., t Amer. Philos. Soc. Proc, ix., pp. 471-475. 'i 266 THE ^ZCAY OP CRYSTALLINE ROOKS. [VU. in 11 concludes that it vas originally deposited in this con- dition.* § 33. But while it is apparent that the ores in question, now found imbedded in clays resulting from the decompo- sitio:. in situ of ancient schists, were, previous to that decay, enclosed therein as massive siderite or pyrites, we must not overlook the evidences that in certain cases a process of segregation of diffused iron-oxyd has played an important part, alike in ancient and in modern times, in the genesis of limonites. Setting aside, as not relevant to our present inquiry, the formation of bog iron-ores and ochres, which are directly deposited from ferrous solutions by peroxydation and precipitation, we here recall the con- tribution to the theory of the origin of imbedded iron- ores made by the late William B. Rogers. The ferrous carbonate found in the rocks of the coal-measures has, as he has endeavored to show, been generated from dif- fused ferric oxyd by a process of reduction, carbonation, and solution, through waters charged with organic matters from vegetable decay ; the carbonate of iron thus formed remaining in some cases diffused through the sediments, and in others becoming concentrated by accrotion.f § 34. This view is to be supplemented by the consid- eration that carbonated solutions of ferrous oxyd formed as above (and often containing organic acids) may, by reacting with beds of carbonate of lime, effect a gradual replacement of the latter by carbonate of iron.J The * Proc. Amer. Assoc. Adv. Science, 1867, p. 114. t Geological Survey of Penn., 1858, ii., 757. X J. Ville found one litre of carbonated water at the ordinary pressure to hold in solution at 2(f C. 1.142 grammes of ferrous carbonate, and at 116° C,1.390 grammes. From these solutions neutral alkaline carbonates readily throw down the ferrous carbonate, themselves passing to the state of bicarbonates; and carbonates of lime and magnesia produce the same effect, though more slowly. (Comptes Rendus del' Acad, des Sciences, October. 1881, vol. xciii.. p. 443. ) The present writer found recently pre- cipitated ferrous carbonate to be temporarily much more soluble, under the above conditions, yielding supersaturated solution^, which in close vessels spontaneously deposit, after many hours, a large part of the car- bonate in a crystalline condition. II. on- VII.] THE DECAY OP CKYSTALLINE ROCKS. 2G7 ion, Lipo- that J, we iea a 3(1 an >e9, in evant 69 and Lutions tie con- id iron- ferrous res l^as, •rem di£- )onation, p matters ^s formed cUnients, "•^ -A e coiisia- ;d formed may, ^Y a gradual n4 ^^' lonate, and at Ine carbonates li2 to the state [duce the same I des Sciences, la recently pte- LoluWe, under Lh\cb in close Vrt of tbe car- transformation of diffused ferric oxyd in sediments into massive limonite, imbedded therein, is thua a twofold process, involving, first, the intervention of reducing solu- tions converting the peroxyd into ferrous carbonate, and the concentration of the latter ; and second, the change of this latter, through peroxydation and hydration, into limonite. [Dr. N. S. Shaler has shown the important bearing of the reaction just pointed out (by which beds of carbonate of lime are gradually changed, through replacement, into carbonate of iron) upon the production of beds of fer- riferous limestone, and of iron-ores, at various geological horizons. This he especially notes in the many limestone layers found in Kentucky and Ohio, in the great mass of sandstones and shales of the carboniferous series, where he points out that the fact that the iron is confined to the upper part of the limestone layers shows their transfor- mation by the action of ferrous solutions from above. He farther adduces the iron-ore beds at different horizons between the base of the Devonian shales and the great sandstone (Oneida-Medina), which in the Appalachian basin forms the basal member of the Silurian. These include, besides iron-carbonate superficially changed into limonite, the widely spread deposit of so-called fossil ore, or Clinton ore, evidently a changed marine limestone, in which the iron is now in the form of scaly red hematite. The genesis of this anhydrous peroxyd is not yet clearly explained, but it is to be remarked that concretions of similar hematite are found, instead of siderite, in certain shales in the coal-measures in Ohio. Shaler is careful to distinguish between ore-beds from replaced limestone and the concretionary carbonate ores which are found in shales, where there is no evidence of the previous accu- mulation of calcareous masses.]* § 35. It is evident that the first stage of the process indicated by Rogers as takmg place in sediments a^ yet * Geol. Survey of Kentucl^y, 1877, iii., 163-167. I IP 268 THE DECAY OP CRYSTALLINE ROCKS. [VII. unconsolidated, may also be set up in the disintegrated ferriferous moterials resulting from the sub-aerial decay of rooks, and still undisturbed ; that is to say, that the infil- tration of waters holding dissolved organic matter may give rise in the decomposed mass to concretions of ferrous carbonate, which are subsequently changed into limonite. In this way, a concentration may be effected, through which rocks originally containing a small portion of dif- fused iron-oxyd come to include masses of limonite. Illustrations of this process are sometimes seen in the decay of ferriferous limestones or dolomites, in the resid- uum of which we find the iron accumulated in the shape of crusts or layers of limonite. An instructive example of an analogous process is seen in the limonite which on Staten Island, New York, is found imbedded in a layer of brownish earthy material, sometimes attaining a thickness of twelve feet. This rests immediately upon the sevpentine-rock of the region, into which it graduates, and from the sub-aerial decay of which it has evidently been derived ; the lower portion of the earthy matrix still preserving the peculiar jointed structure of the underlying serpentine. This decomposed material, though including botryoidal crusts, geodes, and concretion- ary grains of limonite, with occasional druses of chalce- dony and of quartz crystals, retains considerable coherence. The source of this limonite seems to have been the iron-oxyd liberated by the decay of the ferriferous ser- pentine, and the proportion of ore in the superjacent mass shows a direct relation to the color and apparent propor- tion of iron-silicate in the serpentine beneath. This limo- nite, which is now mined to a considerable extent, contains, as several analyses have shown, from one to two hun- dredths of chromic oxyd, which is also known to be present in small amount in the serpentine. An impure argillaceous specimen, containing only 59.63 of ferric oxyd, yielded the writer 2.81 of chromic oxyd in a con- dition readily soluble in chlorhydrio acid. I. VII.] THE DECAY OF CRYST^VLLINE BOCKS. 2G9 id - of ai- . ay ite. \g^ dif- iite. the Bsid- hape seen rk, is :eria^» 5 rests 1, into wbicli of the ucture ,tenal, letion- Icbalce- irence. sen the lu3 ser- it mass propor- lis linio- mtains, o hun- to he impnre ,f ferric a con- § 86. Dr. N. L. Britton, of the School of Mines of Columbia College, New York, in whose company I lately had an opportunity of visiting this interesting locality, published, in 1880, a geological map, with sections, and a description of Staten Island.* He tlierein shows that the earthy material in which the limonite is imbedded is con- fined to the tops of certain hills of serpentine, being absent alike from other similar hills adjacent, and from intervening valleys cut into the serpentine, and he has connected this distribution of the ore-bearing stratum with the facts of the local glaciation of the region, to which he has devoted much attention. It is, I think, evi- dent that the decay of the serpentine, and the concentra- tion, in the residuu n, of its iron in the form of limonite, was a process anterior to the glacial erosion, and that the ore-banks are areas of the decayed material which escaped this action. § 37. Turning now to the valley of the Mississippi, we find that Pumpelly, in his geological survey of Missouri, showed, in 1873, that the decay in situ of granitic rocks, and of quartziferous porphyry, has left great rounded blocks of these crystalline rocks ; while the conversion of the porphyry into clay, and its subsequent removal, have liberated included veins or masses of crystalline hematite, giving rise to an accumulation of detrital iron-ore, such as, at the well known Iron Mountain, forms a covering over the surface of the hill of porphyry. From the pres- ence of stratified deposits of this detrital ore in the ancient Cambrian strata around the base of the hill, Pumpelly inferred that the decay of the porphyry was already complete to a considerable depth, at this early period.f His observations and deductions were not known to me when, in the same year, I published my conclusions as to the great antiquity and the universality of the pro- cess of rock-decay. • Annals New York Acad. Sciences, vol. ii., part 6. t Geology of Missouri; Report on Iron Ores and Coal Fields, pp. 8-12. 270 THE DECAY OP CRYSTALLINE ROCKS. (VH. K I § 88. Proceeding from Missouri northward, we find that in Minnesota, as shown by C. A. White,* in 1870, and by N. H. Winchell, in 1874, the aneien^ granitoid rocks, when protected by cretaceous strata, support a kaolinized layer of considerable thickness.f In Wiscon- sin a similar condition of things is found beneath the Potsdam sandstone in the central part of the State, as described, by Irving, in 1870, J in an essay which is a valuable contribution to the literature of kaolin, and con- tains many analyses of the decayed rocks of the region, by Mr. E. T. Sweet, which have been already referred to (§ 16). Further details of the same region and its kaolins, with analyses, as before, were given by Irving in n^. essay, in 1880, on the Mineral Resources of Wisconsin. § In Jackson and Wood Counties, where the crystalline (Lau- rentian) rocks are covered by a thin sheet of Potsdam sandstone, the river-valleys, cutting through this, expose the kaolin, which "occupies its original position, retaining sometimes the structure of the unaltered rock." This ia derived from the decay in situ of certain bands, which, passing downward, graduate into unaltered feldspathic rock. Save where this mantle of decayed material has been protected by the paleozoic sandstone, the crystalline rocks are there seen for the most part in an undecayed condition, evidently, as Irving remarks, from the removal of the decayed material by " the denuding action of the drift." In some portions of the driftless area of this region the unprotected gneisses still retain their mantle of kaolin- ized material. § 39. From the facts before us, it is clear that the decay of the eozoic crystalline rocks was already far advanced in pre-Cambrian times. I am informed that similar evi- dence is afforded in Sweden by the presence of decom- * Geology of Iowa, i., 124. t Second Annual Rep. Geol. Minnesota, pp. 162, 166, 207; also Hunt, Chem. and Geol. Essays, p. 250. I Trans. Wisconsin Academy, etc., ill., 13. § Trans. Amer. Inst. Mining Engineers, viiL, 103. the ii yn.] THE DECAY OP CRYSTALLINE llOCKH. 271 Ino has line sayed xoval the region :aoUn- decay anced ar evi- ieuom- Huntf posed rock beneatli Cumbrian strata. Prr f. A. Geikie has moreover shown that the sculpturing of the gneiss rocks of western Scothmd, a jjrocess which I have maintained to be dependent on previous sub-aerial decay, was effected before the deposition of the Cambrian sandstones, which there rest upon ancient roches moutonnees* § 40. It might be supposed, from their stability under ordinary atmospheric influences in regions protected by vegetation, that all such portions of decayed eozoic rocks as still exist in driftless or in protected areas date from the (lawn of paleozoic time, did we not know that the same processes of decay have been active in subsequent ages, as is shown by the decay of eruptive rocks of later periods. An example of this, which shows at the same time the little progress made in the process of decay since the drift-period, is seen in Canada, at Montreal, where, to the south of Mount Royal, the nearly horizontal beds of the impure Trenton limestone are found, in sheltered places, deeply decayed, and porous from the removal of their carbonate of lime, and are moreover traversed by dikes of dolerite and other feldspathic rocks, themselves decayed to considerable depths ; while near by, and espe- cially to the north of the mountain, where glaciation did its work of removing alike decayed aqueous and igneous rocks, the eroded surftices of both of these are found to be hard and comparatively unchanged, beneath a thin layer of soil and vegetation, as described by J. W. Dawson. Another instance is afforded by a dike intersecting the Potsdam sandstone in this vicinity, which is found to be converted, to a depth of twenty feet or more, into a plastic, highly aluminous clay, which, from the presence of por- tions of titanium and chromium, is, we may conjecture, derived from a doleritic rock.f § 41. Rigaud Mountain, an igneous mass rising through the Potsdam sandstone, and occupying several square • Nature, August 26, 1880, p. 403. ' , t Report Geol. Survey of Canada, 1878-79, H., p. 7. ,1 272 THE DECAY OP CRYSTALLINE ROCKS. [vn. miles on the south side of the Ottawa, near its confluence with the St. Lawrence, is probably of paleozoic age, and consists in large part of a reddish granitoid orthoclase rock. Considerable areas of its surface, lying lower than the surrounding crests of the mountain, are covered to a depth of seven feet or more, in places, v/ith well rounded boulders from three to eighteen inches in diameter, con- sisting wholly of the rock of the mountain, with the exception of a few masses of sandstone. The areas so covered attain, in their higher parts, an elevation of about 280 feet above the Ottawa, but slope gently both to the south and the north. The boulders are very rare on the north slope of the mountain and at its northern base, but are abundant on the southern slope and in the low-lying clay-covered plains to the southward.* These well rounded masses, spread over so much of the mountain, are apparently boulders of decomposition, still in situ^ having escaped the denuding agents of the drift-period. § 42. Examples of more recent sub-aerial decay of crystalline rocks, under peculiarly favorable conditions, were in 1880 described independently by Jos. LeConte and myself, in the auriferous pliocene gravel of California. The pebbles of feldspathic and hornblendic rocks occur- ring in the portions below drainage-level — the so-called blue gravel — are unaltered, while above that level the similar pebbles, exposed to the action of meteoric waters, are more or less completely kaolinized, exfoliating, becom- ing earthy in texture, rusty in color, and in some cases converted into a clayey mass. The pjrites, so abundant in the blue gravel, has, in these uppc* portions, or so- called red gravely been oxydized, and the accompanying lignites have been silicified, and often incrusted with crystallized quartz, from silica liberated in the process of rock-decay through the infiltration of surface-waters.f § 43. To the porosity of the gravel, and the great * Geology of Canada, p. 896. t LeConte, Amer. Jour. Science, xix., 177; Hunt, ibid., xix., 371. VII.] THE DECAY OP CRYSTALLINE ROCKS. 273 amount of surface thus exposed, is to be added the influ- ence of carbonic acid from the decaying lignite, the carbon of which is oxydized as the process of silicifi cation goes on. The amount of carbonic dioxyd in the air of certain drift-mines in these auriferous gravels is so great that candles will not burn therein. Mr. D. T. Hughes of San Francisco, a well known mining engineer, to whose careful scientific observations I have been much indebted, informs me that in the case of a drift-mine 300 feet below the surface, in Table Mountain, Tuolumne County, Cali- fornia, where the foulness of the air was especially remarked, he satisfied himself, by appropriate tests, of the presence in the air of a large proportion of carbonic dioxyd. If, as there is reason to suppose, the amount of this element in our atmosphere was somewhat greater in former ages than at present, we have in these gravels an illustration of its influence in promoting the docay of sili- cated rocks. It is not improbable that the sulphuric acid generated by the oxydation of the pyrites present in these gravels may also have aided in the process. § 44. The slight evidences of decomposition to be seen in the crystalline rocks of thoroughly glaciated regions, as well as in transported boulders, make it probable that the seemingly rapid progress of decay, occasionally observed on exposure, of similar rocks in other regions, sometimes appealed to as evidence of a decomposition now going on, is really but the mechanical disintegration of masses already partially kaolinized in former ages. The. crumbling of certain apparently unaltered granitoid rocks, in which the feldspar remains bright and hard, should be distinguished from that which follows chemical decomposition. Such disintegration, due apparently to changes of temperature * and the action of frost, is, how- * In this connection I venture to recall the attention ci geologists to a phenomenon already described both by Dr. Shaler and myself, apparently due to superficial alternations of temperature on certain crystalline rocks, which have resulted 'n establishing in them, to a considerable depth, a W ''il iff I-' 1 1 ! I 274 THE DECAY OF CRYSTALLINE ROCKS. [VII. "■' ffiilli ever, important, and deserves farther study from the fact that materials apparently of similar origin enter into the composition of many derived rocks. Lava-flows are, it has been observed, subject to comparatively rapid sub- aerial decay, but these rock-surfaces differ widely in tex- ture, as well as in composition, from most crystalline rocks. § 45. An essay by Professor Pumpelly on Secular Rock-Disintegration, read before the National Academy of Sciences, in April, 1878,* is a very valuable contribu- tion to the subject before us. He cites therein my conclu- sions as to the great antiquity and the universality of the process of rock-decay (to which his own observations in Missouri have contributed importf'.nt data), and also as to the final removal of decomposed material from north- eastern America in the time of the glacial drift. He fur- ther notes the little attention hitherto given to the subject of sub-aerial decay, and points out its importance in con- nection with great problems in dynamical geology. The view that this process of rock-decay is "a necessary pre- liminary to glacial and erosi^'e acdon, which removed already softened materials," receives from Professor Pum- pelly an extended discussion and application. He pro- cc-tds to consider the removal and the re-arrangement of these softened materials by three different agencies. First, the encroachment of the sea upon a subsiding region of decayed rocks ; second the aci/ion of land-glaciers, in which he points out that the groat mass of disintegrated and water-impregnated rock wculd become frozen, and series of lifts or divisional planes parallel to the present surface, which are well known to quarrymen. Instances of this abound ; besides those noticed by me in the Araer. Jour. Science for July, 1870 (vol. i., p. 80), may be mentioned the gneiss on the opposite slopes of Rollestou Hill, Fitclibuig, MassacLviSetcs, and that of Stone Mountain, near Atlanta, Georgia; also a remarkable example of comparatively thin horizontal plates at the outcrop of beds cf nearly veriical micacejus gneiss in the vicinily of Worcester, Massachusetts. ' . * Amer. Jour. Science, xvii., 133-144. I' vn.j THE DECAY OF CRYSTALLINE ROCKS. 276 t )- IC- \e Lar my bu- jlu- the isin as orth- } f ur- ibject 1 con- Tbe y p^^: noved Pum- e pro- ent oi encies. lievs, ii^ jegra in, a ted nd Lee, -which [ides those i., ?• 8«V ir Atlanta, ]borizonta\ leiss in the included, as it were, in the glacier, sharing in its move- ments and forming thus a ground-moraine.* § 46. To these modes, with which are to be included the ordinary action of rivers and floods, all acting on the peripheral areas of continents, he adds, for the central areas, removed from these agencies, and rendered desert by geographical conditions, the action of the winds. By these the decayed rock, according to Pumpelly's extension of the ingenious hypothesis of Richthofen, will be sepa- rated into the fine material of the loess, on the one hand, and the sand and gravel of the desert steppes on the other. He thus explains the condition of the crystalline rocks in northern Asia, from which the decayed mantle has been removed not by glacial 'but by aerial agencies ; while the similar rocks in southern Asia, as in Brazil and the southern United States, are still deeply covered with the products of their own decomposition. § 47. Pumpeiiy farther remarks that the surface of the undecayed rock to be laid bare by erosion is ijecessarily an irregular one, the inequalities depending not upon its original difference! in hardness, but upon its resistance to decay under the influence of atmospheric waters. The effect of fractures, joints, veins, and dikes in the rock, in favoring or retarding the action of this agent, would be manifested by still further irregularities of the plane limit- ing the decomposition of the rock in depth. Thus the * Other agencies than ice may produce a similar displacement of decayed material. Belt, in 1874, in his Naturalist in Nicaragua (page 94), describes a movement of the mantle of decayed crystalline rock on hill- sides, in that country, as due to land-slides in wet weather. The layer of ground and " reworked " decayed material resting on the gneiss, found by Hartt in Brazil (Scientific Results of a Journey, etc., pp. 28, 573), and referred by him to glaciation, may perhaps be a similar phenomenon. More recently, Kerr, in 1879, has described the results of a slow down- ward motion of the decomposed surface on mouni,ain-sides in North Carolina as due to the alternate freezing and thawing of the contained water. To tliis displaced and modified layer, which resembles that pro- duced by glacial action, he gives the name of frost-drift (Proc. Amer. Inst. Mining Engineers, viii., 402.) 1^ 276 THE DECAY OF CRYSTALLINE ROCKS. [VII. rounded surfaces, and the closed rock-basins, so often observed in glaciated regions of crystalline rocks, are seen to be but the natural results of the process of rock-decay, which preceded and prepared the way for denudation. § 48. Similar views as to glacial erosion have since been advocated by Nathorst, and more lately by Reusch, wlio, in a memoir on the geology of Corsica, presented to the Geological Society of France, in November, 1882,* has described the disintegration of the granitic region of Corsica to a depth of several metres, giving to the surface smooth slopes, instead of the bold escarpments seen in like rocks in Scandinavia. He notes in these disintegrated rocks in Corsica, enclosed balls or ellipsoidal masses, fresh in appearance, but like in composition and in structure to the enclosing rock, which, when detached, have been taken for erratic blocks. With this region he contrasts the similar rocks near Christiania, in Norway, with hard, rounded surfaces, marked by glacial scratches, where it is difficult to find any trace of superficial decay. He does not believe that the ice of the glacial period removed any considerable portion of the hard rock, to form fiords, val- leys, etc., but supposes " a profound disintegration of the Scandinavian rocks before the glacial period," and con- ceives the present relief to " represent the surface of the unaltered syenite after the removal by the glaciers of all the decomposed part." The salient rock-masses of the Norwegian coast are, like the fiords, arranged in a north and south direction, and this, according to Reusch, corresponds with that of fissures, more or less nearly vertical, which traverse the rock, and, as he well remarks, prepared the way for its disintegration in depth, the extent of which would depend upon differences in the nature of the rock. Many of the lake-basins of this region were, in his opinion, formed through the removal by glaciers of the decayed material * Bull. Soc. Geol. de France, xi., 62-67. i&. - VII.] THE DECAY OF CRYSTALLINE ROCKS. 277 from depressions, while others are due to the action of moraines, serving as dikes. § 49. It is difficult to state more clearly the conse- quences which follow from the conception that the decom- position of rocks is "a necessary preliminary to glacial action and erosion, which removed previously softened materials." Reusch, however, seems, from some miscon- ception, to regard this pre-glacial disintegration of the rocks as distinct from kaolinization, of which the crum- bling of the granites in Corsica, as in other regions, doubt- less represents an incipient stage, such as we meet with in regions where the superficial, and more completely decayed portions, have been removed (§ 44). § 50. The points insisted upon in this essay may be thus briefly resumed : — 1. The evidence afforded by recent geological studies in America, and elsewhere, of the universality and anti- quity of the sub-aerial decay both of silicated crystalline rocks and of calcareous rocks, and of its great extent in pre-Cambrian times. 2. The fact that the materials resulting from this decay are preserved in situ in regions where they have been protected from denudation by overlying strata, alike of Cambrian and of more recent periods; or, in the absence of these coverings, by the position of the decayed materials with reference to denuding agents, as in driftless regions, or in places sheltered from erosion, as in the Appalachian and St. Lawrence valleys. 3. That this process of decay, though continuous through later geological ages, has, under ordinary condi- tions, been insignificant in amount since the glacial period, for the reason that the timr which has since elapsed is small when compared with previous periods, and also, probably, on account of changed atmospheric conditions in the later time. 4. That this piocess of decay has furnished the mate- rials not only for the clays, sands, and iron-oxyds from the Ik 278 THE DECAY OF CRYSTALLINE ROCKS. [VII. beginning of paleozoic time to the present, but also for the corresponding rocks of eozoic time, which have been formed from the older feldspathic rocks by the partial loss of protoxyd-bases. The bases thus separated from crys- talline silicated rocks have been the source, directly and indirectly, of most limestones and carbonated rocks, and have, moreover, caused profound secular changes in the constitution of the ocean's waters. The decay of sul- phuretted ores in the eozoic rocks has given rise to oxy- dized iron-ores, and also to deposits of rich copper-ores, in various geological horizons. 5. That the rounded masses of crystalline rock left in the process of decay constitute not only the boulders of the drift, but, judging from analogy, the similar masses in conglomerates of various ages, going back to eozoic time ; and that not only the forms of these detached masses, but the outlines of eroded regions of crystalline rocks, were determined by the preceding process of sub-aerial decay of these rocks. m VIII. A NATURAL SYSTEM IN MINERALOGY. This essay was presented in abstract to the National Academy of Sciences, in Washington, April 23. 1885, and subsequently to the Royal Society of Canada, in Ottawa, May 27, 1886. It is published in full in the Transactions of the latter society, vol. III., sec. iii.; in abstract in the American Naturalist for July, 1885; and also, with some farther additions, in the Canadian Record of Science, i., 129-136 and 244-247. I. — HISTORICAL INTRODUCTION. § 1. The examination of the various species of the inorganic kingdom which constitute the crust of the earth has long occupied the attention of students of natural history, and has given rise to descriptive and systematic mineralogy. Botanists and zoologists, by making known the structure, growth, and development of organic species, have meanwhile performs I a similar task for the vegetable and animal kingdoms, and have, moreover, arranged organic species in genera, families, orders, and classes, in such manner as to show more or less perfectly their origin and affinities, so that to-day the received classifications of plants and animals merit the name of natural systems. § 2. Without adverting to the work of earlier students, it should be said that Werner, about a century since, pro- posed for the mineral kingdom a classification which makes an epoch in the history of mineralogy. His system was based on " the natural alliances and differences which exist between minerals," and of him it is said that he " established and arranged the greater number of species in the mineral kingdom solely by agreements and differ- ences in external characters ; " grouping the various mine- rals in classes, families, genera, species, sub-species, and kinds. While chemical considerations were not over- 280 A NATUIIAL SYSTEM IN MINKBAliOGY. tVIIL looked in the larger divisions, Werner, according to Jame- son, regarded the intervention of chemistry as but a provisional expedient, and doubted tlie possibility of con- structing a philosophical system in which the external and the chemical characters sliould be conjoined. § 3. Werner died in 1817, and was succeeded at Frei- berg by Frederick Mohs, who sought to complete tlie work of his great predecessor in mineralogy. His early publica- tions on mineral classification go back to 1805, but it was not till 1822-24 that he gave to the world his " Grundriss der Mineralogie," in two volumes. This was translated into English, with additions, by one afterwards famous in science, William Haidinger, who declared, in the preface to that translation, published in Edinburgh in 1825, that he had been a student in mineralogy with Mohs since 1812. Previous to 1820, however, Mohs had visited Edin- burgh, and had there aided Jameson, then preparing the third edition of his " System of Mineralogy," which ap- peared in three volumes in Edinburgh in 1820. In his preface to this edition, Jameson gratefully acknowledges his aid, and says that the arrangement adopted " is nearly that of my celebrated friend, Mohs, who now fills the mineralogical chair of the illustrious Werner." He adds, " The mineral system, as it appears in this work, is to be considered as realizing those views which Werner enter- tained in regard to the mode of arranging and determining minerals." This system, which was designated by Jame- son the Natural History Method, is, according to him, "founded on what are popularly called external charac- ters, and is totally independent of any aid from chemis- try." It was, moreover, in his opinion, the only method " by which minerals would be scientifically arranged and rightly determined." * * For a further notice of Werner's views of mineral classification, the reader is referred to the preface to Jameson's worlt, already cited, and also to Cleveland's Treatise on Mineralogy and Geology, in 1822, where, in vol. i., pp. 77-83, will be found an excellent analysis of Werner's min- eralogical system, as put forth by him at Freiberg in 1816. ■"i ;! VIII.] A NATURAL SYSTEM IN MINEUALOGY. 281 § 4. The system of Mohs at once found favor witli naturalists, and was adopted by many (notably by his successor, Breithaupt), not, however, without certain modifications as to the divisions, some of wliicli may here be noticed, in order to give a general idea of the plan of classification.* In the order Spar, as defined by Mohs, were included not only all zeolites, scapolites, and feld- spars, with sodalite, nephelite, and leucite, but petalite, spodumene, and cyanite, as well as pyroxene, amphibole, wollastonite, and epidote, the latter four being made si)e- cies of one genus, Augite-Spar (§ 48). Again, in the order Gem of Mohs we find garnet, idocrase, and stauro- lite grouped together as species of the genus Garnet ; chrysolite, axinite, emerald, tourmaline, topaz, andalusite, and zircon, types of as many genera; together with the genus Quartz^ including the species iolite, quartz, and opal. Corundum, chrysoberyl, and spinel are also united in one genus, and boraoite and diamond constitute other genera under this order. § 5. In adopting the system of Mohs, Charles Uphara Shepard sub-divided the order Spar, and established a new order. Zeolite, in which were included with the zeolites, sodalite, nepheline, and leucite, the other genera in the order Spar of Mohs being left as before. J. D. Dana, on the contrary, enlarged this order, renamed by him Chalcinea, by adding t-^ it a large part of the order Mica of Mohs, including all the true micas then known. He, on the other hand, removed epidote from the alliance with pyroxene, made by Mohs, and placed it in its proper position, with garnet and idocrase, in the order Gem, called by Dana Hyalinea. This, for the rest, embraced all the species which had been therein included by Mohs, whom Dana followed by placing cyanite and fibrolite with the Spars, while andalusite was arranged with the Gems. § 6. Bearing in mind the changes just noted, we have * See, for the system as modified by Weisbach and Breithaupt, § 111, note. 282 A NATUKAL SYSTEM IN MINERALOGY. IVIII. r„ ; to record that in 1835 the classification and tlie nonien- clatiire of Mohs, as tranahited into English by Haidinger, were adopted by Shepard in the first edition of his " Treat- ise on Mineralogy." In the second and third editions of this work, however, in 1844 and 1852, Shepard, while retaining with slight modifications the classes and orders of Mulis, abandoned the characteristic specific names of the latter for the trivial names generally accepted. The natural-history system of Mohs was also adopted, in the first and second editions of his " System of Mineralogy," by J. D. Dana, in 1837 and 1844. He, however, devised a Latin terminology for tlie orders, as well as a binomial Latin nomenclature for the genera and species. § 7. In abandoning the natural-history system in his third edition, in 1850, Dana returned to the trivial nomen- clature. Referring to these changes, its author declared in the preface to a fourth edition of his System, in 1854, his opinion that " the system of Mohs, valuable in its day, had s '^served its end, and that, in throwing off its shack- les for the more consistent principles flowing from recent views on chemistry, the many difficulties in the way of perfecting a new classification led the author to an arrangement which should serve the convenience of the student, without pretending to strict science." A so-called " purely chemical mineral system " had been proposed by Berzelius as early as 1815,* and had mean- while found favor with chemists. Towards this, the difficulties of the natural-history method in mineralogy directed Dana, who, in the preface to his second edition, in 1844, gave, " besides the natural classification, another, placing the minerals under the principal element in their composition," adding that " various improvements on the usual chemical methods have been introduced, which may render it acceptable to those who prefer that mode of arrangement." The chemical scheme then given by him was, as he informs us, taken almost entirely from Ramuiels- * Berzelius, Nouveau Syst^me de Mineralogie, Paris, 1819. VIII.] A yATURAL SYSTEM IN MINERALOGY. 283 berg's treatise on Chemical Mineralogy, then recently published. In 1850, in the latter part of his third edition, Dana put forth a new chemical classification, " in which the Berzelian method was coupled with crystallography " ; while in his fourth edition, in 1854, he maintained that *' the classification of minerals must *^'^w directly from the principles of chemistry," and accepted what he now called the Berzelian system, which, as his readers are aware, is retained in his fifth and last edition, that of 1868. It is also followed in the " Text-Book of Mineralogy " of his son, E. S. Dana, in 1883. § 8. The views of Berzelius, as adopted and modified by Rammelsberg, Naumann, Dana, and others, now prevail among students of mineralogy, with whom the results of the chemical analysis of speci s are generally considered as of paramount significance; while hardness, specific gravity, crystalline form, and optical characters assume a secondary value in classification, and are regarded as im- portant chiefly in connection with determinative mineral- ogy. The conception of a true natural method, which, although but partially understood, was at the basis of the system of Mobs, has been lost sight of; the order which the naturalist finds in the organic is no longer apparent in the inorganic world, as presented iii modern mineralogical text-books ; and this state of things has contributed not a little to the comparative neglect into which systematic mineralogy has of late years fallen. As to the complete divorce between physical and chemi- cal characters in the study of mineral species, maintained by Werner, Mohs, and his followers, there seems to have underlaid it the notion of framing a system which, as in botany and zoology, shall be available for the purposes of determination without the destruction of the individual. It is to be noted, however, that characters dependent upon chemical differences, such as the presence or absence of certain acids, alkaloids, and groups uf essential oils, are not without significance in determining the natural affini- t ' i A NATURAL SYSTEM IN MINEUALGOY. [VIII. iCi- m ■ "4\ ties of plants, and, moreover, that as wo descend the scale of being, from the highly orgiinized forms of the animal anr, in other words, vary inversely as their empirical equivalent volumes ; so that we here find a direct relation between chemical and physical properties. . . . § 13. "Chemical change implies disorganization, and all ■i-m I ' ill i VIII.] A NATURAL SYSTEM IN MINERALOGY. 287 so-called cheniical species are inorganic that is to sa}', un- organized, and hence really belong to the mineral kingdom. In i/his extended sense, mineralogy takes in not only the few metals, oxyds, sulphids, silicates, and other salts which aie found in nature, but also all those which are the pro- ducts of the chemist's skill. It embraces not only the few native resins and hydrocarbons, but all the bodies of the carbon series made known by the researches of modern chemistry. The primary object of a natural classification, it must be remembered, is not, like that of an artificial system, to serve the purpose of determining species, or the convenience of the student ; but so to arrange bodies in orders, genera, and species, as to satisfy most thoroughly natural affinities. Such a classification in mineralogy will be based upon a consideration of all the physical and chemical relations of bodies, and will enable us to see that the various properties of a species iire not so many arbi- trary signs, but the necessary results of its constitution. It will give for the mineral kingdom what the labors of great naturalists have already nearly attained for the vege- table and animal kingdoms. § 14. "In approaching this great problem of classifi- cation, we have to examine, first, the physical conditions and relations of each species, considered with relation to gravity, cohesion, light, heat, electricity, and magnetism ; secondly, the chemical history of the species, in which are to be considered its nature, as elemental or compound, its chemical relations to other species, and these relations as modified by physical conditions and forces. The quanti- tative relation of one mineral (chemical) species to an- other is its equivalent weight, and the chemical species, until it attains to individuality in the crystal, is essentially quantitative. It is from all the above data, which would include the whole physical and chemical history of inor- ganic bodies, that a natural system of mineralogical classi- fication is to be built up. . . . The variable relations to space of the empirical equivalents of non-gaseous species. tiUH 288 A NATURAL SYSTEM IN MINEUALOUY. [VIII. i ■ r i ' or, in other words, the varying equivalent volumes (ob- tained by dividing their empirical equivalent weights by the specific gravity), show that there exist in different species very unlike degrees of coiKhmsation. At the same time, we are led to the conclusion that the molecular constitution of gems, spars, and ores, is such that tliose bodies mast be represented by formuiit. not less complex, and with equivalent weights far more elevated than those usually assigned to the polycyanids, the alkaloids, and the proximate principles of plants. To similar conclusions conduce also the researches on the specific heat of com- pounds." In the paper, published in 18G7, from which the above extracts are taken, it was farther said that the views there set forth as " the basis of a true mineralogical classification " were not new, but had been brought for- ward and maintained by the author in various publications from 1853. § 15. The starting-point in this inquiry was the study of the chemistry of carbon. It was in 1852 that I wrote, " We may define organic chemistry as the chemistry of the compounds of carbon," * ?. statement which, though a common-place to-day, was then perhaps made for the first time. I then insisted upon what I called "the carbon series " and " the silicon series," the latter including all the known silicon compounds. This was followed in 1853 by an essay on " The Theory of Chemical Changes and Equivalent Volumes," t wherein the question of equiva- lent or so-called atomic volumes was discussed with rela- tion to the, investigations of Playfair and Joule, and the speculations of Dana. It was then and there suggested that " all species crystallizing in the same shape have the same equivalent volume, so that their equivalent weights * Essay on Organic Cbemistry, forming part iv. of the Principles of Chemistry by B, Silliman; 3rd revised edition, 1852, p. 378. t Amer. Jour. Science, March, 1853 (xv., 226-2.34); L., E. & D. "Philos. Mag. (4), v., 520, and in a German translation in the Cheraisches Central- blatt of Leipsic for the same year (p. 849); also in the author's Chem. and Geol. Essays, pp. 427-437. l«. mi. VIII.] A NATURAL SYSTEM IN MINEKALOGY. 289 ^ob- iby rent the those iplex, those id the Lisions • com- which lat the ilogical ;ht for- Lcations (as in the case of vapors) are directly as their densities, and the equivalents of mineral species are as much more elevated than those of the carbon series as the specific gravities are liigher." § 16. Another principle there set forth was the general application of the law of progressive or homologous series, first enunciated in 1842 by James Schiel of St. Louis, and soon afterwards adopted by Ch. Gerha.dt, but hitherto applied only to hydrocarbonaceous or so-called organic species. It was now said that " it may be expected that mineral species will exhibit the same relations as those of the carbon series, and the principle of homology be gn^atly extended in its application. The history of mineral species affords many instances of isomorphous silicates whose formulas differ by WO2M2, as the tourmalines, and the silicates of alumina and magnesia ; while the latter, with many zeolites, exhibit a similar difference of WO2H2 [O in these formulas = 8]. The relation is in fact that which exists between neutral, surbasic, and hydrated salts." It was further declared that the carbon-spars must be repre- sented as polycarbonates, having not less than from " twelve to eighteen equivalents of base replaceable so as to give rise to a great number of species " ; while the variations in the calculated atomic volumes of these car- bonates were said to "indicate the existence of several homologous genera, which are isomorphous." § 17. These conceptions of progressive series of more or less highly condensed molecules of polycarbonates and polysilicates, and of similai'ity of volume for isomorphous species, were developed more at length in a second paper published in the same year, 1853, on "The Constitution and Equivalent Volume of Mineral Species." * It was therein explained that the formulas of homologous bodies may be represented as series in arithmetical progression, in which the first term may be either like or unlike the * Anier. Jour. Science, 1853 (xvi., 203-218), and, in abstract, in the author's Chem. and Geol. Essays, p. 438, etc. 290 A NATURAL SYSTEM IN MINEKALOGY. IVIII. i!(ili common difference ; both cases being, it was shown, illus- trated in the chemical history of mineral species, in- cluding carbonates, silicates, and oxyds. Similar views were also then extended to nitrates and sulphates, as well as to chlorids and to sulphids. The simplest atomic formula of the carbonates being CMO3 (C = 6 and = 8, according to the molecular weights then in use), the rhombohedral carbon-spars were referred to three genera represented by ^(CMOa), namely: (1) calcite, w = 30 ; (2) dolomite, siderite, and diallogite, w = 36; and (3) smithsonite and magnesite, w = 40. For the prismatic species, aragonite, like calcite, belonged to a genus with w = 30 ; while for strontianite, cerusite, and bromlite, n= 25 ; and for witherite n = 22. The volumes of the rhombohedral species deduced from these formulas were from 550 to 560, and for the prismatic species from 500 to 510. These arbitrary molecular weights and vol- umes were, at the time, supported by comparisons with those deduced from the formulas of the rhombohedral red-silver ores and the prismatic bournonite, and farther by the volume of the compound of glucose and sodium- chlorid, regarded as homoeomorphous with calcite, with a density of 1.563, which, doubling its empirical formula, gave a volume of 558.5. The various alums, if their for- mulas be doubled, give in like manner, as was shown, vol- umes of from 543 to 561. § 18. Extending to the silicates the same notion of polynierism which had just been applied to the carbon- ates, the existence of various polysilicates was admitted. Thus the formulas of spodumene, diopside, hudsonite, and woUastonite were described as pr» anting a homologous series of the first kind, in which the first term is the same as the common difference, "represented by ^(SiaMOa), the respective values of n being 30, 26, 24, and 22." Spodumene was then, chiefly on crystallographic grounds, compared with the pyroxenes. The excess of silica above the bisilicate ratio, met with in some amphi- VIII.] A NATURAL SYSTEM IN MINERALOGY. 291 boles, was referred to as an example of a homology of the second kind, in which the common difference is unlike the first term. To these species there was assigned an equiv- alent volume approximating to 460. In support of this vt)lume it was noted that the various orthophosphates and ortharseniates of sodium, with I2H2O have, according to Playfair and Joule, equivalent volumes of from 233 to 235, while ferrocyanid of potassium gives 230, lactose 234, and piperine (with a density of 1.244) 476, or about double tiiese numbers. Other species, as it wtis pointed out, iuive apparently an equivalent volume of 430, and still others about 200, or some multiple of this number. Whether the weights thus assigned to various silicates iuid carbon-spars might represent their chemical equiva- lents, or some portion thereof, they in any case served to show the relative condensation of matter in the different s[)ecies compared. § 19. This subject was continued a few months later, in a paper read at Washington in May, 1854, before the American Association for the Advancement of Sci- ence, entitled "Illustrations of Chemical Homology."* Therein were reviewed and re-affirmed the teachings of the two papers of 1853, while the principles of homology were farther exemplified, and it was maintained that homologies may exist alike between species differing by /((M2O2) and ^(HaOa), and even between those related s])ecies which differ in the proportion of silica, so that the liitio between silica and bases has but a specific value, it was farther contended that the water contained in a jj^i'eat many hydrated species often described as altered silicates, was to be regarded as not of subsequent intro- duction, but an original and essential element cf the spe- cies, as is admitted to be the case in the zeolites. § 20. In the second paper for 1853 was considered the * Proc. Amer. Assoc. Adv. Science, 1854, pp. 2.37-247; also, in abstract, Atiier. ,Jour. Science for September of the same year, and noticed, with extracts, in tlie autlior's Clicm. and (Jeol. Essays, p. 4:58, et seq. I,*;- ' llllfti 292 A NATURAL SYSTEM IN MINERALOOY. [vm. question of chemical notation and formulas, which was farther illustrated in the paper of 1854. At this time the question of the atomicities of the elements had n ">t yet been discussed, and the distinction between univalent and bivalent metals, suggested by Cannizaro in 1858, was un- recognized. The symbols then used for both of these stood for one atom, or for the proportion which in the so- called protoxyds is united with eight parts by weight of oxygen. In sesquioxyds like alumina, however, recog- nizing the trivalent character of AljO^ (27-|-24), it was by the writer regarded as corresponding to three atoms of oxyd of aluminium = 3alO. Silica, which, following Berzelius, was then generally written SiOg (21-|-24), became 3siO. With this notation were constructed atomic formulas, the elements now regarded as diatomic bejng confounded with monatomic elements, and, like these, represented by capital letters. Thus the common atomic formula for bisilicates, as given above, was written ^(sigMOg), and spodumene, w = 30, was made si6o06o(al24Li4Na2)03o. Similar atomic formulas are still employed in these pages, using, however, small letters to represent an atom of any element, whether univalent, like sodium or chlorine ; bivalvent, like calcium or oxygen ; trivalent, like boron ; quadrivalent, like silicon, titanium, and carbon; or sexvalent, like the double molecule of aluminium. The above general formula is thus now written w(sir.mi03), and that given for spodumene (si6oal24li4na2)o9o. In order to distinguish the atom of ferrosum = ^Fe, from that of ferricum = ^Fe, the former is written fe, and the latter fi, while manganicum, corres- ponding to manganic sesquioxyd, is mni. § 21. The M in the general formula M2O2, employed in 1853, was thus made to represent an atom either of prot- oxyd or sesquioxyd, and in 1854 a farther generalization was attempted. The boric, titanic, tantalic, and niobic an- hydrids were reduced to the same atomic formula as silica, and, moreover, in view of the variations in the silica-ratio II. ras bhe yet and un- tiese B so- \i of ecog- t was ^toins awing +■24), vucted atomic d, like ommop ^e, was g made Lxe still tters to ent, like oxygen ; Itanium, ecnle of us now lodumene atom of e former , corres- iployed in Ir of pi'o^ jralizati'^'^ Iniobic an- la as silica, Isilica-ratio VIII.] A NATURAL SYSTEM IN MINERALOGY. 293 in related silicates, like feldspars, sc*ipoHtes, and micas, and the supposed rei)lacement of silica by alumina in cer- tain amphiboles, it was suggested that the old distinction of acid and base, recognized in the dualistic hypothesis, might be set aside. M, in the generalized formula as then written, n(M02), would then represent not only Na and Ca, but al, si, b, ti, and ta, as well, and "to this type, which is also that of the spinels, all silicates may be re- ferred, except a certain number which, like eudialyte, sodalite, and pyrosmalite, contain metallic chlorids; hauyne, nosite, an'^ lapis-lazuli, which contain sulphates; and cancrinite, which holds a portion of carbonate. These are respectively basic chlorids, sulphates, and carbonates, and are represented by (MaOa)^. MCI, by (M202)w. SaMjOg," etc. To these should of course be added the basic fluorids, or oxyfluorids, like chondrodite and topaz ; and oxysulphids like helvite and danalite. It was then said, "The above formulas are intended to involve no hypothesis as to the arrangement of the ele- ments, for in the author's view each species is an individ- ual, in which the pre-existence of different species that may be obtained by its decomposition cannot be asserted." The importance of this notation, proposed in 1854, will be apparent when we come to consider farther the question of atomic volume in its relation to mineralogical classifi- cation. § 22. Another and an important question, connected with the complex constitution which had been assumed for silicates and carbonates, was considered in the paper now under review. The high molecular weight assigned to the polysilicates admitted the presence therein of many atoms of base, and of partial replacements ; while the existence in crystalline species of visible mixtures of for- eign matters also served to explain the presence of small portions of many elements detected therein by chemical analysis. It had, however, become apparent that there are variations in composition which can scarcely be ex- 294 A NATURAL SYSTEM IN MINEKALOGY. tVIlI. !iifi:;ii"«fi plained in either of these ways. Delesse had already noticed that in the homoBomorphous tiiclinic feldspars the silica-ratio appears to vary continuously between albite and anorthite, and was disposed to regard the feld- spars intermediate in composition between these two as varieties only.* Scheerer, also, had in like manner ex- pressed the opinion that the various feldspars were to be regarded as combinations of anorthite with labradorite, albite or orthoclase, or of labradorite with albite. Yon Waltershausen had, however, given a more definite shape to the notion already in the minds of chemists, when, in 1853, he proposed to admit three typical triclinic feld- spars, anorthite, albite, and krablite; the latter a sup- posed highly silicious species with the atomic ratios, 1:3: 24, since generally regarded as a mixture of albite with quartz. These three feldspars, according to him, " alone have the right to be regarded as species in miner- alogy. . . . All other feldspars, labradorite, andesine, oli- goclase, etc., are merety mixtures of these," and were conceived by him to be built up " of infinitely small crys- tals of anorthite and krablite, or of anorthite and albite." f § 23. At the time of writing, in 1854, I was ignorant of the lately published conclusions of Von Waltershausen. I had then made an extended series of analyses of these feldspars, from the Norian recks, and, rejecting the hy- pothesis of Scheerer, to which I referred, attempted to give the matter a more definite form by pointing out that anorthite and albite might be represented by a common formula, which, if a molecular volume of about 402 were assigned, would be 32(M202) ; the two polysilicates being respectively, in the atomic notation adopted, (sigiala^Cag) * Delesse, Ann. des Mines, 1853 (5) iii., 376. Scheerer, Pogg. Ann. Ixxxix., 19, cited in L. and K. Jahresbcriclit for 1853, p. 105. t Sartorius von Waltersliausen, Uber die Vullcanisclien Gesteine in Sicilien und Island, GiJttingen, 1853. For this reference, and for other notes on the literature of this question, I am indebted to my friend, Dr. G. F. Becker. 1. VUI.J A NATUKAL SYSTEM IN MlNEltALOGY. 296 ,V3 en Id- as ex- be lite, Von inipe n, iu feia- sup- atios. albite , him, miner- ne, oU- { were ,11 crys- be n^nd rnovant ^bauiscn. ;{ tbese the by- mted to out tbat jcommon t02 were ies being Ig^aliMCas) •ogg. An«- Jtesteine i" [a for otlier my friend, 0^4 and (si48ali2Na4)Oe4. Petalite having the vohime of these, and its composition not being then definitely set- tled, was referred to the same general formula, while orthoclase, from its less density, was conjectured to be 30(M./)2). As regards the homoeomorphous triclinic feld- spars, it was then said that " between anorthite and albite may be placed vosgite, labradorite, andesine, and oligo- clase, whost composition and densities are such that they all enter into the same general fornn^la with them, and have the same equivalent volume. The results of their analyses are by no means constant, and it is probable that many, if not all of them, may be variable mixtures of albite and anorthite. Such crystalline mixtures are very common ; thus in the alums, aluminium, iron and chrom- ium, and potassium and ammonium, may replace one another in indefinite proportions. . . . Heintz has shown by fractional precipitation that there are mixtures of homologous fatty acids, which cannot be separated by crystallization, and have hitherto been regarded as dis- tinct acids. The author insists that the possibility of such mixtures of related species should be constantly kept in view in the study of mineral chemistry. The small portions of lime and potash in many albites, and of soda in anorthite, petalite, and orthoclase, are to be ascribed to mixtures of other feldspar-species." § 24. These conclusions were reiterated, in 1855, in a paper giving the results of my chemical studies of these feldspars (when Scheerer's hypothesis was noticed), and it was said that similar views " must also be extended to the scapolites." * Some years later, in 1864, Tschermak f put forth a view similar to that advocated by Von Wal- tershausen and myself, and maintained that the feldspars proper were reducible to three species, adularia or ortho- * Examinations of Sc ne Feldspathic Rocks; L., E. & D. Philos. Mag., May, 1855. t Tscliennak, 1864, K. K. Academie Wissenschaft, Wien, and Pogg. Ann., 1865, v., 139. See also tlie author's Chem. and Geol. Essays, p. 444. 296 A NATURAL SYSTEM IN MINERALOGY. IVIII. li/-: clase, albite, and anorthite. While recognizing the fact that certain potash-soda feldspars (such as j)eithite) are made up of alternations of orthoclase and albite, he fur- ther concluded, as I had already done, " that oligoclase, andesine, and labradorite appear to be members of a great series, with many transitional forms, and may be regarded as isomorphous mixtures of albite with anorthite, some- times with small admixtures of orthoclase." § 25. With regard to this conception of the nature of these intermediate feldspars, it should be noted that the chemical difficulties in the way of verifying it are much greater than in the case of soluble compounds, where, as in th^ case of the fatty acids just mentioned, solution and separation by fractional preci^^itation are possible, or where differences in volatility may be appealed to. While a definite feldspar-species having the composition assigned to labradorite doubtless exists in nature, it is nevertheless true that a mixture of proportions of anorthite and albite containing equal parts of alumina would give a, centesi- mal composition identical with that assigned to labra- dorite, just as the composition of a fatty acid may be simulated by a mixture of its higher and lower homo- logues. In so far as the view of Von Waltershausen and myself, since adopted by Tschermak, is true, the action of acids capable of attacking the basic feldspars will enable us to discriminate between admixtures and definite inter- mediate species. That the latter should occur in nature is, a priori^ probable from the composition of the parallel series of the zeolites, in which appear well crystallized species, having the atomic ratios (excluding the water) of the intermediate feldspars, and also from the evidences of species like hyalophane and leucite. The late observa- tions by Tschermak as to the action of acids on various intermediate scapolites, to be noticed farther on (§ 75), go far to show that these are not admixtures but integral compounds. § 26. In concluding the paper of 1854, which I have 11. ict ire ar- ise, •eat (led tme- te of b the nuch i-e, as I and e, or While signed ;hele8S albite entesi- labra- nay be homo- en and 3tion of enable inter- nature parallel staUized water) vidences observa- i various 1 (I 75), integral ;e VIII.] A NATURAL SYSTEM IN MINEIlALOGY. 297 here reviewed, it was said with reference to the problem of a natural 8y.stein in mineralogy, then, as now, before the writer: — "No mineralogical classification can be complete which does not take into account both the chemical and the l)hy8ical characters of species ; and the connection be- tween these, which is shown in the relation of equiva- lent weight to specific gravity, must constitute an iir^^or- tant element in a natural system. Guided by their physi- cal characters and composition, we bring together such homoeomorphous species as belong to one chemical sub- type, and from the densities fix their formulas ainl com- parative equivalent weights. From the ct)mparison of the formulas, and the associations of these different min- erals, we must also decide which are to be considered as mixtures and which are true species. Until we shall have determined with certainty the comparative volumes of dissimilar crystalline forms, the relations of species differ- ing in this respect must be decide'^, by their affinities, and their places in a homologous series must remain unde- termined. In this way we may hope to arrive at a miner- alogical classification which shall satisfy alike the chemist and the naturalist." § 27. Before going farther, it seems proper to advert to the history of the notion of polymerism in silicates and carbonates, which enters into the views maintained in the author's papers of 1853 and 1854 ; and to show its rela- tion to the views previously put forward by Auguste Laurent. He had already, in 1847, proposed to reduce all natuial silicates to a small number of types, correspond- ing to the observed atomic ratios. These yield both neu- tral and basic salts, according to Laurent, who, moreover, in his notation, admitted, in order to explain the complex results of chemical analysis, a divisibility of molecules to which he assigned no limit, and supposed that protoxyds and sesquioxyds might, within certain limits, replace each other indefinitely. He also extended a similar view to 298 A NATUUAL SY8TKM IN MINEKALOOY. ivm. i the bonitos,* In a Hubsticjuent memoir, in 1841), Laurent criticised the iul)itrary formulusj proposed ly chemical miiiendoyists, and showed that tlie rehitions therein bet forth were often but ai>proxiinations. It was j)ointed out by liim that in many rehited species, as for exami>le in the various micas, the atomic relations between sescjuioxyds and ])rot()xyds were not constant, and it was argued that •tliese two chisses of bases, and water, were ca[)able of rephiclng each other mutually, within certain limits, in ratios which, as represented by him in atomic formulas, seemed to be indelinite. He also insisted on the impor- tance in silicates of small portions of water, which, though generally neglected in the formulas, ought not to he regarded as accidental. This later pai)er,t however, while rellecting the perplexed state of chemical mineralogy, fails to propose any solution of the difficulties. The reader will note the broad distinction between the simple formulas, with an indefinite divisibility of molecules, adopted by Laurent, and the complex formulas, necessa- rily including many atoms of base, employed by the writer, further supplemented by the conception of crystalline admixtures of homceomorphous species. § 28. It was not until 1860 that the doctrines of liigh equivalents, and of the existence of polycarbonates and polysilicates, maintained by the writer in 1852 and 1853, found an advocate; when Ad. Wurtz again put forth the notion of polysilicates, explaining their genesis from the union of several molecules of silicic hydrate and the suc- cessive elimination of water. He cited in this connection the example of the metastannates of Fr(imy, which contain five quadrivalent molecules of tin. Wurtz did not, how- ever, attempt in any way to discuss the difficulties pre- sented by the composition of the native pol^-silicates (for * Comptcs Rendus des Trav. de Chimie, July, 1S47, from Comptes Keiidus de I'Acad. xxiii., 1050, and xxiv., 5)4. For an analysis of Laurent's memoir by the writer, see Amer. Jour. Science, 1848, v., 405. t Sur les Silicates; Comptes liendus des Trav. de Chimie, 1849, pp. 250-288. VIII.) A NATURAL HYHTKM IN MINICItALOOY. 299 certain of which ho proposed Btriustiirui fonuuhis), or to fix their inolecuhir weights, uiul he seeiiis to hiive over- looked the previous contributions to tlie subject by the present writer.* § 2\). In 1859, in n paper on "Euphotide and Saussu- rite," t the writer, luiving made an exteniK'd ciuMuical and luineralogical study ol' the typical saussuritc, as found in the euphotide of Monte Rosa, in Switzerland, showed that it was not a feldspar, as generally supposed, but a finely granular or compact silicate having the hardness of quartz, a specific gravity of 3.305-3.385, and the composi- tion of a lime-soda ej)idote, or a zoisite, to which latter species it was referred. In this connection he called attention to the observation of Rammclsberg, that zoisite is apparently identical in centesimal composition with meionite, the most basic of the scapolites, which has a hardness of 6.0, a specific gravity of 2.6-2.7, and is readily decomposed by strong acids. It was further noticed that, while boiling concentrated sulphuric acid did not attack pulverized saussurite, " it was, however, partially decom- posed by tliis acid after having been strongly ignited." Attention was then called to changes produced in the denser silicates by heat, and it was noted that epidote, according to Rammelsberg, has its density reduced from 3.40 to 3.20 b}"- ignition, while saussurite, accoiding to the original observation of Saussure liimself, is converted by fusion into a soft glass having a density of 2.8. The specific gravity of garnet was found by Magnus to be reduced one fifth by fusion, and that of idocrase from 3.34 to 2.94.| The silicates thus modified by heat are, like * Ad. Wurtz, Rep. de Clilmle, 1800, ii., 464; also .lour. Chem. Soc. of London, 1862, p. 387; and Lemons de Philosophic Chimique, 1864, p. ISO. See farther, on polysilicates, Naquet, Principcs de Chimie, 1867, i., 175. t Contributions to the History of 2uphotide and Saussurite, Amer. .Jour. Science, ia">(), xxvii., .3:l6-a4!). t The observations of Greville Williams on beryl show that this min- eral, having a density of 2.65-2.01), when fused before the oxyhydrogen blowpipe, gives a clear glass, which may be scratched by quartz, and has a density of 2.40-2.42. The fusion of quartz gives in like manner a glass -ill hi \\ it,' 300 A NATURAL SYSTEM IN MINERALOGY. [VIII. f^HfJtf 11 i| meionite and nephelite, decomposable by acids, and all these facts were adduced as evidences that the action of heat is to reduce such complex silicates to simpler and less dense forms. § 30. In conclusion it was said that " the two silicates zoisite and meionite offer a remarkable example of that isomerism in mineral species upon whose importance I have long insisted. The relation of the specific gravity to the empirical equivalent weights of minerals must enter as an essential element into a classification which shall unite the chemical and natural-historical systems. Simi- lar isomeric relations exist between cyanite and sillimanite (fibrolite), rutile and anatase, and, as I have elsewhere endeavored to show, among the carbon-spars. It becomes necessary in the study of mineral species to determine their relative equivalent weights, to which specific gravity must be the chief guide." § 31. The relations of the members of the scapolite group * as a series parallel to the feldspars, already pointed out by the author in 1855, were not lost sight of, nor their connection with saussurite, but were the subject of a communication to the French Academy of Sciences, in 1863, which was translated by the author and published at the time, in the American Journal of Science, as already to which a density of 2.22 has been assigned. Williams, in repeating the experiment with rock-crystal, of density 2.65, obtained before the oxyhy- drogen blowpipe fused globules, which in five experiments gave a specific gravity of 2.17-2.21. He noted, moreover, that alumina thus fused, as in the experiments of Gaudin, becomes crystalline on cooling, and has a density of only 3.45; that of corundum being about 4.00. The crystals of alumina got by the method of Fremy and Feil, which consists in decom- posing an aluminate of lead by fusion in contact with silica, have, how- ever, all the characteristics of corundum, and a density of 3.9-4.1 (Gre- ville Williams, Proc. Roy. Soc. London, 1873, p. 409, and also Fouqu^ and Michel L^vy, Synthase des Mineraux, etc., p. 222). * The scapolites have very lately been taken up and discussed from the author's point of view by Tschermak, Monatshefte der Chemie, December, 1883, as will be noticed farther on. The slight change in the empirical formula of meionite suggested by ' schermak does not atfect the present argument. VIII.] A NATCJIIAL SYSTEM IN MINERALOGY. 301 ng the oxyhy- jpecific as in has a stals of decom- e, how- 1 (Gre- Fouque .d from hemie, e in the ffectthe noticed in § 10.* In this paper, after recalling the general argument so often set forth as to the principles of a new system of mineralogical classification, it was said, " Meio- nite, with the oxygen-ratios 3 : 2 : 1, is the most basic term known of the series of the wernerites (scapolites). The proportion of silica in these minerals augments until we reach in dipyre the ratios 6:2:1, with a density which does not exceed 2.66. We might then expect to find a silicate which should be to dipyre what zoisite or saussu- rite is to meionite, and Mr. Damour has recently had the good fortune to meet with such a mineral in a specimen of jade from China, of which he has given us the descrip- tion and the analysis. (Comptes llendus, May 4, 1863.) This substance closely resembles in its physical and chem- ical characters the saussurite or jade from Monte Rosa, of which it has the density, 3.34. It is a silicate of alumina, lime, and soda, and gives the same empirical formula as dipyre. We may expect to find between saussurite and this new species, to which Damour gives the name of jadeite, other jades, having formulas which will correspond with the wernerites intermediate between meionite and dipyre. ... By its hardness, its specific gravity, and its indifference to acids, jadeite is completely separated from the wernerite group, and takes its place alongside of zoisite or saussurite, with the garnets, idocrase, and epi- dotes." § 32. To this last succeeded the paper of 1867, on "The Objects and Methods of Mineralogy," already noticed (§§ 12-14), in which was given a review of the subject as discussed by me in various publications from 1853 up to that date. Before proceeding to show the systematic application of the principles already set forth, it is now proposed to consider farther the question of the relation between the atomic weights and densities, so often insisted upon in the above publications. The study * Compte Rendu de I'Acad., June 29, 1863, and Amer. Jour. Science, , 1863, xxxvi., 426-428; also the author's Chem. and Geol. Essays, p. .446. ff'i ■-[ :^5|,| 302 A NATURAL SYSTEM IN MINERALOGY. [VTII. of the so-called equivalent or atomic volumes of solid and liquid species, got by dividing the assumed equivalent weight of these by their specific gravities, — water being taken as unity, — has occupied the attention of many chemists since the early investigation of the subject by Le Royer and Dumas. The application of this method to hydrocarbonaceous bodies, or to hydrated or double salts of admitted high equivalent, is comparatively simple, but it becomes more difficult when, wo have to deal with such compounds as mineral silicates, for which, as in the case of feldspars, micas, epidotes, and tourmalines, the ingen- uity of mineralogical chemists has devised chemical formulas often exceedingly complex and difficultly com- mensurable. For all such cases I have shown that the atomic formulas already described furnish a simple solu- tion. § 33. In the atomic notation adopted by me since 1853, the ordinary cliemical symbols of the elements are em- ployed to represent one part by weight of hydrogen, or eight parts by Aveight of oxygen, and the proportions of other elements which unite with these respectively. In other words, the coefficients of the symbols of the elements in the ordinary notation are multiplied by the atomicities of the respective elements, for the atomic notation. The symbols in the latter are distinguished from those repre- senting molecular weights by the use of small letters, and, to |)revent the confusion which might otherwise arise from the abseiice of capital letters in the formulas, a coefficient is in all cases employed after the symbol of the element ; while in constructing condensed formulas the values of "m" may be represented by fractions. Thus, the general formula of pyroxene in the atomic notation being w(si2mi03), if the value of n be 30, and m =(ca6mg^fe6), the proper atomic formula of pyroxene will be si6o(cai5mgiofe5)o9o. § 34. But, as we have elsewhere shown (§ 21), the variable relations between silica, alumina, and protoxyds, in closely related species ; the intervention of boron and VIII.] A NATURAL SYSTEM IN MINERALOGY. 808 II. nd ng my by I to alts but iucb case igen- iiical com- ,t tbe solu- titanium, on the one hand, and of sulphur, fluorine, and chlormc, on the other, permit a farther generalization, by which silicates are affiliated to quartz, on the one hand, and to corundum and spinels, on the other. We thus arrive at a general atomic formula n(A-}-E), in which A represents an atom of silicon, boron, or titanium, or of hydrogen or any metal, and E, an atom of oxygen or sul- phur, or of fluorine, chlorine, or oxysulphion. Dividing now the molecular weight of the compound bj- w, we get the value of A+E, which is the mean weight of the indi- vidual or atomic unit of the species, whether this be oxyd, silicate, oxyfluorid, oxychlorid, or oxysulphid. It is this weight, designated a;-; P, which for each such species must be the term of comparison in fixing the atomic condensa- tion of the spec'cs. The mean unit-weight thus deduced, divided by the specific gravity of the species, Avater being unity, gives the v^dume, V, of the atomic unit. In sili- cates, the value of P is deduced by dividing that of the empirical atomic formula by the whole number of oxygen atoms, to which, in the case of oxyfluorids, oxychlorids, or oxysulphids, the number of atoms of fluorine, chlorine, or sulphur is to be added. In this way only is it possible to obtain direct comparisons of volume between different mineral species, as was indicated in 1852, and will be fully shown in the third part of this paper.* § 35. The principles hitherto maintained by the author as the basis of a natural system in mineralogy, may be resumed as follows : — 1. The conception of high equivalent or molecular * Dana, in his inquiry into the subject of atomic volumes in 1850 (Amer. Jour. Science, ix., 221), proposed to divide tlie volumes tieduced from the empirical chemical formulas by the number of atoms of ele- ments in these formulas. Thus (0=8), SiOa, AljOi, and CaO, were, in the notation adopted by him, supposed to contain respectively four, five, and two elemental atoms, whereas in atonuc notation they evidently correspond to thre;-, three and one oxyd-units. Ilenc(>, as I long since showed, the results obtained by such a discussion of atonuc vohunes were fallacious. (Amer. .lour. Science, 1853, xvi., 214.) '■'i,: 304 A NATURAL SYSTEM IN MINERALOGY. [VIII. 'nil weights like those of the carbon series in so-called organic chemistry, extended to all mineral compounds; as was especially maintained for th^ carbon-spars, the spinels, and tl'o various natural silicates, and illustrated by the hypo- thesis of polysilicates and polycarbonates, with many atoms of base. 2. The conception that the laws of progressive or ho- mologous series, previously recognized only in hydrocar- bonaceous bodies, must be extended to mineral species, and are of universal application. 3. The conception that the variations observed in the chemical composition of such mineral species are due, not only to their highly poly basic character, but also, in cer- tain cases, to indefinite admixtures of homoeomorphous species, as pre piously indicated by Delesse, Scheerer, and Von Waltershausen, extended and generalized by myself, and subsequently adopted by Tschermak. 4. The attempt to fix the molecular weights of such compounds as the polysilicates and polycarbonates from their densities as compared with those of species the mini- mum molecular weights of which are otherwise determined ; and the assumption that for homoeomorphous solids, and probably for all solids, tlie molecular volumes are identical. 5. The adoption of atomic formulas to represent the •composition of mineral species, and the showing that com- parisons of the volumes or spatial relations of complex species, like the silicates, should be based on the numbers which are deduced from these atomic formulas, and which lapresent the relative volumes of the unit-weight in the species compared. P being the unit-weight got by divid- ing the empirical molecular weight by the number of oxyd- atoms in the formula (including any chlorid, fluorid, or sulphid atoms which may be present), and D the specific gravity, the volume of the unit, designated as V, is repre- sented by the quotient obtained by dividing P by D. 6. The showing that, in related and homologous species, the hardness and chemical indifference are inversely as the VUI.] A CLASSmcATIOK OF SILICATES. One value of V: or in ni-u the co„de„sati„;, tkTrZZ7^' "'"' f'^ "'"'^"^o "ith ?36 rn''^"" "'"''"™^™'^ ™' ™'CATES oWfi,,„„„ „, „ ■» J;h -eo„d part of this papL. to "t'C Silicates, for the reason tLt '', I J' ''^°"'" *''e Natural <="..T.lex a„d the la.^es"': /''^^ '"«-'' "''^ ">"«' P'ysical and their chemieal^v? ''^ ""*■>'<' species, their t'oiong,,,, studied tClor:?:*^ .T" ''»^° ■»°- this It may be added that the t f 7 "**'' S™"P- To employed a olassifloat on blsedl H '"' '°^ """'/vears, """.gement of his own pr vate e 7!-''""""'"''^ ^"^ *he »"'oates. These „,ay be re^J^d '°" "' '^' """^ great natural order and t-If*' ^'' "^ '^onstitutinj; one i'"Po.;ta„ee of coSdert TrLr'", '"^^'■■«"=" *° "- chenueal and tl,e physical hil? f ™' '>''*em alike the that a fundamental SuonZr T''''' " """^ "^^ ^-^ hy their chemical eons t"Hn •°^"' '^ """ Presented o^yl or sesqnioxvd°d o^'h^i""'"*'-^ <='*-' P™t- «aso„theorlrSmca^1s^^L^^''"''^''• ^""^ ^hieh P™t„siHcate,Protope:S"1rd^:!^'''7 ™ The names of protow.! 7 J ^eradicate, pounds, and of peCyd^i'lP-'^lt for ferrous com- ' a"d sesquisalt, for fe,i.io com„ ' ' ""^^ »f sesquioxyd ■sfs: and when, in naminrr K"'"'f'^""""'"- *» <=hem- hecame necessaiV to sele fa er™". T''^^ "^ ^'"^"tes, it impounds, aluminic comlu 1 'Vr®"""' "'^^ ferric »"d aluminie ^esqnioxyZ partMv 7""' ''"^'' fe™" sutured to substitute for sesoS ? " '"'^ °'^'"' I "■hoate the shorter and mornun)^'-" "'"^ P^tosesqui- ^f e and Protopersilicate With r""" "'""' "^ P^'i"- ''h.oh also include chrome and 1''"'"°''^'' hases, '-.rco„i«, since, uotwithstond'nt?h *'""; "^''^' '^ '™ged ^"""■""-■'-■^■-■-^""^atsraSr::,:^^^^^^ 306 A NATUllAL SYSTEM IN MINERALOGY. [VIIX. ( i i;; place it at ihe side of alumina. Here also bismuthic oxyd probably belongs. Boric, titanic, niobic, and tantalic oxyds, all of which are found in silicates, are ranged, as already stated, with silica, which they are regarded as replacing. § 37. Inasmuch as zirconia and chromic and manganic oxyds are but exceptionally present in silicates, and ferric oxyd, though more commonly found than they, is much less frequent therein than the alumina which it sometimes replaces, it may be said that it is essentially the relations of alumina to the protoxyds and to silica which we are now called to consider. Native silicates may be divided into those with and those without alumina, the latter di- vision constituting the first sub-order — Protosilicate. Again, the aluminiferous silicates either contain combined protoxyds, constituting the second suu-order — Protoper- aillcate ; or are without protoxyds, making the third sub- order — Persilicate. The presence or absence of combined water, — it being an element widely diffused in nature, — is of subordinate importance in the study of the silicates. Upon the general distribution of silica and alumina in the crust of the earth, and the relations of these to each other, to protoxyd-bases, and to igneous and aqueous solvents, is based the whole genetic history not only of the three sub-orders of silicates, but of quartz, and the non-silicated oxyds. The .affinities which determine the nearly contempora- neous formation of protosilicates and of protopersilicates are displayed in many different and unlike conditions, which merit especial consideration. This distil ction is well seen in the basic crystalline rocks, wherein pyroxene and chrysolite, often with magnetite, are found side by side with feldspars. Whether this separation, which may be supposed to have taken place in a plutonic magma, was effected with or without the intervention of water, is im- material to our present inquiry, since we know that the chemical affinities involved lead to similar results alike iu \* vm.] A CI,ASSmcAT,ON OP SIUCATKS. »;.ecessively, chrysolite" mateUe^^'''"^ basic ,„ag,„a, That 3m.ilar affiniUes eonS'l^^™''""' "'"' '"'^ar f : '^™'^<' temperatures s sW^'^^ "i ''™'''"=^ "^ ^v^ter cret,o„ary granitoid veins wW \^ ""'^^^^ '« "on- »d pyroxene are fo„„d ~ilT 1'?"''' '"''P''i'>ole, •te, ,„icas, garnet, and epidote of '""j '!,'*pai., scapo- tlie one hand, and with n amet,/ •'""' '''^ 'J'"«'t2. on on the other. "'aguetUe, spinel, and oorundLm" P™topm-siHoaTst''prt!nTed''r'l" °' P™'o«ilicates and "■^t- . In the veins rdgeodst T"''""' *"■» ^^^ seen, side by side fhl ^ ? "'""<' '" such rock, »,.„ "atolite, a„dV:';h„ : ^d'th "'"' P-'^^e. otnite sented by prehnite,'^>pid;te and thr"*"?'™"""''^ 'W "■ore rarely by orthoclase and lu 7"""' '"'^'''^ and ™aime,-b„th quartz andtt^e ^ V'?"™' ""<• *°- The same distinction is owf. ''""«^ "'«<> P™en' forming in the channels of .."' .'" ""^ P'oducts now P!-tolitic and ^eoliti 1 lilte" Ir *''"™'" "^""^^ wherl ^'fferentiation less marked wL' "'^•'"'f '^d. Nor is the Da»br«e, water at a high teml '."^ '" "'" «'^Perin,ents of P''otosilicate allied to Xnitl? If "! ""'^ "P°« g'«««- A PKoxene and quartz, fte X f/" '"""'"'' '»ge?her with ^'l'°a, with a portion of »,„'""■' '°'»««n ^etainin^ i'eated solutions, holdinJ oh'"'- ^"'"' ^™ilar super *-»encs, crystal's of t fo Se 0?^"°"" "^ '"-^ X 8 39 Tn fV. '^ known that thc^ula ^1 r"^ ""'f *™"«f<'™atirs ft "■» -- of atmos'pht::\ttreCir*-^-«'' - ^"^^^« *^ieir complete 308 A NATURAL SYSTEM I^ MINERALOGY. [VIII. decomposition. The lime and magnesia of amphibole, pyroxene, and chrysolite, are thereby dissolved, together with a large proportion of the silica itself; a part of this, however, according to Ebelmen, remains behind, together with the iron, changed from a ferrous condition to that of ferric hydrate. In the sub-aerial decay of such protoper- silicates as the feldspars and closely related species, the protoxyd-bases, chiefly alkalies and lime, pass into solu- tion, together with a large part of the silica ; and the alumina, united with the remainder of the silica, and with a portion of water, remains as an insoluble compound, which in many cases has the composition of kaolin. This decay of the feldspars plays an important part in terres- trial chemistry. The process is slow and gradual, and the feldspar softens and becomes disintegrated before the loss of protoxyds is complete, so that the clays thus formed still retain, in many cases, a portion of alkali, which may amount to two or three hundredths (a?i^e, page 254). The decomposition of the more basic feldspars and feldspathic minerals will be considered farther on, as also the genesis of various micaceous and colloid or clr.y-like persilicates. § 40. From the subsequent transformation of clays more or less completely deprived of alkalies, are appar- ently derived, in many cases at least, muscovitic micas and tourmaline, together with the crystalline persilicates, kaolinite, pyrophyllite, andalusite, cyanite, fibrolite, and related species. The micas just mentioned are more stable under atmospheric influences than the feldspars, while those which, like phlogopito and biotite, abound- in protoxyds, yield readily to decay. The harder and gem- like protopersilicates resist to a greater extent this pro- cess, and the more common species of these — garnet, epidote, and tourmaline — are found unchanged in sands, together with persilicates, such as andalusite, topaz, and zircon, and with quartz, corundum, spinel, and menaccanite. Thus the natural processes of sub-aerial decay destroy the protosilicates, and transform the predominant types of VJII.j A CLASarriCAl..^ OB- slLrOATES. 309 protopersilicates either inf^ more stable ty,.:'*/ ,''::;:, '■"'i'^ "'"■'""'- nnd oate», m all cases with the stm,.,, f ' '"' """ l'««"i- protosilicates, that is to savT x""' "^ "'" "^"'"^"^o of W these, v.l,ile iron I lilt''" "'"' ''■'""-y'l-bases , »tote, the alkalies ami li„ e " 'f/ , '" ;,'" """'"We ferric protoslioates, pass i,.,„ he „I litio' 1 '""S"'^'" '^ ">« s.I,ea hei.^ liberated in a so,„h :'!' ^ "' ""■'"'""'-■ "- -ereted i„ basic rrcks, IjZ":' -"-- -cl those I>rocess of solution of s licatL „ ""'"' ^'■l'to™nea„ conditions as yet in.ne f^! .T'"'"'"''' ^oes on under •■■" er the inmrence oHlS S.r'"'"' """^ '^''"•'ly •;atters thus dissolved co^e not ont'?, "'"""''*• J"™" "'e ''ei>os.ted in the forms of zeo 14!,°^ . !■ f "'"l^^'ioates garnets, and tourmalines, but a so tl °\ '''''^"■■"•^' '«'<"'^' protos,lieates which take the fo"f ""'''"= »'"' »"«'ine okemte,apoi,hyl,ite,and wollalle 'f 'f *' ™"-"«te, solutions, coming in content ^""^ Protosilicate dioxyd, would bf decompred w^i "'"""P!'"''' """""'io »tes, but with dissolved ZlZ- "T"'""" <"' »•■"*"»- ;!ouble exchange silic^ts^slc, as'sf'^r'''' ^'^"' by fne, enstatite, chrysolite am 7h ^^P'^''^' We, serpen- and pyroxenes. It wi C menXrrlf ^"^ "'"P'"'^"'- s.l.cates may, like the felds Z b/f ."' ''"'' P™'"" 0..S and by igneous processeTand th fT' '"''''' ""^ "'l''^- ■■>» to origin must be drawn betwltt " '"'""' <"«'in<=tio.. cates, of both 8ubK,rdeTwhic,rr ] "'" »■' Vdrous sili- S'tion, and are oftl? " ''""' '" "q^oous de--,- q»rt., and the sa, e sprcir'";:' J"""' -'-'« "■"• with "'cks, and may be thfrl Z'f ^'^ *""'"' ''» P'ftonic -'oling igneous n, ss Cone tT'""""'"" *™" " reactions with magnesian saft? "u '""''""°'''' ^y "''^ S've rise to eompomX i" ' T^' ^ ''°""'' ««^<=hange, and the chlorites" whi e th r.' '''.*''' "■■'^"^*'"'> »*»« Wy to be sough TZ relr'^t "^ »'''™'""'<' '« Proba^ 'Iff hh M. 'r'i !l.. I ' 310 ^ ^ATUBAL SYSTEM IN MINEUALOOY. [VIII. U giving ovigiu to the l^^fji'^^.e latter, by »ub-ae.>al oates, it is the ^"^Z^^,.^ «- Fot^yi""--' aquenus action, wluUi, uj generates the V^'-f"'f'-.„„^ between the sub-orders ^ « 42. While the d""" "* °™ ^,,1, genetic history a e wllh have been ^/^rj,*' " a t holevev, remarkable generally well -l^fi''^'^' ^^'..ect the protosilieates with Ixanrples which serve *» «°^^^"^^ ^ i^ species of am- the protopersilieates. ^''"^'tosiUoates, there have phibole and pyroxene, wh.ch a e V ,„ j, „hich, while hitherto been »«''«l^'l "^''^^^ ;„ 'external characters, apparently identica «''h ^f^^^;,.,. Taking as a type of contain notable P0'-«^»« ° f ,"„e-magnesia pargas.te, we the aluminons a^P'''^"''''; „*J" yds, alumina, and silica, td for the atomic '»'- " ^mdle, silica 42.2. This, 2 : 1 : S, which >^«f«^\Sw, as has been suggested, it the alumina ^^P^^f ^^^Xwe tbe ordinary amphibole a portion of the silica, ^™"W S'v .„topers.l.ea e, ratio of 2 : 4. It is, l^^^f f ' „,,ilite; while the alu- having the same f °""^, *'\: ^the analysis of Liver- „i„ous species, S^<^:'"°f^'''fl^i gastaldite, a still more ridge, gives the ratios 3 ^ 2 j^8, a_^ J ^^^^^^^^ „a,,,ed by aluminous species, \-f-^- ^ ical characters have species like these, which from p y writing man- b'een compared with amphAole ^^^ .,^ ^j^,,, . „er the importance of «^«™"^^';°„ „gasite iu its ato™'" ogy. Analuminousa.igit«=ueartop g 213). "fations has also been examined by^Fo"q^ J^ ^^ ^^^ ^^^ From gastaldite in "'^1* the ato ^ ,^ ^^^ ^^^. and alumina are 1 : 2, '» J '"""P^hich the ratio becomes, gasite 1 : i. we h"^, " f ' d o o" "> *'= ''^^ '^""T"' lin humboldtilite,! : iV^d^o ° ^^^^^^^ ^ . ^ „, [I. VIM.] A CLASSIFICATION OF SILICATES. 311 ,te iW, Ate jili- ises, :der9 y are kable with i ixm- ebave , w\iUe cacters, type of site, we Ld silica, ». ThiSi [ggested, npWbole •vsUicate, B the aUi- oi Liver- stiU more ffered by •ters bave ^^ minerai- its atomic If protoxya k and par- ^o becomes, aluminous ,s 1 •• 1^' "' aese species to the protosilicates. In like manner, towards the otlier limit of tiie prot()i)ursilicates, we tiiul this ratio changing from 1 : 6 to 1 : 9 ayd 1 : 12, in iiulieolite, rubellite, and the muscovitic micas ; thus marking the transition to gem- like persilicates like andalusite and topaz, and to persili- cate micas like kaolinite and pyrophyllite. § 43. We may conceive the relation of the three snb- orders to each other to be represented by a design of two bands of equal breadth, but of unlike color, and of dimin- ishing intensity of color, protracted in opposite directions alowg a common course, for a considerable part of which the two bands overlie or rather blend with each other. The unmingled iJortions of these two color-bands repre- sent the protosilicates and the persilicates. As a result of such an arrangement, the protopersilicates, towards the protosilicate-end of the continuous series, include com- paratively little alumina, as in melilite, pargasite, and phlogopite, while towards the persilicate-end they hold but little protoxyd, as seen in indicolite, rubellite, the musco- vites, and pinite. § 44. It follows from what has been already set forth that the more or less arbitrary ratios generally assigned by chemists to various silicates, and deduced from empiri- cal formulas which in many cases represent but approxi- mately the results of chemical analysis, are not always to be regarded as exact. Thus, in examining the various for- mulas hitherto devised for protosilicates, we find that for the whole succession from chondrodite to apophyllite, the atomic proportions between the bases and the silica may be represented by some twelve simple ratios between 4 : 3 and 1:7. It is probable, in view of the complex constitu- tion, involving from twenty to thirty atoms of base, which we have assigned to these polysilicates, that, while some of these ratios are exact, others represent but approxima- tions to the truth. The same remark applies with equal force to the persilicates, where a like number of similar ratios is made to include all of the known species. In ,he ;^ 812 ^ NATURAL RVai^.i , 1 .f thpir composition, how- 'ever, U,e "-''^"V "-l-'^'';2;;:v'"^ tor th. t.l.u.a. silica are .eta.nca m ,,i,i„,tV gWen farther on views ot i,roto».l.«. U« • "U 1 <^ sui,-order» we note « 45. In tl.o study "f >>»''' ecies which „„f„y n,iuevalogical -»en.b - ',^. ,.,,.,. ,,,„,„v,,ances .litfer widely in '*"» " 'f ^ „„ ,ve exau.ine the lavgev become »tiU move -M'!""""' , . ^ protoneraiUoates. Here, '„d more complex B;-"') . "J, ('1,^ .eolites, the feldspars, £„r example, in the ''"' "^^ ^^^^ similar and hom-BO- „,d the "capolites, we nd l,hj^ J , „Uo o£ prot. „orphou» species "^ ^^^^ ,„i„uie. The same tlnng oxydsa«dalununa,thes.Uca s ^^,^ ,„,,.ovder; as may be observed among the n .^ ^,^ „,„.g„. X eorundophibte - -"f '^e latter also present ite with certain ■»«,»™;f;.^; .opposition, for in ditferen another type ot variations m eo P ^^ ,„,^„i yd and analyses oi "">»<"'""=•;„,!, it of the protoxyds, repre- siUca remains unchanged, that J ^^^^^ ,,,„,i,, sented by alkalies, IS vaiiahle. ,^^ physieally pear in the t""'™'' " XSmervariation, in the ratio ot similar present, at tlie ^"'^^iZ that of alumina to sd.ca. pMoxyd-bases to »l"™»';;f;^^ Zlv^^on, without sen^- These divergences in fT^^ ^^ovi strong arguments ^^0^ i;:::r:lt^"' »^ -^^ -^ «— '^ '- ^^ "^ o,|d.metals, on the »- >>»'l;-f, j^j devices have been placing elements, on the ofter .^ ^^ ,, proposed by ehemists, °* y'^^'jt^ue letters respectively. Ihe employment of R;~?j!;''i, „sed for an atom of Accordingly, in these P^S^' ehromicum, bismuth. \' vin.] A CLASSIFICATION OK SILICATKS. aia I li- te ch jes ger 3ve, »UBO- n"ot- iiing .; as ivgar- •eseut Eerent d and vei)ve- lea ap- ■sically Iratio of silica. X sensi- •uments a a true of m : si ; tliose of tlio iicM-siliciitcs, m : si ; and those of the protopersiliciitcs, in : m : si. § 47. Having sliovvn the wide chemical differences existing between tlie three great divisions of the (.rder Silicate, we [)roceed to consider tiiose dillerences, alike chemical and physical, which are found hetween species often having identical or similar centesimal composition. Physical characters, irrespective of chemical composition, constitute, in the language of Mohs, the " characteristic " of mineral species, aiid served as the basis of his system of classification. We propose to show that, by a re-examina- tion of these characters in the light of modern chemistry, it is possible to devise a new mineralogical method, which shall letain all that was good in the Natural History System, ai I at the same time bring it in accordance with the facts of chemistry, thus giving a veritable Natural System to Mineralogy. § 48. The great divisions marked by external charac- ters were made by Mohs and his school the basis of a system of classification, as is exemplified in his orders of Mica, Spar, and Gem, already noticed in §§ 5-6. We have there seen the heterogeneous nature of the order Spar, wherein — besides the genera, Sehilltr-Spar ; Dis- thene-Spar, including cyanite ; Triphene-Spar, comprising spodumene and prehnite ; Petal ine-Spar, for petaiite ; Azure-Spar, for lapis-lazuU and lazulite ; Augite-Spar, in- cluding pyroxene, amphibole, wollastonite, and epidote ; Felcl-iSpar, embracing adularia, albite, anorthite, Icorado- lite, and scapolite — there was a genus, Kouphone-Spar, in which were gi'ouped not only leucite and sodalite, but the characteristic zeolites, mesotype, laumontite, harmotome, analcite, chabazite, stilbite, and heulandite. With these was also placed apophyllite, while datolite was assigned to another genus, Dystome-Spnr. When, in 1844, Shepard divided the order Spar, by the separation from it of a new order, Zeolite, he transferred to this the whole of the species of the latter two genera. The order Spar of 6". ■i* ! Mil .;M'fi:|; MMM mtm 1*11*' > l< M' 'J 1 Cii 314 A NATURAL SYSTEM IN MINERALOGY. [VIII. Mobs, and the united orders of Spar and Zeolite of Shepard, thus included alike protosilicates, protopersili- cates, and persilicates of very various degrees of hardness and chemical unlikeness ; since not only datolite, apophyl- lite, and pyroxene, but mesolite and stilbite, leucite and albite, spodumene and epidote, and even cyanite, found a place therein. A still more heterogeneous assemblage was seen in Dana's order, Chalcinea, which comprised not only the order Spar of Mohs, but also the protopersilicate micas of his order Mica. § 49. We propose, while keeping in view the great chemical sub-orders already defined in »ur system, to group mineral species with more regard to these external characters than has hitherto been done. The obvious dis- tinctions of structure, hardness, and density, which separate protopersilicates, like garnet, staurolite, and tourmaline, from the micas, on the one hand, and from the feldspars, scapolite, and zeolites, on the other, though but imperfectly appreciated, underlay the division by Mohs into Gem, Mica, and Spar, and the necessity of a sub-division of the sparry or spathoid type was soon felt by Shepard. The need of this is most apparent in the great sub-order of the protopersilicates, where it will be seen that, alike on chemical and physical grounds, the natural line of division coincides with that betAveen hydrous and anhydrous spe- cies, — the latter including the feldspars, leucite, sodalite, and scapolites, and the former, or hydrospathoid, the zeo- lites. A similar distinction of hydrous and anhydrous spathoids is equally marked in the protosilicates. Upon the foregoing distinctions, and upon the still farther one which, in each sub-order, separates all these crystalline species from amorphous colloid compounds, we may pro- ceed to divide the various sub-orders into Tribes. § 50. Beginning with the Protosilicates, we recog- nize first among them a type of crystalline hydrous species, of inferior hardness and comparatively low density, which are decomposed by strong acids with the formation of a «♦ a ot ite eat to i-nal dis- irate iline, ipars, fectly ai tlie The er of ke on tvvisiou s spe- dalite, tie zeo- ydvous Upou ler one stalline ay Pi'o- e recog- species, ly, wbicb lion o£ a vm.] A CLASSIFICATION OF SILICATES. 315 jelly, or, to use Graham's phrase, pectise with acids. These hydrospathoids, which are represented by pectolite, apo- phyllite, datolite, calamine, etc., may be conveniently designated as the tribe of the PectoiitoUh. A second type, not very dissinnlar lo the first, but somewhat harder, and anhydrous, though still pectising with acids, is represented by willemite, tephroite, helvite, wollas- tonite, etc. These anhydrous spar-like species we designate as the tribe of the Protospathoids. In the third place, we note a group of species not unlike the second in general aspect, and, like them, generally anhydrous; which are, however, harder, and considerably denser, as appears from the reduced value of V. This group is represented by chondrodite, chrysolite, phenacite, amphibole, pyroxene, danburite, titanite, etc. Many of these species present a hardness and transparency which caused them to be in- cluded by Mohs and his school in the order Gem, and this gem-like or adamantoid character suggests for them the tribal name of Protadamantoids. As regards their rela- tion to acids, it may be noted that while the spathoid wollastonite is readily decomposed thereby, the corre- sponding adamantoid bisilicates, amphibole and pyroxene, are unattacked. The highly basic chrysolite pectises with acids, but the more condensed phenacite and bertrandite, with the same atomic formula, resist their action. The case of titanite, a titanosilicate, is peculiar, for the reason that the titanic oxyd, of which it contains so large a pro- portion, is soluble in chlorhydric acid. This deconiposi- tion of titanite was long ago studied by the writer, who showed that the titanic oxyd thus dissolved presents chemical reactions very different from that got from menaccanite by the same solvent, or from titanite itself by the action of hot sulphuric acid, and then described it as a peculiar modification of titanic acid.* § 51. We recognize in the fourtli place among the pro- tosilicates a group characterized by a hardness less than . * Araer. Jour. Science, 1852, xiv., 346. I W t ?' , iH li i.U ill 316 ^ NATUBAL SYSTEM IN MINEKALOGY. [VIII. bib -^ ^^" 1 J tHatof t.e t,.ee J^^S^ltX^^^ basal cleavage, y"'''!''* ™X foliated sevpentiueB H. ^ ^ell seen i.i talc, and "> the i j,^,jer, is largely lyve, l.nt sparingly « '-'" " ^^^^^ eWorites ot the second, fevelopcd in t^>e '""';'« J^j'f, „,;u,.aUy designated as Stay Ue ^^X^C'^^'^^ ff^ in this ordev a «'"«>'l<='*'; C by much that is called sei- chiefly ™»g"^«>»\.?P'!!!"ute chrysocolla, etc. To these pentincbydeweyUe, ee.ohte c^^y^^^ ^^^^^^^^ ^ the tribal dos'gna""" f P They are, for the G,eek name for ^"Te"*'"';;^ L acids witlrout peotasa- rnost part, readdy d^c" "P^W substances. The cryst"^" tion, and are amorphous collma jj^;^, i„ physical line silicates which apP>^oach thes^ ^^^ .^ ^^ and chemical eh»acters--m pat p^y^^^ .^ ^^^ j^^^,^,, nerhaps, spathoids,— will oe discussion of the ophitoids. ,„,,^,der, that of the R 52. Piissing now to the secona ,<,to,iUoates, P Lov.KB,ucATES, we recognue. as .^^^^ J^^^ ^^ five tribes, which ^oP^f.^^^/Jf ti,e great family of the •ust noticed. The fi.^t«*'"^^\ Jed species, wh.ch Lolites, with f o^-"*; ^"/^„ ;:thoid tribe, convemently together constitute a ''jdre^P* resemblances, as regards designated as ZeoUt^.d^- J^^J^\, „e,„,,ence. between . hardness, density, aspect, MIC m pectolitoids are this tribe of hydrous "/^j^e difference in chemi- .uch that, -*-*tTi?eou: pectolitoids have generally ^3 ::2dtith - — ,„,er. constituting the "'The spathoids of the second ^^ ° , ,,,ge number of tribe of the •?"«»?/ '''"''itiMe, gehlenite, ilvaite, the it VIII,] A CLASSIFICATION OP SILICATES. 317 I s is d, as to- ins tes, ser- lese tlie • the itisa- ystal- ysical , part, iirther of the Ucates, those of the which nieixtly regivi'ds etweeTi Olds are chemi- enerally iting tlie [umber of laite, the leucite, lecies are, ^^,e Proto- peradamantoids form a large and important tribe of hard and gem-like species, inchiding pargasite, glaucophane, gastaldite, idocrase, garnet, beryl, euclase, ardennite, axinite, epidote, spodumene, sapphirine, staurolite, and the tourmalines; besides allanite, the titanic species, keilhauite and schorlomite, and the remarkable ferric species, segirite, acmite, and arfvedsonite. These, though differing in this regard among themselves, have all of them a more condensed molecule than the densest of the spathoids, and it is to be noted that their resistance to acids is correspondingly greater. The highly basic ada- mantoids of this sub-order, such as garnet, epidote, and zoisite, are not attacked, while the basic spathoids (as scapolites and feldspars) are readily decomposed, by acids. The phylloid type in this sub-order is represented by the great group of the micas and chlorites, constituting the tribe of Protoperphylloids and including a large number of species, both hydrous and anhydrous, which are more condensed than the spathoids, though less so than the adamantoids. In the fifth place, we find the uncrystalline colloidal species of this sub-order represented by the tribe of the Pmitoids, named for the typical species, pinite, and corre- sponding to the ophitoids, with which they have man;" analogies. This tribe includes several species which are essentially hydrous silicates of alumina, with more or less alkali. With the true pinitoids are probably confounded other substances which are compact forms of the corre- sponding phylloids. The h3^drous silicates palagonite and pitchstone, and the anhydrous tachylite and obsidian, though not definite mineral species, are placed in this tribe, as being colloidal protopersilicates. § 53. The hydrospathoid and spathoid tribes are scarcely represented among the less protobasic silicates of the second sub-order, and, with the exception of westanite, which seems to be a Pcrzeolltoid^ and the bismuthio silicates, eulytite, agricolite, and bismutoferrite, apparently !• 318 A NATURAL SYSTEM IN MINERALOGY. [VIII. 'J' . I" ' Perspathoids, are as yet unrecognized in the sub-order of the Persilicates. The Peradamantoids, however, con- stitute an important tribe, including andalusite, topaz, dumortierite, fibrolite, xenolite, cyanite, and the zircons. The Perphylloid tribe is rep'-'^sented by a few micaceous species, such as pholerite, talcosite, kaolinite, and pyro- phyllite ; while the uncrystalline or colloid type in this sub-order, which we have designated the Argilloid tribe, includes the various clays or amorphous hydrous silicates of alumina, from schrotterite through allophane .aid halloy- site to cimolite and smectite, together with wolchonskoite and chloropal. § 54. In the preceding scheme it will be seen that the first place has been given to the great chemical distinctions which are embodied in the three sub-orders of silicates. It might be thought that the well marked physical types which we have seen recurring in the different sub-orders should be made the ground of a first subdivision of the order Silicate, rather than the chemical distinctions here adopted. These resemblances, dependent upon similar molecular aggregations, and upon physical structure, are, however, less fundamental than those based upon elemental consti- tution. These, as we have sought to show, are genetic, and should, therefore, have assigned to them a greater signifi- cance than the analogies based on similarity of aggrega- tion and structure, which, although of much importance in chissifica^ion, are essentially mimetic. The foundations alike of the order and the sub-orders are wholly chemical, and the division of each of the sub-orders into tribes is primarily and essentially chemical and genetic. On the other hand, the remarkable resemblances between the cor- responding tribes in the different sub-orders, which are chemically distinct, is imitative or mimetic, and should, therefore, be assigned a subordinate rank in classification. § 55. In arranging still farther the different families, genera, and species in each tribe, the question arises, what kind of chemical variation should take precedence. Con- 1i. VIII. A CLASSIFICATION OF SILICATES. 319 a- IS. us ro- his ibe, ites Aoy- Loite sidering the general persistence of type in series of proto- persilicates like those of the zeolites and the feldspars, in each of which the ratio of protoxyds to alumina is con- stant, that of the silica being variable, I have, in a tabular view of the sub-order, arranged species so related on the same horizontal lines; while species belonging to the same tribe, but having different relations between the protoxyds and the alumina, are arranged in successive horizontal lines ; those with the larger proportion of protoxyds being above, and those with the smaller proportion below, so as to represent the passage towards protosilicates, in the one direction, and to persilicates, in the other. It should here be remarked that in many cases, as in tourmalines and in micas, the species thus vertically arranged present physi- cal resemblances not less close than those between species on the same horizontal line, within the tribe, as may be seen in the synoptical table of the protopersilicates, mentioned below. As regards the relative condensation, the suc- cessive species or genera of a tribe on a given line may be placed with regard to the value of V, — the denser, or those with the lesser atomic volume, following those which are less dense. § 56. For the better understanding of the formulas given in the accompanying tables of the various tribes of silicates, it may be well to recall the values of the chemi- cal symbols here employed, which are atomic, — the small letters representing atoms of the elements. Hence, while for univalent elements, or monads, like sodium, chlorine, and fluorine, the symbols represent the received molecu- lar weights, these weights for dyads, like glucinum, cal- cium, ferrosum, oxygen, and sulphur, are divided by two ; for triads, like boron, aluminium, chromicum, ferricum, manganicum, and bismuth, by three ; for the tetrads, sili- con, titanium, zirconium, and thorium, by four ; and for a pentad, like niobium, by five. Thus the numerical values of the symbols here used, hydrogen being unity, are as follows : — A NATUKAL SYSTEM IN MINEKALOGY. IVIII. Atomic Symbols and Weights. . 9.00 . 8.00 . 10.00 . 19.00 . 35.50 . 7.00 . 23.00 . 39.00 cs . gl . nig ca . sr . ba . fe . mn 3.3.00 cu . . 31.05 4.50 ni . . 20.00 12.00 zn . . 32.50 20.00 ce . . 47.00 43.75 yt • . 44.50 68.50 b . . 3.66 28.00 al . . 9.00 27.50 cri . 17.33 fl . inni bi . si . ti . zr . th . nb . 18.66 18.50 69.33 7.00 12.50 22.50 58.00 18.80 § 57. The sub-orders and tribes of tlie order Silicate, as already set forth, are here presented, and are followed by a list of the principal species in each tribe. The several minerals of the various tribes, in their sequence, will then be briefly noticed, and tables of them will be given, show- ing the atomic formulas of the species, and the values of P and V as calculated therefrom. For the crystalline tribes, the form, when known, will be designated in these tr,bles under X, initial letters being used, as follows : I, Isometric ; T, Tetragonal ; O, Orthorhombic ; C, Clino- rhombic ; A, Anorthic or Triclinic ; H, Hexagonal, and R, llhombohedral. This will be followed, in a fourth part of the essay, by a brief discussion of the non-silicated oxyds or Oxydates, and the non-oxydized metallic ores or Metallates, together constituting two additional orders, the places of which are then assigned in a general scheme of classification that includes all native mineral species. Following this, is a discussion of the question of molecular weights, and its bearing on a new departure in chemistry. Finally, the principal minerals of each sub-order, arranged under their respective tribes, in the sequence already explained, will It ate, as vedby several ill then ^, sbow- lalues of stalline ^1 these o\vs". ^-1 ,, CUno- nal, and VIII.] A CLASSIFICATION OF SILICATES. 821 be presented in synoptical tables, giving at a single view the new classification of the silicates.* Order SiLtrATE. StTB-OBDEB I. PkoTOSILICATB. Tribe 1. Hydroprotospathoid (Pectolitoid). Tribe 2. Protospathoid. Tribe 3. Protadaniantoid. , Tribe 4. Protophylloid. Tribe 5. Protocolloid (Ophitoid). Sub-Order II. Protopersilicatb. Tribe 6. Hydroprotoperspathoid (Zeolitoid). Tribe 7. Protoperspathoid. Tribe 8. Protoperadainantoid Tribe 9. Protoperphylloid. Tribe 10. Protopercolloid (Pit toid). SUB-OrDEB III. PiJBSILICATE. Tribe 11. Hydroperspathoid (Perzeolitoid). Tribe 12. Perspathoid. Tribe 13. Peradaraantoid. Tribe 14. Perphylloid. Tribe 15. Percolloid (Argilloid). Tribe 1. Pectolitoid. Calamine, Thorite, Cerite, Gyrolite, Friede- lite, Pyrosmalite, Xonaltite, Plombierite, Dioptase, Pectolite, Datolite, Apopliyllite, Okenite ; together with Villarsite, Matrieite, Picrosmine, Picrolite, and Chrysotile. (Table I. ) Tribe 2. Protospathoid. Danalite, Willemite, Batrachite, Tephro- ite, Knebelite, Gadolinite, Helvite, Leueophanite, Wollastonite, Tscheffkinite. (Table II. ) Tribe 3. Protadamantoid. Chondrodite, Monticellite, Chrysolite, Phenacite, Bertrandite, Amphibole, Rhodonite, Pyroxene, Enstatite, Guarinite, Titanite, Danburite. ( Table III. ) Tribe 4. Protophylloid. Thermophyllite, Marmolite, Talc. (Ta- ble IV.) Tribe 5. Ophitoid. Serpentine, Retinalite, Deweylite, Genthite, Aphrodite, Cerolite, Chrysocolla, Spadaite, Rensselaerite, Sepiolite, Glauconite. ( Table V. ) * As regards the designation of the tribes, the use of a term which ends in a syllable expressing likeness, to include not only bodies resem- bling a given type, but the type itSelf, is justified by the meaning given to such words as haloid, albuminoid, and colloid, and also by the use in botany of the name of Aroideos for an order which comprises not only Araceae, but the typical genus Arum. 1 1 T m y iSwi' i fiSI k ^^^^V 1 1*" 822 A NATURAL SYSTEM IN MINERALOGY. tvin. . I ■'. Tkibe 6. Zeolitoid. Xanthorthite, Hamelite, Catapleilte, the various Zeolites; wltli Cancrlnite ai i Ittnerite, Edlngtonite, Sloanite, Forestlte. ( Table VL) Tbide 7. PnoTOPEKSPATHOiD. Melilitc, Eudialyte, Wohlerite, Hum- boldtllite, Ilvalte, Gehlenite, Sarcolite, Mllarite, Barylite, Meionite, with Marialite and intermediate Scapolites, Sodalite, Nosite, Hauyne, Lapis-lazuli, Leuclte, Hyalophane, Orthoclase, Microcllne, Neplielite, Paranthite, Eucryptite, Anorthite, Albite and intermediate Feld- spars, lolite, Petailte. (Table VII. ) Tribe 8. Protopebadamantoid. Pargasitc, Eeilhauite, Idocrase, Glaucophane, Schorlomite, Garnet, -.Egirite, Allanite, Beryl, Euclase, Prehnile, A-rfvedsonite, Ardennite, Axinite, Epidote, Zoislte, Jadeite, Gastaldite, Acmite, Spodumene, Sapphirine, Staurolite; and the Tourmalines, including Coronite, Schorlite, Aphrizite, Indieolite, Rubellite. (Tar'eVIIL) Tbibe 9. Pbotopebphylloid. Astrophyllite, Phlogopite, Pyroscle- rite, Penninite, Ripidolite, Prochlorite, Cronstedite, Leuchtenbergite, Venerite, Corundophilite, Biotite, Voigtlte, Cryophylllte, Seybertite, Thuringite, Jefferisite, Annite, Willcoxite, Chlorltoid, Lepidomelane, Zinnwaldite, Oellaclierite, Lepidollte, Margarlte, Euphyllite, Cooke- ite, Damourite, ParagOiilte, Muscovite. (Table IX.) Tbibe 10. Pinitoid. Jollyte, Fahlunite, Esmarkite, Bravaisite, Sorda- valite, Hygrophllite, Pinite, Cossaite; with Palagonite, Tachylite, Pitchstone, and Obsidian. ( Table X. ) Tbibe 11. Perzeolitoid. No species known except perhaps Westanite. Tbibe 12. Perspathoid. No species known to represent this tribe except Eulytite and other related bismuthic silicates. Tribe 13. PEBADAMANTOin. Dumortierite, Topaz, Andalusite, Fibro- lite, Cyanite, Bucholzite, Xenolite, Worthite, Lyncurite, Malacone, Zircon, Auerbachite, Anthosiderite. ( Table XI. ) Tbibe 14. Perpuylloid. Pholerite, Talcosite, Kaolinite, Pyiophyllite. (Table Xn.) Tbijb 15. Abgilloid. Schrotterite, Collyrite, Allophane, Samoite, Halloysite, Kaolin, Keramite, Hisingerite, Wolchonskoite, Montmor- lUonite, Chloropal, Cimolite, Smectite. (Table XIII.) Tribe 1. Pectolitoid. § 58. We notice first in this tribe the hydrated sili- cates of lime, often with alkali, most of which are fre- quently found among the secretions of basic rocks, and which include pectolite, xonaltite, gyrolite, plombierite, datolite, okenite, and apophyllite. The name selected for the tribe recalls at the same time the most common of these species, and also the property which belongs to most of them of pectising or being decomposed by strong VIIIO A CLASSIFICATION OP SILICATES. 323 us te, im- ilte, yne, jUte, acids, such as chlorhydric, with the separation of gelati- nous silica.* It has also the advantage of distinguishing them from the zeolitoids, the corresponding type in the next sub-order, with which they are generally associated, and sometimes confounded. Differing considerably in the proportion of combined water, the pectolitoids have a irase, iclase, idelte, d t^e . IcoUte, yrosc\e- ybertite, , Cooke- ,e, Sorda- lacbyWe, 'estanite. this tribe Malacone, Satoolte, [' Montmor- Table I. — Pbctolitoids. Species. Formula. P D V .! Calamine . . . (zn,8i,)o, -f- Jaq 24.00 3.60 6.87 0. Thorite . . (th,si,)Oj-(-Jaq 32.62 5.30 6.16 I. Cerite . . . (ceiSi,)oj-f-Jaq 29.80 4.90 6.08 P Gyrolite . . . (ca,8i3)06-|-laq 18.33 .... .... P Friedelite . . (mnj8i3)06 -\- 2aq 19.14 3.07 6.23 R. Pyrosmalite (fea8i3)0s + laq 21.68 3.17 6.80 H. Chrysotile . (mg3si4)0: + 2aq 15.33 2.22 6.98 P Xonaltite (caisi,)os-f-{aq 18.53 2.71 6.83 p Flombierite (caisij)03 + 2aq 15.20 • • • • • • • • P Dioptase . . (cnisij)08 + laq 19.67 3.34 6.88 R. Pectolite . . (cassiij)iK + laq 18.27 2.78 6.57 C. Datolite . . (cas8i4b3)oo +laq 16.00 2.99 6.35 C. Apophyllite . (ca,8i4)Os + 2aq 16.14 2.35 6.44 T. Okenite . . , (c^si4)04 + 2aq 15.14 2.35 6.44 0. hardness below that of quartz, and, with but few excep- tions, a comparatively large atomic volume. In the case of apophyllite a little fluorine is present, and in datolite a large amount of boric oxyd, which in our atomic formula * The name pectolite is said to be from the Latin pecten, in alhision to the comb-like structure of some varieties of the mineral, but it at the same time suggests the Greek thjxtos (curdled or congealed), from which have been derived the chemical terras pectose and pectin, and the verb to pectise, employed by Graham to denote the gelatinizing property of certain substances. Ik 324 A NATURAL SYSTEM IN MINERALOGY. [VIII. is represented as replacing a portion of silica. They are all native species, some of which 'lave also been artificially formed, and at least one of them, apophyliite, is found of recent origin in the channels of the thermal waters of riombi^rcs, in France, where another species, plombierite, his also been met with. An unnamed pectolitoid was got by Daubr(;e as a pioduct of the action of super- heated water on glass. Delonging to this same tribe are : the zinc-silicate, calamine ; the rare species, thorite and cerite ; the manganeslan silicate, friedelite ; pyrosmalite, a ferro-manganesian species containing ciilorino ; and the copper-silicate, dioptuse. The composition of tritomite is not certain, but appi caches that of cerite. Here, also, mosandrite probably belongs. § 69. We place here also chrysotile, which constitutes the common amianthus, and has hitherto been regarded as a variety of serpentine, with which it agrees in centesi- mal composition. It is, however, distinguished therefrom by a lower specific gravity, and by its fibrous character, which, like that of ar.uanthoid amphibole, indicates a pris- matic crystallization. As will be shown farther on, at least two other species, one phylloid, and another ophitoid, have been confounded under the name of serpentine. While the density of these last is 2.60, or higher, that of chrysotile, according to three determinations, is 2.142, 2.220, and 2.238, the first and the last of these being by E. S. Smith, and according to his analyses, corresponding to specimens containing respectively 2.23 and 3.36 of ferrous oxyd.* If this oxyd be to the magnesia as 1 : 30, it would give for P a value of 16.51, which, with a density of 2.22, would make V=6.98. Fibrous silicrles having the same centesimal composition as the last are, however, met with, having a much higher specific gravity. A well defined mineral, described many years since by the writer, from Bolton (Quebec), under the name of picrolite, is separable into long, rigid, elastic * Amer. Jour. Science, 1885, xxix., 32. !♦ LI. re o£ o£ •ite, iper- are : and lalite, d the nite is , also, ititutes jgarded centeai- leveirom lavacter, ;s a pris- sr on, at )phitoid, jrpentine. L that oi is 2.142, being by ponding to oi ierrous ) itwonld ty of 2.22, imposition iclx bigl^er tibed many ,^^ndertbe Igid, elastic vm.i A CLASSIFICATION OP SILICATES. 325 fibres, and lias, with a specific gravity of 2.G07, the com- position, silica 43.70, iimguesia 40.68, forroua oxyd 3.51, with traces of oxyds of nickel and chromium, and 12.45 of water, = 100.34.* § 60. While the above species of unlike density agree in having the serpeiitine ratio, 3:4:2, there are several other hydrous silicates of magnesia which present other ratios, and should, like these, be included among hydro- spathoids. Such are the orthorhombic sparry villarsite, with D = 2.98, which has been described as a hydrous chrysolite, and is represented by the atomic formula (mgi8ii)o3-}-^aq; and the fibrous crystalline matricite, with I) = 2.53, more hydrous in composition, with the formula (mgisix)o2-l-laq, nearly. The sparry orthorhombic picros- mine, with D = 2.66, which is sometimes fibrous and asbestiform, is a hydrous bisilicate, represented by (mgisi2)o3-|-^aq, and Terrell has very recently described as chrysotile, from an unnamed locality in Canada, with D= 2.56, an asbestiform silicate, which is at once more basic, more hydrous, and heavier than ordinary chrysotile, and approaches matricite in composition. His analysis gives silica 37.10, magnesia 39.94, ferrous oxyd 5.73, alu- mina, traces, water 10.85=: 99.62. This corresponds very closely to (mg(5.5feo.5si8)oi5-f-6aq.t These various prismatic hydrous silicates of magnesia, including chrysotile and picrolite, constitute an important group of what may be designated as magnesian pectolitoids, which have for the most part an atomic volume approaching to dioptase and to datolite, and demand farther study, but, with the excep- tion of chrysotile, have not been placed in our table. § 61. In the accompanying table (No. I.) of the prin- cipal pectolitoids, are given their atomic formulas as deduced from chemical analysis, the unit-weight, P, cal- culated from these, the density, D (water = 1.00), and the atomic volume, V, which = P -;- D. ^n. calculating the * Geology of Canada, ISfiS, p. 472. + Coiupte lleiulu Ue "Acad, des Sciences, January 26, 1885. i l« A NATURAI. SYSTEM IN MINER ALOOY. tviii. value of P for these silicates, we have to consider that two or more protoxyd-bases are often present, and that the proportions of these must be estimated as nearly as possible. As the specific gravity of species is in many cases inexactly determined, we have, where more than one value of D is given by mineralogists, selected tliat which seemed most probably correct, and, where determi- nations of density are wanting, have left a blank in the table. Of the species in this table, datolite is a borosili- cate, pyrosmalite a chlorosilicate, and apophyllite a fluoro- silicate. § G2. It has been thouglit well, for reasons which will be apparent when we compare the pectolitoids with other tribes, to represent their contained water by the symbol aq, preceded by the sign -}-• It will be noted that in the atomic formulas here employed, the symbols of the metals, with those of silicon, boron, and titanium, afe placed within parentheses, and those of oxygen, sulphur, fluorine, and chlorine, together with water, without. From this it will be clear that the atomic weight deduced from these formulas must, in order to arrive at P (the weight of the atomic unit), be divided by the number of these units; that is to say, by the sum of the coefficients of the ele- ments outside of the parenthesis. The present table is far fiom complete ; the determinations of density are in many cases uncertain, those assigned to the same species by dif- ferent observers often presenting wide variations. Again, the value of P in cerite is calculated as if it were simply a silicate of cerium, while it contains unknown proportions of lanthanum, didymium, and samarium. In calculating the value of P for pyrosmalite, it is regarded as a ferrous silicate in which p^ is replaced by cl|, equal to 3.46 of chlorine. The general agreement in the value of V is noticeable, save in two cases, — that of dioptase, for which another recorded determination of D = 3.28 gives V= 6.00, and that of datolite, whose volume shows a condensation approaching to that of the adamantoid protosilicates. w •(A VIII. A CLASSIFICATION OF SILICATES. 827 .t it an i\at lui- t\ie .»iU- Loro- i wiU other in tVie netalSi placed uori^e, this* it ft these of the units; [the ele- Ae is far tn niany Is by dif- ^gain, isimp^y '^ kportions [iculating la iervous 3.46 of p of V is [for which V = 6.00, .densatiou jates. Tribe 2. Protospathoid. § 68. In the second tribe, whicli wo have called Proto- spathoids, shown in tuhlo No. II., are tlie sparry silicates of zinc and manganese, willoniite and tephroite, and the ferro-niangane«uin species, knebelite, — all having the ratio of unisilicates. To these are joined tlie iloiihlo silicate of lime and magnesia, batrachite, with gadolinite, a silicate TaBLK II. — PuOTOSPAXnOIDS.* Species. Form i; LA. Danalite. - - Willemite. - Batrachite. - Tephroite. - - Knebelite. - - Gadolinite. - Helvite. - - Leucophanite. Wollaatonite. - Tscheffkinite. (m,8ig)o„ - (m = gl, fe, an) (zn,8ii)o, (m,8i,)o, - (m = cao.smgo.5) (innisi,)0j - - - - - (m,8ii)0j - (m =feo.5mno.5) (mi8i,)o, - (m =gK yt, fe) (mi8ii)0j - (m == gl, mn) - (mjsiiioT - (m = gl, ca, na) (ca,8ij)o, (m,sij)os - (m s= ce, ca, fe) 22.15 27.75 19.50 25.25 25.37 25.60 20.16 18.05 10.33 27.00 D 8.43 4.18 3.03 4.12 4.12 4.20 3.30 2.97 2.92 4.26 0.76 0.63 6.43 6.13 6.15 «.10 6.r 6.07 6.62 6.34 I. n. o. o. o. 1. o. c. ? chiefly of yttria, giving apparently the same atomic ratios, and helvite, a silicate of glucina and manganese, remark- able for containing a large amount of sulphur ; in which respect it resembles the more basic silicate of glucina, iron and zinc, danalite, belonging to the same tribe. With these are also placed leucophanite, which is interest- ing as being a fluoriferous silicate of glucina, lime, and * The formulas employed in calculating the values of P and V for the following species are a. Danalite — (gl3.ofe3.onino.5zni,68i8 ,,)oi9.oSi.o. 6. Gadolinite — (gl».noyt2.oofeo.T6ceo.26si4.oo)oB.w . c. Leucophanite — (ca5.ogl5.onaa,o.«:ij.o)o»8.o' d. Tscheffkinite — [Damour] (ce9.i»feo,»(,ca^.,6si,.aotio.7o)Qi.oo. 328 A NATURAL SYSTEM IN MINERALOGY. [VIII. soda, having the same atomic ratio for its bases as serpen- tine and chrysotile. Among bisilicates we find, in this tribe, wollastonite, a simple lirae-silicate, and tscheffkinite, a titanosilicate of lime with cerous and ferrous oxyds. All of the silicates of this tribe are decomposed by acids with pectisation. Tribe 3. Protadamantoids. § 64. We next proceed to note the adamantoid proto- silicates or Protadamantoids, closely connected with the Protospathoids, but distinguished from them by a more condensed molecule and a greater resistance to acids. First in order comes the fluoriferous magnesian silicate, chondrodite, next the double silicate of lime and magnesia, monticellite (which, from its recorded specific gravity, would seem to be a denser silicate, isomeric with the spath- oid batrachite), and the chrysolites, belonging to a more condensed type than either. The genus chrysolite in- cludes not only the ordinary more or less ferrous species, but forsterite, on the one hand, and hortonolite and fayalite, on the other. To this succeed the two glucinic species, phenacite and bertrandite, the former of which is the most highly condensed protadamantoid known, while the latter is remarkable for containing a portion of water. Next in order comes the manganesian species, rhodonite, together with the amphiboles and pyroxenes, two important genera, or rather families, which (with the apparent exception of certain amphiboles having the atomic ratio of bases to silica of 4 : 9) are bisilicates. While rhodonite and pyrox- ene are clinorhombic in crystallization, the magnesian species enstatite, with hypersthene and diaclasite, is or- thorhombic. Anthophyllite appears to be an ortho- rhombic species having the composition of amphibole, and kupfferite, a magnesian amphibole. Their very varied composition, and the great number of bases which enter into the composition of some of the amphiboles and the pyroxenes, are illustrations of the polybasic character of .. ^ I- is >e, Is. ds VIII.] A CLASSIFICATION OF SILICATES. 329 the silicates. With the pyroxenes, some mineralogists have grouped spodumene, gegirite, arfvetlsonite, and acmite, the association being based on similarity of crystalline form, and supported by a misconception of their chemical relations. All of these species find their position in the '' '"ill ['•■:re Table III. — Pkotauamantoids. Lte in- .e Spbcihb. Formula. F D V X. Chondrodite. - - (mg^Bia)©; 18.64 3.20 5.82 0. Monticellite. - - (misil)©., - (m = mgo-jcao-s) - - 19.50 3.25 6.00 0. Forsterite. - « (mgisijoj 17.50 3.30 5.30 0. Chrysolite (1). - (miBii)Oi, - (m = mgo.»feo.i - - 18.30 3.40 5.38 0. Chrysolite (2). - (misi J02 - (m = mgo-gfeo-j - - 19.10 3.50 5.45 0. Bertrandite. - - (gli8i:)02+Jaq 13.22 2.59 5.10 0. Phenacite. - - (g^lSJiK 15.75 3.00 4.58 R. Amphibole (1). - (mi8ij)03 - (m = mgo.75,cao.25) - 17.33 2.97 5.88 C. Amphibole (2). - (misij)03 - (m=mgo.6cao.3feo.i)- 18.00 3.00 5.88 C. Rhodonite. - - (mnisi2)o, 21.83 3.00 6.06 C. Pyroxene (1) - (misij)03 - (m = cao-smgo-s) - - 18.00 3.27 5.50 C. Pyroxene (2). - (misi2)03 - (m = cao.5mgo.5) - - 18.00 3.28 5.48 C. Pyroxene (3). - (m,8i2)0s - (m = cajmgs) - - 17.55 3.22 5.45 C. Pyroxene (4). - (mi8i2)03 - (m = cajmgffe J) - 18.66 3.41 5.47 c. Enstatite (1). - (mi8i2)o3 - (m = mgo-gfeo-i) - - 17.20 3.10: 5.54 1 0. Enstatite (2). - (m,si2)03 - (m = mgo.gfco.j'' - - 17.73 3.25 5.45 0. Titanite. - - - (ca,8iiti2)05 19.80 3.50 5.65 c. Guarinite. - - - (oaisi2ti2)o5 19.80 3.50 5.65 1 T. Danburite. - - (cai8i4b3)08 15.37 3.00| 5.12 0. next sub-order, and the place of the last three is near to garnet and to epidote. We have already shown, in § 42, how, on similar grounds, the aluminous species pargasite, glaucophane, and gastaldite have been erroneously placed with amphibole. tammm 330 A NATURAL SYSTEM IN MINERALOGY. [VUL *; .* ' (i!;!: § 65. The relations of amphibole and pyroxene to each other and to wollastonite, as shown in the unlike degrees of condensation made evident by the different values of V, were pointed out by the present writer, in 1853, as ex- amples of isome-ism in polysilicates, when the three were represented as belonging to as many homologous types (§ 18). These relations, so far as amphibole and pyrox- ene are concerned, were mentioned some years later by Dana, in 1868, when he noticed that the pyroxenes have a spccitic gravity about one tenth greater than that of the corresponding amphiboles.* The chemical difference between these species and the corresponding spathoids is seen in the resistance of both amphibole and pyroxene to acids, which decompose wollastonite. Rhodonite, a manganesian species with the crystalline form of pyroxene, appears, from its volume, to be more closely related to amphibole, and is partly decomposed by acids. Different and unlike varieties of pyroxene agree closely with each other, with enstatite, and with chryso- lite, in the value of V, as will be made evident by ihe accompanying table. No. III. In this, the four pyroxenes compared were examined and analyzed by the writer. To this tribe of Protadamantoids we add titanite and guarinite, two titanosilicates of unlike crystalline form, but of identical composition and specific gravity. The solubility of titanitt in acids has already been noticed, in § 57. Here, also, is the place of danburite, a borosilicate, remarkable for having a value of V near to that of the pec- tolitoid borosilicate, datolite. The amphiboles, rhodonite, chondrodite, and monticellite, are the adamantoids which approach nearest to the spathoids, from the denser species of which, tephroite, hel . ite, and leucophanite, they are not far removed in volume. * System of Mineralogy, 5th ed., p. 240. ;• I. lb es ot JX- ere pes •ox- ve a the ence ds is )xene onite, m of jlosely ed by 5 agree chryso- ent by e iour by ^'be Lite and [e form, y. The Iticed, in )silicate, tbe pec- lodonite, Is wbicb jsr species tbey are VIU.] A CLASSIFICATION OF SILICATES. 331 Tribe 4. Protophijlloids. § 66. The phylloid type in the protosilicates is repre- sented by a small number of magnesian minerals, of which the best known is talc, apparently including two species with different atomic formulas, but indistinguishable save by chemical analysis. To these must be added one or more of the species generally classed under the head of serpentine. Among them is thermophyllite, having a recorded density of 2.56-2.61, while marmolite, with a simi- lar composition, should, if its density be really 2.41, con- stitute another Protophylloid species, as indicated in Table IV. — PROTOPHYLLOios. Species. FOBMULA. P D V X. Thennophyllite. Marmolite. - - Talc. .... Tala - - - - (mgjsii)©, + 2aq - (mg3si>j + 2aq - (mg,8iio)Oj4+laq - (mgjSijK + f aq - 15-33 15-33 16-93 15-82 2-61 2-41 2-70 2-60 5-87 6-35 5-90 6-07 ? ? 0. o. Table IV. From the structure of these minerals, Dana has suggested that serpentine may be micaceous in crys- tallization, like talc and chlorite.* This is so far true that a silicate having the centesimal composition of serpentine assumes a phyllcid type, as seen in thermophyllite and in marmolite ; but it also takes on a prismatic fibrous type in chrysotile and in picrolite, silicates of unlike density, already mentioned in § 65, and is, moreover, found as an amorphous colloid species included in the next tribe, that of the Ophitoids, of which it may be regarded as the type. * s System of Miueralogy, 5th ed., p. 465. . jflS A, '.ill a u ■■: V m. ''y. ''its mSSSSSBk\ I I 832 A NATU' ^Ji SYSTEM IN MINERALOGY. [VIIL '* Trife 5. OpJiitoids. § 67. Irx considering this tribe, we begin by noting certain differences in composition and in specific gravity among the maguesian silicates, wnich (besides thermo- phyllite, marmolite, picrolite, and chrysotile) have hitherto been grouped under the name of serpentine. A density of from 2.60 to 2.70 is generally assigned to this silicate, but bowenite, according to the analysis of J. Lawrence Table Y. — Ophitoids. Species. Formula. P D V Serpentine. - - (mg.,si4)07 + 2aq 15-33 2-65 5-78 Eetinalite. - - (mgaSiOo; + 2Jaq 15-00 2-40 6-2.5 Deweylite. - - (mgjSislOj + Saq 14-00 2-25 6-22 Genthite. - - - (nij8i3)05 + 3aq 18-25 2-40 7-60 Aphrodite. - - (mgisi2)03 + ^aq 15-13 2-21 6-84 Cerolite. - - - (mgisi2)03 + ljaq 14-11 2-30 6-13 ChrysocoUa. - - (cuisi2)03 + 2aq 17-53 2-24 7-82 Spadaite. - - - mg58i,2)o„ + 4aq 15.04 ? • a • • Eenaselaerite. - (mg4siio)oi4 + laq 15-93 2-70 5-90 Sepiolite. - - - (mgisi3)04 + laq 14-80 ? • • > • Glauconite. - - t • • • o » m • • ' • • Smith and Brush, is a nearly pure serpentine, with a density of 2.69 to 2.78, and a hardness of 5.5 to 6.0. Rt'i- nalite, a clearly marked ophitoid or amorphous species, which includes much of the serpentine of the Laurentian limestones, is a very pure magnesian silicate, distinguished from ordinary serpentine by its lower density, and its larger proportion of water, which, from several analyses, the writer found to be over fifteen hundredths. The spe- cific gravity of retinalite is 2.^6 oo 2.38, or nearly that assigned to the phylloid species marmolite. The name of lao- vto sity ate, jnce f8' i5 22 60' 84 13 82 • • • •90 witli a lO. Be'i- VIII.] A CLASSIFICATION OF 3ILICATES. 833 serpentine may, perhaps, be retained for the amorphous silicate with density 2.6 to 2.7, which must be dis- • tinguished from retinalite, as well as from chrysotile, from picrolite, from thermophyllite, and from marmolite. This last requires farther study, as does, likewise, bowe- nite, which merits particular notice from its superior density and hardness, and requires optical examination. § 68. Following serpentine and retinalite in Table V. are deweylite an J. genthite, — the latter a niccoliferous ophitoid, as chrysocolla is a cupric one. With the latter are placed the bisilicates aphrodite and cerolite, which last appears to have the volume of retinalite and of dew- eylite. After these we have placed spaJaito. as also rensselaerite or pyrallolite (which is, perhaps, a compact phylloid rather than an ophitoid), and sepiolite. Along- side of this, a position has been conjecturally assigned to glauconite as not improbably a ferrous potassic ophitoid, of which a large part of the iron has subsequently passed into the ferric condition. ^See, for a discussion of its composition, pages 196-198.) § 69. The significance of this tribe of amorphous hydrous silicates in mineralogy will be more apparent when wo come to study the corresponding tribes among the othcx two sub-orders of silicates, and among the non- silicated oxyds. In each of these we find a group of com- pounds which, although, in parallel tribes, occasionally assuming crystalline form, require for their crystallization conditions not always present. The particular silicate of magnesia which constitutes serpentine, although some- times crystallizing in hydrous forms, as in thermophyllite and chrysotile, appears incapable of forming an anhydrous species like the more and the less basic crystalline sili- cates of the same base, such as chrysolite and enstatite. Hence we often find the hydrous colloid, serpentine, still associated with the one or the other of these, into a mix- ture of which it is resolved when its dehydriitiou and fusion are effected by heat. ' if 834 A NATUKAL SYSTEM TN MINERALOGY. [vm. Tribe 6. ZeoUtoids. § 70. The sixth tribe, being the first in the sub-order . of the Protopersilicates, has been designated Zeolitoid for the reason that it includes, and is chiefly represented by, that large family of silicates familiarly known as zeolites, which have been aptly described as hydrated feldspars. These are double silicates of a protoxyd-base and alumina, the atomic ratio between the two being 1 : 3, and the pro- toxyds essentially lime and alkalies, occasionally with baryta and strontia, — magnesia being for the most part absent, or found only in traces. The proportion of silica varies from that of thomsonite, which gives the ratios 1 : 3 : 4, to stilbite and related species, with 1 : 3 : 12. The water is also subject to great variations, and is held with different degrees of force — some species, such as lauraontite and chabazite, parting with a portion in dry air at ordinary or slightly elevated temperatures, while others are much more stable. The intumescence before the blowpipe-flame, which is characteristic of many species of this family, and which suggested the name of " zeo- lite," would seem to indicate that a partial melting of these takes place before the complete expulsion of water, or, in other words, that the silicate fuses in its water of crystallization. The zeolites are attacked by acids, gener- ally with pectisation, and are but little condensed, having high values for V. We have given, in the accompanying table (No. VI.), some of the more important species of this large family. Pollucite, from the analysis of JKam- melsberg, is a zeolite, in which two-thirds of the protoxyd- base is oxyd of caesium. § 71. In the same tribe of Zeolitoids we place several other hydrous silicates, which are distinguished from the zeolites by presenting different ratios between the prot- oxyd and sesquioxyd bases. The species here called hamelite was described by the writer many years since as a crystalline hydrous silicate of ferrous oxyd, magnesia. a. ler . for by, tes, lars. Ana, pro- witli part silica ratios 1 : 12. J held icli aa in dry ,, while , beiore species " zeo- Iting of water, rater oi ^, gener- having [panying rjecies oi [oi Kara- Irotoxyd- VIII.] A CLASSIFICATION OF SILICATES. Table VI. — Zkolitoids. 835 Species. Formula. P D V X Xanthorthite.- ( mialisia )o4 + 2aq - ( m=< 3e, fe) - - - • • • • 2.90 • • • c. Hamelite. - - Catapleiite. - (m,al,sis)08 + laq-( ,m- = mg,fe,na) • « • • 18.09 • • • 2.80 • • • ? TT ^mjZrgSitjOg T^ zaq - " 0.40 ii.. 1 Cancrinite. - (na,al,88i„)05,+ 3CiCaiO,+4iaq - - .... 2.42 • « * II. Thomsonite. - (mials8i4)08 + 2}aq- (m = cainai) - 13.58 2.38 6.54 O. Gismondite. - (caial38i4.5o)o8.5o + 4Jaq 14.38 2.26 6.36 o. NatroHte. - - (naialjSi8)Oio+2aq 15.83 2.25 7.03 0. Scolecite. - - (caialj8i8)o,o + 3aq - 16.08 2.40 6.28 c. Mesolite. - - (mial3Si8)Oio + 3aq - (m =» cafnaj) - 15.15 2.40 6.31 c. Levynite. - - (caial3ei8)o,o+4aq - 14.64 2.16 6.77 K. Pollucite. - - (mjalasig)©,, + laq - (m a= csSnaJ) - 21.46 2.90 7.40 I. Analcite. - - (na,als8i8)Oij + 2aq - - 15.71 2.29 6.86 I. Eudnophite. - (naial38i8)Oij + 2aq - i .71 2.27 6.92 0. Laumontite. - (caialjsig)©,, + 4aq - 14.68 2.30 6.38 C. Herschelite. - (mialsBi8)o,2 + 5aq- (m: = na|ki)- - 14.76 2.06 7.16 O. Phillipflite. - (mial3si8)Oi2 + 5aq - (m =cai|naj) - 14.41 2.20 6.55 0. Chabarite. - (caial3si8)oi2 + 6aq- - 14.05 2.19 6.41 R. Gmelinite. (mi8l38i8)oij+6aq- (in = = cajnaj) - 14.11 2.17 6.50 R. Faujasite. - - (mial3si,)Ois-h9aq- (m =rnajcaj) - 13.45 1.92 7.00 I. Hypostilbite - (caial38i9)Oi3+6aq ■ - 14.10 2.20 6.40 ? Harmotome. - (mial38iio)Oj4+5aq .(m sasba^-nai^o) 16.73 2.45 6.82 o. Epistilbite. - (caial38ii2)oi8 + 5aq - 14.47 2.25 6.43 0. Brewsterite. - (mial38ij2)Oi6+5aq . (m = BrSnaJ) - 15.27 2.45 6.23 c. Stilbite. - - (caial38i„)oi6+6aq • 14.23 2.20 6.46 0. Heulandite. - (caial3sii2)Oi8 + 5aq - 14.47 2.20 6.58 c. Edingtonite. - (baial48i7)oi2 + 4aq - 17.84 2.71 6.58 T. Sloanite. - - (caial58i7)oi3 + 3aq • 15.31 2.44 6.27 O. Forestite. - - (caial«|Bi,2)Oi9 + 6aq - 14.56 2.40 6.06 0. and soda, which is found filling the pores of a paleozoic crinoid,* while catapleiite is a zirconic zeolitoid, in which zirconia takes the place of alumina. Here also we have * Ampf. Jour. Science, 1871. i., 379; see also ante, p. 104, for details. 336 A NAT U UAL SYSTEM IN MINEIIALOGY. [VIII. placed xanthorthite, a hydrous species which, by its com- position and its low density, is widely separated from the a; lyd-ous dense adamantoid orthite, or allanite, to be > >uce(l farther on. While the species just noticed are f 'e protobasic than the zeolites, there are not wanting t.:ui, ijples of zeolitoid species less protobasic than these. Sucii ■ the curious barytic silicate, edingtonite, to which analysis assigns for protoxyds and alumina the ratio, 1:4; sloanite, zeolitic in habit, with a ratio of 1 : 6, and for- estite, a species closely resembling stilbite, to which is given the ratio, 1 : 6. The hydrous carbosilicate, can- crinite, and the sulphatosilicate, ittnerite, which properly belong to the zeolitoids, will be noticed under the follow- ing tribe, in §§ 83, 84. Tribe 7. Protoperspathoida. § 72. We Tiave next to consider the Protoperspathoids, which include, besides the feldspars and the scapolites, a number of other species of double silicates, chiefly alumi- nous. The species of this tribe are distinguished from the preceding by their higher density, superior hardness, and greater resistance to acids ; since, while the whole of the zeolites and zeolitoids are decomposed thereby, generally with pectisation, only the more basic of the protoper- spathoids are thus attacked. The feldspars, like the zeolites, have the atomic ratio between the protoxyds and alumina represented by 1 : 3, the silica in both being subject to the same variations. As in the zeolites, the protoxyd-bases are alkalies and lime, rarely with baryta, while magnesia and ferrous oxyd are but exceptionally present. Unlike the zeolites, they are anhydrous, or contain occasionally one or two hundredths of water. § 73. The feldspar family includes, first, the feldspars proper, represented by the anorthite-albite genus; sec- ondly, orthoclase, microcline, and hyalophane, near which may be placed nephelite and paranthite; and, thirdly. Hy vep vm.] i- 36 re ng ich :4; for- h. is can- pevly bhoids* lites, a alumi- om til® ss, and of tbe nerally rotoper- jic ratio Iby 1 • ^' Itiations. [lies and ferrous zeolites, or two feldspars inus; sec- lear Nvliicft thirdly^ A CLASSIFICATION OF SILICATES. Tablk VII. — Protoperspatkoids. 337 Species. Formula. P D V X Melilite. - - (oa2m,8i3)Og-(m = alifiJ) - - 20.46 3.10 6.60 T. Eudialyte. (in4zrj8ii,)o,8 - (m=nai.5cai.5fei.o) 20.80 3.00 0.70 R. Wohlerite.- - • • ■ t • 3.41 • • • C. Humboldtilito. (oa3al2si5)oio I'lTO 2.00 0.05 T. Ilvaite. - - - (m;,fli8i3)oio ■ (m = feScai) ^ . iii 3.71 0.15 O. Gehlenite. - - ';ca,.om,.o8ii.3)o3.s - (m == al« ^ ., .,.? 3.00 6.48 T. Sarcolite - - (ca,al,si2)04 18.75 2.93 0.40 T. Milarita - - (in,al,8i8)Oio - (m = cao.8ko.j) 1;-J.88 2.59 0.51 0. Barylite. - - (bajal38i,)Oia ... - - 25.75 4.0;^ 6.38 ? Meionite. - - (ca4al98i,j)025 - - - - 17.80 2.74 0.49 T. Wernerite. - (m4al98iie)029 17.41 2.70 6.44 T. Ekebergite. - (m4al9sii8)03, 17.42 2.74 6.32 T. Mizzonite. (uiialBsiaOosi 17.20 2.62 6.56 T. Dipyre. - - - (m,al98i24)037 - ------ 16.89 2.64 6.39 T. Marialite. - - (m4alaSi3e)049 16.43 2. .57 6.39 T. Sodalite. - - (na,al98i,2)024cli 19.88 2.30 8.28 I. Nosite. - - - (naial38i4)08 + Jnai8i04 - - - 20.28 2.40 8.25 ^, Hauyne. - - (naial38i4)08 + ?caiSi04 - - - 21.60 2.50 8.64 ^* Lapis lazuli. - • • • • 2.45 LeiTcite. - - (kial3si8)Oia 18.16 2.56 7.09 I, Hyalophane. - mial38i8)Oi2 - (m sbajkj) - - 19.39 2.80 6.92 C. Orthoclase. - (kial38i,2)Oia 17..37 2.54 6.83 C. Microcline. - (kial3sii2)Oi8 17.37 2.54 6.83 A. Nephelite. - - (naialjSi4.5)08.6 17.58 2.66 6.00 H. Paranthite. - {caial38i4)08 17.37 2.75 6.31 T, Eucryptite. - (Iiial38i4)08 15.75 2.67 5.93 H. Anorthite. - - (caial38i4)08 17.37 2.75 6.32 A. Barsowite; (caialasi5)09 17.11 2.73 6.27 ? Labradorite. - (mial3si8)oio - (m = cafnaj) - 16.97 2.70 6.28 A. Andesita - - (inial38i8)oj3 - (m = cajnaj) - 16.70 2.68 6.23 A. Oligoclase. (inial3si9)oi3 - (m = nafcaj) - 16.63 2.65 6.27 A. Albite. - - - (naial38ii2)Oig 16.37 2.62 6.24 A. lolite. - - - (mial38i5)09 - (m = mgjfej) - 16.81 2.67 6.29 H. Petalite. - - (Ii,al48i2o)o25 15.32 2.42 6.83 C. ^' 'it '3 if J- f : ' 1 ^rsM 338 A NATURAL SYSTEM IN MINEUALOGY. [vm. 1 n u leucite. These distinctions, as may be seen from the table No. VII. correspond to different values ,of V. lolite, a ferro-magnesian feldspathide, tliough peculiar in composition, and differing in crystallization from thp feld- 8i)ar8, agrees in volume with anorthite and albite. Eu- cryptite, which has the formula of a lithia-anorthite, seems to differ from these in possessing a more condensed mole- cule. The possibility of a more silicious feldspar than albite, corresponding to the supposed krablite of Forch- hammer, with its ratios of 1 : 3 : 24, should not be over- looked. The specific gravities of orthoclage and micro- cline show for these species a considerably greater atomic volume than for albite and its related species, a fact which was noted in 1854 by the writer as a reason for referring orthoclase to a less condensed molecule than these (§ 30). Nephelite also shows a volume near that of orthoclase, aa does the baryta-potash feldspar, hyalophane, which has the same general atomic formula as andesite ; while leu- cite, with the same atomic formula, has a still larger volume. § 74. The history of that feldspar genus which in- cludes the species anorthite and albite has been noticed at length on pages 294-296, where was discussed the view that the feldspars intermediate in composition be- tween these may be mixtures of two homoeomorphous species. The notion was there expressed that while such mixtures are, as was long since suggested, not uncommon in nature, many if not all of these intermediate feldspars are really definite species. The careful studies of the late George W. Hawes have thrown much light on this subject, by showing that in similar and apparently identi- cal rocks the feldspathic element may be represented by two associated feldspars of the same genu.s — in one case, apparently, anorthite and albite ; in another, labradorite and andesite. The diabase which along the Atlantic border of North America is found irrupted among meso- zoic strata, from Neva Scotia to North Carolina, is singu- VIU.J A ci^sair,c.ATro« o. s.ucatks. 839 wh.oh ..,« bee,, represe.t C 1 7''." '™""'« fel^ '"o spe ■ntermediate species He '„^': fT'™ "' •"'' «' »ore «qu,site balance of comnnV ' '*'' '''*■■« 'n^'ght, "An he necessary to orystalfee' su h » '""^,'''-»"^'-™ee Wo,dd par," and conceives tl we tve'T 7'*'' " ^'"g"^ ^d- focks , rarely simple arret"!':;""'?, ">">' ""•»>« -ent,t as was previously shoXby FouT,. f T'"" '"''■ • See J. D. Dana .„ , ^ ^""^ *» ''eeent urn lor J8bl, pp. J29-134. ; *( -^ I: M, I' i- 840 A NATURAL SYSTEM IN MINEUALOOY. tviir. lavas of Santoriii (ante, page 213). Meiinwhile it will remain to be decided for each individual case, whether a feldsjpathic material intermediate in composition between albite and anorthite is an integer, or an admixture of two integers, which may themselves be either the extremes of the series or integral hitermediate species. Ment'oTi should here be made of petalite, a species in many resi)ects closely related to the feldspars, but present- ing the ratios 1 : 4 : 20. As regards the proportion be- tween protoxyds and alumina, it is important as the one spathoid which corresponds v/itli the rare and less proto- basic zeolitoids ; while if krablite be rejected, petalite is the most silicious species known. Its atomic volume is identical with that of anorthite, albite, and iolite. § 75. The scapolites apparently constitute a single genus of silicates which, approaching in composition, hard- ness, and density the feldspars, were, from an early time, compared with them, so that when, in 1854, the writer attempted to generalize the notion of Von Waltershausen as to crystalline intermixtures in the intermediate feld- spars, he extended, as already shown (§ 31), a similar view to the scapolites. The ratio between the protoxyds and alumina in the scapolites has, until recently, generally been regarded as 1 : 2, and the writer, in 1863, in farther discussing the relations of the scapolites, described them as a group of which the extreme terms were meionite, with the ratios 1:2:4, and dipyre, with 1:2:6; including intermediate species which might be regarded as crystal- line admixtures of the two isomorphous silicates. Very recently, however, Tschermak has reviewed the scapolites,* and has reached the conclusion that the atomic ratio of the protoxyds to alumina therein is not 1 : 2, as hitherto supposed, but 4 : 9, or 1 : 2\. In other words, if we would compare them with the feldspars by multiply- ing their atomic formulas so as to get in each the same amount of silica, while anorthite becomes (ca3al98ii2)024, * Monatshefto fiir Chemie, December, 1883. seii 1,1^1 VIl!.] A CLASSIFICATIOX OF BtLlCATEH. 841 »'""> of ca,o,. liu'l f ' ■■",■" " »" "*''""■' ..f a *enn of ".e ^ene.'r ' ' J j," ' "- .""•■e t,,„ „t,.:'; lli>».i.. acco,-,l„ncf with Xr "''""" ^^polito. ^ = 2 .■ 9, „,, muUiplvill ,.„ '" '^^'""».'- """"i"". would be "'X^' into accorc/anoe ,v 'tT '''^.*" '"""'6 "'« fo" 9 = 36. lieverting to til' ' '',''''""»1< » conclusion, 4- d.£ferfromanorthite,„„ L ' °'/''-'!°''' "*<"■■■ ,erie, °'"' tlm-d of an a to"'! d ioC"°',"' ""■"■■'""ng each (o».al,„,)o„ + i,„,„,„„^, "^'°"»' 'i ,''™"'V'l. being «8 by three, to com,,are v ■H^^h^^■ + ^'l""- *'"'"My ov nnnonite, ,ve have for th L '!, '',"'"""" «''™" "l"'"' num. deseribod by Von, KaU,' '" '""""'"o "^ fi»- ^MoHne,l":hr::;tp'Ts rr -- - - ">;mall quantities in meio'nUeT; ""? ■^' ""'"'■ *^"""<1 ?;48 per eent, though an .„.,'",'"'""'"«' <»""nples The theoretica chlorifLt ' i^. "''''"" "' """-'"lite ■"g to Tschermak. a mt ° L r""?'.'*? '"' '""'""- accord- «uia, one atom of oxygr,! '','"?','' '""' "'« »'«'ve for- cent), or, in other worfs L 7Tt ^^ '^'""""^ (^-20 per ' the additional basic elem™t hi ' ,<^'""''»'*''")".. + Jna/cl„ ^yd- In the sca„on erTs i^;;'^ ?',? "'''™'' "«to»< o . the series, there a .pears a ml ^"^'^"P"''' "' "lending ^hich gradually rlE ".P™?"^^'™ '"crease in alkalies ".arialite ,ve fild cotfeaWe od """! '" '"^^"""o » general decrease in dens ^ fe at th;" '°"" P"'"^''' ^ but more accurate determiLatio", '„^\r'; """^ "P'^rent, fo>: the scapolites. VVe have T t. " *"=''"' "<> needed 83T, revised the atomic o™," ' ""'""^'"S t'^'*- pago the ratio of 4 : 9 for prott^^^ '!a .d^t '"■ "'^^P™" -W' The tntermediate sennolftes of H """"■'• «e.'.es are imagined by I'ohe^ak t be'Tr""""'''^'"* ^^' ^""^^ as proposed ft. >:'-^Mm«^i^7t&ifr-:^'^''^fSi'3m 342 A NATURAL SYSTEM IN MINERALOGY. [VIIL by Von Waltershausen and myself, crystalline intermix- tures, but binary combinations, in different proportions, of the two silicates, meionite and marialite. He notes, (1) compounds holding one equivalent of marialite to two of meionite, which are almost or completely soluble in acids ; (2) compounds with one of meionite to two of marialite, — incompletely soluble ; and (3) compounds with less than the latter proportion of marialite, — insoluble in acids. This variation in solubility will, in the chemist's eyes, be, as already shown (§ 32), a reason for rejecting the notion that they are admixtures, while he will at the same time repudiate the attempt to perpetuate in their formulas the dualistic notions of a former day. These intermediate scapolites, like the feldspars, labradorite, and oligoclase, and the various zeolites between thomsonite and stilbite, must be regarded as distinct species. § 77. In close relation to the scapolites comes a remark- able group comprising sodalite, nosite, and hauyne. Soda- lite has the atomic formula of a cblorinated soda-meionite, being (na4al9sii2)o2icli. Nosite is a similar species, in which the chlo iie is replaced by oxysulphion, while hauyne is another species, in which the proportion of protoxyd-base is greater than in these, giving the ratio 5 : 9 : 12. The relations of these various species to an- orthite and to each other, may, if anorthite be writ- ten (ca3al9sii2)o24., be represented as follows: meionite, (ca3al9sii9)o24 -f- caiOi ; sodalite, (nai,al9sii2)o24 -j- najcli ; nosite, (na3al9sii2)o24 + naiSi04; hauyne, (na3al9sii2)o24 + 2(caiSi04). Both of these sulphatic species -contain also small amounts of chlorine. Ittnerite is a hydrous species related to these, but containing a smaller proportion of sulj>hates than either, and, like the associated scolopsite, requires farther study. Lapis-lazuli, a sulphatic and sul- phuretted species. ' 3 composition of which is not accu- rately determineu, is apparently related to the sodalite group. Notwithstanding the resemblance in composition between these silicates :ind the scapolites, they differ very !• Vill.J A CLASSIFICATION OP SILICATES. *J» Wte, „. .„aeea tin L^wS™ itJrJ elude under the general nami \°1'""'3"'<', which we in- -^ to understand^Te .aturfofl ^'"'■''"'^ ^""P. '>Z to nephelite. This latter spterwhfh T' ''^ ''""■- volume near that of orthoZe s a tn ' "" '"°™'' than anorthite, and its atominfl " """'" «''<=ioi« ".ay be multiplied by six, taUnJ • '"" "• f"''-<^'^-^''->)'>,.,, more simply its relatioi to "?„ P'.!'^''««'^')o=„ to show »hich was formerly im™„,*^;X" % '""^ '"''"^"J. donate of lime wi'th a hydtted n T ,'!''""«"'^ "f carl rece«t studies to be an hftel, "?:'''' "PP'^"^ &»» *.th the sulphatosilieat s LTdthl ,°? ■*• "°"P^™'"<' ^odahte and soapolito grou™ Th ""°™^"i'=''tes of the . *'th a speeific gravity of 245 Jl' ""f "'"'^ <>* Miask, «p.esented by (na.al Ji^xfro"' """'f ^^ "-r RaulF, is while the canerinito o" Dtoo to!o '1 """'^ +''^'"1 (° = 6). - portion of potash, and XkZt^ *" ^och, contains f r *"™"''' '» the amounte of r ™"'',"'™ '™'° «>« hydrous carbosilicate, life ?he h i"™ ""^ ^'"«'- This ntnerite, will fi„d « ilace ''„ ^ 'T ^'P^atosilicate we\Ive thrrcmarkaUe° b"mio'^"^ ?"''"^'"'^'<= ^P^^oids, -g.the anomaly of a hig2 L'' '"? ''"'•^"''> P™»ent ''""ug the ratios 2:3:7 Z^T "'^T' *W<^''. "'hile spars and scapolites, is said tn ! ™'''™<' "^ the feld- M'larite, sarcolite, and g hie, if'' *'^ """"^ "^ -oids. f o«p in which, the rJoZ .LJTT "" '"^^^ting 1 ■■ 1, there is a great variation ,r^'' *° "'""""" b^'ng fr""' 1 = 1:8, i.. milarite toT- .'J' • f ™P"«» "^ -"oa! •'•-'"' sarcolite, aiuU : 1 . 'Hi.*! Wl Hi il ''^"***^'"'''?ii'?ri"-"r'Trr 344 A NATUHAL SYSTEM IN MINERALOGY. [Vlll. 1^ in gehlenite. In the native gehlenite a small portion of alumina appears to be replaced by ferric oxyd, but the artificial gehlenite from furnace-slags, analyzed by Percy, is without iron, and is an oxysulphid, containing 1.50 per cent of sulphur. Melilite, a spathoid silicate, is also found as a furnace-product, and, according to Percy, contains a variable amount of sulphur, equal in one case to 1.62 per cent, while the native melilite is destitute of sulphur. Under this name are, perhaps, confounded two distinct species. The luUficial melilite, whicli approaches the so- called humboidtilite in composition, has an atomic for- mula near to (ca3al2si5)0j , and a volume almost identical with gehlenite, while the rative melilite is more nearly (caoalisig)©^. Similar atf>mic ratios to the last, as regards the bnecs, are presented b; eudialyte, a zirconic spathoid, the composition of which is nearly represented by (m^zr2sii2)oi8, and which contains much lime and soda, with a little chlorine. The atomic formula of wohlerite, another zirconic spathoid, containing some niobic acid replacing silica, also with lime and soda, is not well estab- lished. In these two species we have examples of the com- plete replacement of alumina by zirconia. In melilite as analyzed by Damour, and also in gehlenite, a partial re- placement of alumina by ferric oxyd is shown, and a com- plete substitution of this kind appears in ilvaite, a spathoid species having a density of 3.71. The higher specific gravity of 4.00, observed for some examples of ilvaite, may show a related species, or more probably, as suggested by Dana, may be due to an admixture of gothite or other iron-oxyd. Of the species included in this tribe, wohlerite, is a niobosilicate, eudialyte, the scapolites, and sodalite are silicates more or less chlorinated, nosite and hauyne are sulphatosilicates, while lapis lazuli is also sulphuretted. Tribe 8. Protuperadamantoids. § 80. We come next to the Protoperadamantoids, a very important tribe. Beginning with the most highly »t /liV VJII. Species A CLASSIFICATION OF SILICATES. 345 D J'itrgusite Ke/lbauite. I Schorlomite. jldocrase. - I Garnet. JAllanite. - l^giiite. - I Beryl. - . j Euelase. - jArfvedsonite, JArdennite. jAxinite. - jEpidote. - jZoisite. - - I Jadeite. - I Gastaldite. I Glaiicophane I Prehnite. - I Acmite. I Spodumene. I Sapphirine. I Staurolite. I Coronite. - ISchorlite, - Aphrizite. - I Indicolite. JRubellite. - (oaaaljsi5)ojo - ■ (caialisijo^ - . (°iiaI,si,)o,.(m=ceJca^fej) (m3fisi,)o,,.(na,==„a,ca,fe.) (bo3al3si,,)oj3 .... (be,al3siJog + laq . . _ (inn^alasOog + iaq ... (^•axm,si3)o,+Jaq . K=.alffif ) (c%aIjsi3)oe ...... (na,aI,si,)oa (mialosiojog (nisalasiajois - . . _ _ (c32al3sifl)o„ + laq . . Kfi.«i.)o,.(m, = na,.,fe,;) ("-a],siio)o,5 (mg,a],si,)os (feial,siV5)o,.5 + ^aq - - . (m,al3si5)09 (miaI,ai6)o„ (m,al68i8)o,5 (mial„si,2)o,3 (m,ali2si,5)oj8 ).63 ? '.67 T. i.53 3.10; 3 20 I 5, ' y.08 : 5, ' .'3.00 I i.90 O. t.92 O. *-3« R. '.38 I R. -38 j R. ■33 R. .35 R. l^ole, connecting tl,e piotonp,«!i ! ^^""""^us amphi. iriiBiii m ■" ('■ cf > 1 mi ^" 346 A NATURAL SYSTEM IN MINERALOGY. [VIII, silicate, which, like titanite, is attacked by chlorhydric acid, a character not coiumon to many adainantoids, except the more highly basic species of the first sub-order, and already noticed (§ 57). In keiHiauite, one-third of the alumina is replaced by ferric oxyd, and in the titaniferous schorlomite, which is also attacked by the acid, and has a ratio very near to keilhauite, the whole of the sesquioxyd base is ferric, while a partial replacement of the same kind is observed in some varieties of idocrase. Next in order comes garnet, including many species, in some of which ferric or chromic oxyd replaces, more or less com- pletely, alumina ; while the protoxyd-base is either wholly lime, or in part magnesia or manganous or ferrous oxyd. The single example of garnet give". \n table A'^III. i.-; that of a pure lime-alumina species oxarii'uod by the writer. We have placed allanite near to garnet, for the reason that, according to Rammelsberg, the best determinations give approximately the garnet-ratio, 1:1:2, rather than that of epidote, 1:2:3, not vvitl 'standing that the species is homoeomorphous with epidote, a!)d is often spoken of as a cerium-epidote, to the atomic ratios of which some analyses f;i;:"ircntly conform. A farther study of the group * [ mi' ,rals commonly included under the name of allanite or orthite is required. The great differences in density; the facts that some resist the action of acids, while others are attacked thereby ; that some are anhy- drous, while others are more or less highly hydrated, — all lead to the conclusion that several species are here in- cluded. We have already separated therefrom the so- called xanthorthite, as a cerium-zeolitoid, and it is probable that besides one or two hydrous species, and a true ada- mantoid, there will be found at least one intermediate spathoid species. The alumina in th3 allanites is often in part replaced by ferric oxyd. A pure alumina-allanite, with the garnet-ratio, in which the protoxyd-bases are equally divided between cerous and ferrous oxyds and lime, gives the value for V as here calculated. Of the ;,l »• VIII.] A CLASSIFICATION" OP SILICATES. 347 illi species in the table, keilhauite and schorlomite are titano- silicatas, ardennite, a vanadosilicate or arseuosilicate, and axinite, a borosilicate, while both boric oxyd and fluorine enter into the composition of the tourmalines. § 81. The glucinic species, beryl, is generally regarded as having the atomic ratio 1:1:4, and has a volume near to garnet. The late analyses of Penfield* have, however, shown that beryl contains a small and variable amount of alkalies, replacing glucina, besides a portion of water varying from 1.50 to 2.50 per cent. He finds that the composition of the mineial is best expressed by the more '^ implex formula (gl5al6si22)0334-4aq, a change which, however, affects very slightly the values calculated in the table, that of Y being thereby changed to 5.48. Euclase, though closely related to beryl in composition, and, like it, liydrated, shows a much greater condensation. Ardennite, which presents the atomic ratio of euria?;. , and is also hydrated, is essentially a manganese-alivnuna sili- cate, with some magnesia and lime, besides a small portion of vanadate, more or less completely replaced, in s ;me instances, by arsenate. These latter elements are prdba- bly comparable, in their relations, to the sulplirtes m nosite and hauyne. Abstracting them, we find f tlie silicate essentially the formula given in the table, which can, however, only be regarded as ajf "oximate. Preh{;ltc.; although classed by Shepard in tl order Zeolite, be- longs to the present tribe. It has i .; ratios 2:3:6, which are those of euclase and ardennite, and like those, and epidote, is hydrated, while its voluuie is near to those of beryl and idocrase. The species axinite is notict u.e for containing some boric oxyd. The formula which we have deduced in the table, in which one eighth of the silica is thus replaced, and one third of the sesquioxyd is ferric, is, also, but an approximation. The composition of this, like that of beryl, of ardennite, and a great number of polysilicates, * Amer. Jour. Science, 188^ \xviii. , 25. ift. i jrs:SSSiiii:^ii..,.^fiMth3ta»ii ak,^ 348 A NATURAL SYSTEM IN MlNKllALOGY. [VIII. Jll cannot be accurately represented by such simple formulas, which, however, suffice to show, with sufficient exactness, the atomic volume and the place of the species in the system. § 82. We come next to epidote, the composition of which presents many variations, due in part to a greater or less replacement of alumina by ferric oxyd, and, in the so-called piedmontite, by manganese sesquioxyd. The presence of a small amount of water, equal to about 2.0 per cent, seems, as in beryl, euclase, and ardennite, to be essential to the composition of the species. The atomic formula for a pure lime-alumina epidote, as imagined by Ilammelsberg, is (caial2si3)0(; ; but such an epidote is unknown in nature, and we have, for the purpose of determining the A'olume of the species, selected a variety in which one third of the sesquioxyd is ferric. The for- mula, morei ver, takes no note of the small amount of water present in the species. Zoisite is essentially a lime-alumina silicate, seldom con- taining over five or six hundredths of ferric oxyd, and often traces only. It is not improbable that the true ratio of the protoxyd and sesquioxyd bases in these two species, as in meionite, with which they have been paral- jcled, may be represented by 4 : 9, rather than by 1 : 2. We note next the more silicious jadeite, whose formula, as already pointed out (§ 81), is related to that of zoisite as tliat of dipyre is to meionite. While zoisite is essen- tiiilly a calcic species, seldom containing over three or four hundredths of soda, iadeite is sodic, and it appears, like the comi^act zovAte or saussurite, to be anhydrous. The atomic volume of jadeite, as shown in the table, apj)ears to be less than those of garnet, epidote, and zoisite, showing a more condensed molecule. Gastaldite has the atomic formula of jadeite (mial2si6)09, but, with a dejisity 3.044, gives a volume of 5.61. § 83. We lia-^^e next to notice three remarkable adamantoids, closely related to those just mentioned. \ J li- as, !8S, the . of ater the The t2.0 to be toinic id by :>te is ose of variety he for- unt of am con- yd, and [lie true ese two n paral- y ^ •• 2- iformula, i zoisite is esseiv three or appears, hydrous, e table, ote, and astaldite t, with a hmarl?:able fientioued, VIII.] A CLASSIFICATION OF SILICATES. 349 and alfio to the spathoid, ilvaite. In garnet, axinite, epidote, and keilhauite, the sesquioxyd nuiy be in large part ferric, and in schorlomite and ilvaite it is entirely sd, the protoxyd-bases in these being cliietly lime, magnesia, and ferrous oxyd. We have in tegirite, arfvedsonite, and acmite, three well defined protopersilicates in which the sesquioxyd is entirely ferric and the protoxyd in large part sodic. These three species, which have hitherto been little understood, will be seen from the table to be related, respectively, regirite to garnet, acmite to epidote, and arfvedsonite to euclase, and to have a common value for V very near to that oi" garnet and epidote. The presence in each of these ferric species of large amounts of soda, equal to ten or twelve hundredths, is the more remarkable since the aluminous silicates with which we have compared them contain little or no alkali. This association recalls the highly alkaliferous character of an- other iron-silicate, glauconite. While these three homoeo- morphous species, all ferric bisilicates with soda, having very different ratios between protoxyds and sesquioxyds, are, from their condensed molecules and their indifference to acids, assigned a place among adamantoids, the related species, ilvaite, with a larger volume, has been placed among the spathoids. It is possible, from the analysis of Rammelsberg, that babingtonite may be a ferric species belonging to the one or other of these tribes, but without farther analyses it would be premature to fix its place. § 8-4. We come next to spodumene, a lithia-alumina species with the atomic ratio 1 : 4 : 10, remarkable for its great condensation and its volume of 4.88. It is instruc- tive to compare it with the still more silicious lithia- alumina silicate, petalite, which, with its lower density, has a volume of 6.33, and takes its place among the spathoids. The relations between these two silicates are analogous to those between zoisite or jadeite and a scapolite like raari- alite. While these two lithia-bearing species, with the ratio of protoxyd to alumina of 1:4, are among the 350 A NATURAL SYSTEM IN MINERALOGY. IVIII. most silicious known, sapphirine, which has the same ratio, is the most basic, and, with its atomic formuhi of (mgial4sii)ou, serves to connect the silicates with the spin- ellicls, while, by its great condensation, it takes a place by the side of spodumene. Staurolite is essentially an aluminous double silicate, with the ratios of 1 : 4 : 2|, the protoxyds being ferrous oxyd with a little magnesia and, rarely, a portion of oxyd of zinc. In one variety it would seem that manganese- sesquioxyd replaces a portion of alumina, and a small portion of water appears to be an essential element. Omitting the water, we get a volume of 6.01. § 85. We come now to the tourmalines, a family of silicates which, perhaps, might be called a sub-tribe, ^ince the five distinct species, representing as many genera, differ from each other not only as the related silicates, albite, labradorite, and anorthite, or as zoisite and jadeite, but also, at the same time, as anorthite differs from meio- nite, or as lime-garnet from idocrase or epidote. In other words, not only the relations of the protoxyds to the ses- quioxyds, but the relations of both of these to the silica, are subject to notable variations in species of tourmaline which so closely resemble each other that it is difficult, if not impossible, to discinguish them by physical characters alone. The studies of Rammelsberg, which first clearly showed the varying composition of the tourmalines, ena- bled him to divide them into five species, each of which is the type of a genus, distinguished by the ratios of pro- toxyd to sesquioxyd. We follow him in regarding the boric oxyd, which is fiot constant h\ amount, as replacing silica, and recognize the fact that the tourmalines are oxyfluorids containing a small and variable amount of fluorine. For the brown magnesia-tourmaline, the most highly protobasic species, with the atomic ratios of 1 : 3 : 5, a trivial name was needed, and we have ventured to suggest that of " coronite," from Crown Point, in New Yo.rk, a ■iJso for tl ^e ad ieast new s| fJuced sniall VIII.] ^ "'^ SILICATES. Of "Xi " t r ,;";'^, o"- "pec r""t" "■"" f ;."clico,ite " Ct,:ti^™- '""""aline/ ^ TT V5 iioni coninouiirI« h.. • ^^ ^^e pass in fu- amount of altnK P^otoxvd-basps ih^ ""^uiui- ratio of the sesni,;^^ i "^ ' "^^^'a-betirino- Thn * »f bono oxyd, is, moreover nnf. "'"' ™'T"'? amount »neraical variables to be tilt.,, ■ '. ^™ "'« ""is maiiv of a group of .mnerals^S frf T"""' "> «=e Xdy to'nal cbaraoters, were Irevt " *'""'' ^^Uarity i„ ex S 86. We have given i„ t,^ ^ differences in color also in table No Vm I ^^ Preceding |,aram-.,r j ^"^p-SoiiS^fti^^-r™-^^^^^^^^^^^^^ '°"' •■""^' '^"" '<>« exce/trs^^j.tTit;;' ;■; 352 A NATUKAL SYSTEM IN MINEllALOGY. IVIII. l:\l TTn Hi el ^ I ii'! ferriferous aphrizite, which is 3.20, range from 3.10 to 3.04.* These figures are adopted in the table, save that for rubellite the density has been placed at 3.00. The equivalent volumes, as calculated by llammelsberg from his own arbitrary formulas for these five species, respec- tively, gave him the discordant and apparently incommen- surable numbers, 144.6, 167.3, 241.0, 117.0, and 148.0, whicli fail to show any relations between the species com- pared. These, however, from their close resemblance in external characters, on the one hand, and their chemical differences, on the other ; from their varying relations between prot- oxyds, sesquioxyds, and silica ; from the partial replace- ment of silica by boric oxyd, and of alumina by fei Ic and manganic oxyds, are peculiarly fitted to test the correctness of the new method of study ; and this the careful determinations of Rammelsberg enable us to apply to the tourmalines with guaranties for accuracy not often to be met with. The results of such a study of these five species, as set ^wn in the table above, may here be stated at greater length. In calculating the mean weight (P) of the oxyd-unit for each species of tourmaline, care has been taken to get the nearest approximation to the results of Kammelsberg's analyses of that species. The manganic oxyd in indicolite and rubellite is included with ferr.c oxyd, and the various protobases always present are grouped under the heads of ferrous oxyd, magnesia, soda, and lithia. The formulas thus arrived at, with their frac- tional coefficients, and the value of the oxyd-unit, got by dividing the calculated equivalents by the number of units in eacli formula, are subjoined. But these formulas do not take into account the fact that all of these tourma- lines contain from 1.5 to 2.5 of fluorine, replacing a por- * For the original memoir of Kammelsberg, see Pogg. Ann., 1850, Ixxx., 449; and for a summary of his results, Amer, Jour. Science, xi., 257; also a farther discussion thereof by the present writer, ibid., xvi., 211. For later studies of the tourmalines by Kammclsberg, see Annal. Phys. Chem., 1870, cxxxix., ;37'.» and 547. VHI.J A oi,ass,k,ca™k or sa,c..x.«. S. CL 'P. a o 5- ? £^ o" S" o 3 i & a a s. ? is- p !r jr tion of oxygen, tl,e mean of .1- 1-90 of «uori; "'■ '■"'"^"''te 0.17 and 18?' u" '"''"""" "« ».;'- wit„ont"; ,""-/'"■■ values arp i.i..^ i ^iiese t'.-es, wl ,e' i rl,™' '■; "— ■ fof Uiis mean n '^""f'*'' values ployed for J ' ? ' "'"' "'""o "m- volumeo/tt^t! "^"""'*°'>»« values of V !l, °w '"'"^"- The species a rl„, Zl > ''"'' «^« rectness of ih^ - ° ^"^ cor- "■e aeeurar„"«,^^''; "■"•*» «««^ of Ra„f„elL"g. "''"""»"■ P^UoIdJpUtrm''''/"'''!'-- ""the p/noi;r«:^:r;«''- of endeavored to set forth 7„ T accounts of r,i.„„.j- " ""> allows ffrmt r°<"^'"S tribes. It wi, gieat variations in tl,. '"t'ons of protoxyds to''"" ^Worites/t^h^'lirror.",? Muscovites in , 1 • 1 ^ • ^' to Ferric and cln-omic oxyds l\' ^^ I « « S §§ « , ^ , species, replace, moreTt "'^ ^-^~^-^d_ - less co.pieteI,, alumina, and t ^ ■>. IMAGE EVALUATION TEST TARGET (MT-S) // ^ J^^4^s 1.0 I.I l^|2.8 ■ 50 l"^" 116 2.5 lii£ 12.2 Hf 1^ 12.0 1.8 "" ■ / 1.25 1 u 1.6 < 6" ► ^/^ >>; ■c*l >tv >> V /A Photographic Sciences Corporation \ J. .V NJ :\ \ V -'^\ c> '<^ 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) •72-4503 '""^ ""' 354 A NATURAL SYSTEM IN MINERALOGY. [VIII. Table IX. — Pkotoperphylloids. ' Species. FOBMULA. P D V X Astrophyllite. - in m SI ^rt B»--.Mv>v . . . 3.32 • • • 0. l™5"*2«"ioPn --------- Phlogopite. - - (m,al28i8)0ij-(m4=mg3.jko.j) - - 18.12 2.85 6.35 0. Pyrosclerite. - - (mg^aljSiaKa + 3aq 15.40 2.74 5.62 0. Penninite. - - (mg,.oal,.osi<.5)o,o.5 4-3aq - - - - 15.40 2.67 5.70 R. Ilipidolite. - - (mg5al38ie)Oi4 + 4aq 15.38 2.70 5.70 C. Prochlorite. - - (m4al3si<.„)0ii.j8 + 3aq - (m.^mg^fej) 17.72 2.96 5.98 H. Leu(!htenbergit«. (mg4.5al3.(^i5.o)Oij.5 + 3Jaq- - - - 15.46 2.65 5.8;-) H. Venerito. - - - (m4m3sie)oi3 + 4aq 16.84 ? CorundophUite. - (m^al^siOoij + Sjaq - (m4= rnggfe,) 15.20 2.90 5.21 C. Biotite. - - - - (m4m4sie)Oia-(m4 = mg3.5ko.5) - - 18.18 3.00 6.06 H. Voigtite. - - - (m4m4sig)oi8+ 4aq - (1114 = mgjfe,) - 16.48 2.91 5.66 ? Cryophyllite. (m3al48ii4)02i-(m3=nfe,kilii) - - 17.90 2.91 6.15 0. Seybertite. - - (maal9Si5)02o + Jaq - (mg = mg4ca2) - 17.97 3.15 5.70 0. Thuringite. - - (fe8m8sig)024 + 6aq - {m^ = algflj) - 19.56 3.19 6.13 ? Jeflerisite. - - - (rnggwigsijojo + 7iaq - (m = alBfi, ) - 14.92 2.30 6.50 0. Annite. - - - - (mgWiuSiisKg -(m6 = fe4ks) - - - 20.84 3.17 6.57 ? Willcoxite. - - (m6di28iio)o28+ 2aq - (mg = mgjna, i 16.76 ? Chloritoid. - - (feial38i2)og + laq 18.00 3.55 5.07 C. Lopidomelane. - (m,m38'4)0g 3.00 H. Zinnwaldite. - - (mial38i8)oio-(m = ko.5lio.5) - - - 17.20 3.00 5.73 0. Oellacherite. - - (mialjSi6)Oio +Iaq - (m =kibain)g^) 17.33 2.99 5.79 ? Lepidolite. - - (mi.oal4.88i8.o)Oi3.5 - (m = Kbh-i) • 16.85 3.00 5.61 0. Margarite. - - (caialai8i4)Oii + laq 16.58 2.99 5,54 0. Euphyllita. - - (m,al8Si8)o,8-(m = ko.s3nao.6.) - - 17.07 3.00 5.6!) * Cookeite. - - - (miali,8i8)o2o +5Jaq - (m^lio-Tsko-js) 14.80 2.70 5.48 ? Muscovite. - - (kialgsig)©,, 17.75 3.12 5.68 0. Muscovite. - - (kialgSigK + 2aq 16.77 2.85 5.88 0. Muscovite. - - (kialgsiuloa 17.27 0. Damourite. - - (kial98ii2)022 + 2aq 16.68 2.79 5.94 0. Muscovite. - - (ko>5al8.o8l9.o)Oi5.5 ------- 16.80 ... ... 0. Muscovite. - - (ko.5alg.9Si,.o)o,5.5 + 2aq - - - - 15.91 2.75 5.78 0. vm.j A CLASSIFICATION OF SILICATES. 355 Protobases vary from «n. • • P-sent to otVTn S^^^ f ^^^ ^^^^^- only are P aces being partially oT^tllfsuZ- ?k ""^^"^^' ^^e- cupric oxyd, magnesia hZTn/^^'^^ ^^ ^^^^^us oxyd «f -lica to the base?; res\vTder"^ '"'^^^- ^^^ '«^^o same proportions of protoxvd '^^^^^ '^''''' ^^^^^g the "moreover, species hfvW^f, ^"^ sesquioxyd bases ; and compositionrand siX L^^^J *^^ -me ehe'm af presence or the absence of t"S^ TT'^ ^^^^ ^^ the condensation, as shown byTh. , ?"^' *^^ degree of er^hly among ph^lloil! I f V£ ^^ ^' -"- -nsid well marked in th^^r physical A '' "^^^^'^heless, so n^icaeeous or phylbid typf fine 7^^? '"^'^ he by the student. ^^ '^ """^ «^ the first recognized § 88. We have sookpn nf • *sti„ctio„ is an arbtoLVL""" T^^orites, but the hydrous magnesian imZiZ' '■""" ^^ "^^""n from "tes, is not so great as 't at frri?'"°S°P"«^. '« ohio Th^^i *« ""-^"ovitie type oH '^^ ?»« »>icas to -llie difficulties of arf»n.,.; > ^: " *''einseives Jjvdrat.,) -e increased hy ZZTit7''Z'"« ""^ S'-'tfbe PWogopite, biotite, and musofvit T^ ^P^^^" "ame "ni.ke compounds, which houU T' '"^'"''^ "'•^">-^% The rauscovites present ,„ °™ ""^t'lct specie, f'fering Hke theCmlnri t^™"'™'^ ■"»"?"-* the micas included under the „ *''^"'/'<>">ie ratios, and variations in the ratio of ^X?/, ^^ P^logopite ^how ^•lto8:2; while there aret- ? '° ''""l'»<"'yds from f d 1 : 2. More highly L,l'"'' ^""" ^ = 1 to 1 l" however, astrophylmf l^^'^T "''"' ""^ Phlogopite' h 2 ferric and zirconie oxvd? « i^^ "'"■"'"" and in part ' the specimens from CoSd, ' T*"^ '» «»"'• whileTn I present in traces ont with K^!?" ''y ^oenig alumin" ■ formn a, 1 : j , 2, we ma/note tt mf f "^'''"^ *e atom"; ^•^=^---»--^hitrrr:ft:;--a 356 A NATURAL SYSTEM IN MINERALOGY. [VIII. quioxyds, come the highly silicious fluoric lithia-mica, cryophyllite, 3 : 4 : 14, and die basic seybertite and will- coxite. The ratio of protoxyds and sesquioxyds in the latter, 1 : 2, is that of some biotites, and of the ferric spe- cies, annite, near to which is the still more ferriferous lepidomelane, apparently 1:3:4. A like ratio appears in the dense basic chloritoid, 1:3:2, and the more silicic zinn- waldite, 1:3:6, followed by the barytic species, oellacher- ite, 1:4:6. In biotite and voigtite one fourth, in annite, thuringite, and jefferisite one third of the sesquioxyd is ferric. After lepidolite, probably 1 : 4|^ : 8, and, like zinnwaldite, a highly fluoriferous mica, remarkable for containing lithia with caesium and rubidium, we come to the muscovites proper, with which the last two species are connected by the fact that their protoxyd-bases are alkalies only. The variations noted in the ratio of these to the sesquioxyds (in which ferric oxyd replaces a small portion of alumina) are from 1 : 6 to 1 : 9 and 1 : 12, and the ratio of the sum of these to the silica in different analyses is from 1 : 1| to 1 : 1^. From various muscovites have been deduced the atomic ratios, 1 : 6 : 9, 1 : 9 : 12, and 1 : 12 : 18, with others intermediate, and a careful study would probably show, as in the case of the tourmalines, the existence of a series of muscovites. Near the muscovite with the ratio first named must be placed the less silicious and somewhat calcareous species, margarite, 1:6:4, and farther on, euphyllite, 1:8:9, and cookeite, 1 : 10 : 9. Of the phyl- loids, phlogopite and cryophyllite contain more or less fluorine. In calculating the value of P for both biotite and voigtite, W4 = algfi^. § 89. It will be noted that in this list we have included both hydrous and anhydrous species, between which it is impossible to draw a line of demar'Vj,tion. Phlogopites and biotites are reputed anhydrous, but, as is well known, contain in many cases from two to four hundredths of water, while corundophilite, willcoxite, seybertite, chlori- toid, oellacherite, margarite, euphyllite, and cookeite are vin.j 867 all more or less hydrous • th„ ' «x hundredths i„ euphvliVl ' "?'""'* »' water risi„e t„ oooteite A„,„ ^^ •'^ "'"' "id to twien tK.f "'"8 to , ™- Among inuscovitp« :„ ,.,"^ '«»« amount in f°"nd .n all proportions "1;, '*\ banner, water i" damourite and pa'ragonite 7bill Tsf '^t""^ ^P'-- «' t f •'""' ^""la-rausoovite Th. "^ "'^ described as :^ater„,phy„„id,„^j;*^ The pre^^^ „^ ^^^_^ d as tt classihcation among nhvl^J ^'°'""' "^ a distinction »antoids, where we 6nd ber ° Iv'T "'"^ ^an in ada »>te,p,eh„ite,epidote"a„d m2 ^''' ^^'•^'' ^"lase, arden en ers as an essential 'ZeT"''''' '" "" »' -hioi water Tn^^'^^ZS^^"^-- the Ohio- to the phlogopites and the b; ff '™^' "eariy related IrtT'^ V"^"'"- PWogopt ttth ?"'' ~-^to apart from the water, of 4. o « '"*.tl>e atomic ratios mous species, represented by 4 27/""}! '^ * '-' ^ the closely related ripidolii '. ' t*" ^*'<"' *I>ese come and prochlorite. Cron S ttth Ti" '"^'*^' «"-"«! able as an example of a well Tfi!' . ^T"^ « ■•e-nark- nt.c species very near tolchtorl '"'' ""■^^•^"'"^ ^ll eonta.„.ng a large proportfon ofXpe" ?"^t"' ">»' • Thi, apecie,, „i|„i "PP*"^- CorimdophUite from six to seven nir ^P"" ^'•"' ^^ ^^ich severaUho.? .^ ^'*^"^° ^een schists of the reS" th ''" ""''' ^'^ ^o^ncl ?„ the Sn ■''' '"^"'^ «»d eral in question r^wu""-^*""^ ^^^^^l slates ofJ*^^"'^" crystalline eral feetVth Ttrata : "^"ted ^° S'-^-ter o t^s abtS.r'' V *^« '°'°- Poor in copner f'!**' *^™ating with layers of » Z^^*'^^^ through sev- with the S;i ' ""^'^ '"^'•^«d ^ith feSL.'TT*""^^'-«^aterial an inch or xnoretTv. T '""'^^'^^ted with velTn "'^^ ^'"^^ ^°'ncide g'-ams of quartz and a small 358 A NATURAL SYSTEM IN MINERALOGY. [VUI. follows, with 4:4:4, and the more silicious species, voig- tite, 4:4:8, a hydrous biotite. From this we pass to portion of magnetite. A qualitative examination of this material showed that it contains no carbonates, and is not of the nature of a clay, but con- sists of a hydrous silicate of magnesia, copper-oxyd, alumina, and iron- oxyd, constituting a kind of copper-chlorite. It is but feebly attacked by dilute acids, while strong acids, and notably sulphuric acid diluted with two or three parts of water, and aided by a gentle heat, readily and com- pletely decompose it, with separation of flocculent silica, which, by solu- tion in dilute soda-lye, is readily separated from the accompanying quartz and magnetite. A single somewhat rough analysis made in this way gave me, for 100 parts: insoluble sand, 14.10; silica, 24.60; alumina, 13.00; magnesia, 15.15; ferric oxyd, 7.11; cupric oxyd, 15.30; water, 11.50. =100.70. The qualitative examination of a considerable portion of an- other and less pure specimen, gave an appreciable quantity of zinc, and a distinct trace of nickel. A portion of the specimen of this copper-silicate of which the analysis is given above, was freed by careful washing alike from the coarser grains and from the lighter portion, which remained long suspended in water. The material thus purified was somewhat richer in copper than before, and has been carefully analyzed by my friend, Mr. George W. Hawes, of New Haven, who found: insoluble sand, 6.22; silica, 28.93; alumina, 13.81; ferric oxyd, 5.04; ferrous oxyd, 0.27; magnesia, 17.47; cupric oxyd, 16.55; water, 12.08=100.37. This, deduct- ing the insoluble matter, gives, for 100 parts: silica, 30.73; alumina, 14.67; ferric oxyd, 5.35; ferrous oxyd, 0.29; magnesia, 18.55; cupric oxyd, 17.58; water, 12.83=100.00. This, as remarked by Mr. Hawes, gives, on calcula- tion, an oxygen-ratio between protoxyds, sesquioxyds, silica, and water, of 4:3:6:4, very nearly, which puts this mineral, if it be a homogeneous substance (as its microscopic characters would indicate), among the chlo- rites, some of which it resembles very closely in its atomic ratios. Before the blowpipe, on charcoal, it swells, then fuses quietly into a black globule, giving the usual reactions for copper. The iron is almost wholly in the state of sesquioxyd, as shown by two determinations of the amount of protoxyd of iron, which gave, respectively, 0.27 and 0.29 per cent. This copper-chlorite appears, alike from its physical and chemical characters, to constitute a distinct mineral species, for which I propose the name of Venebite, in allusion to the mythological and alchemistic name of copper." ( " A New Ore of Copper and its Metallurgy." ) Trans. Amer. Inst. Mining Engineers, iv., 325. The atomic formula for venerite given in the table above represents it as a chlorite in which a part of the sesquioxyd is ferric and a part of the protoxyd is cupric. This formula (mg».r6CUi.26al3.6ofio.6osi6.oo)oi3.oo+4aq, requires: silica, 31.4; alumina, 14.8; ferric oxyd, 4.6; magnesia, 19.2; cupric oxyd, 17.4; water, 12.6=100.00, which agrees very closely with the numbers deduced by Hawes from his analysis, and varies but little from my own analysis, given above, of a less pure specimen, when calculated for 100.00 parts. A microscopic examination of this curious chlorite will VIII.] A CLASSmcATION OP SILICATES. 359 -»e ratb „f piVx^d 'sf'"' ^P^''- having the »«-otoperphylI„id triL frl" 17'- '''.^-'gh-'ut the whole auhydrous and a hydm'jT- ^'''''g^P"'' *« musoovite an ""de up of the Xe h; ', r "1'!"'^ '^•"<>»«o group i" »«-. In these eomparS ^^ -" """"'«'«''* '»« tie atomic ratios from th formal ™ ^"'™™"y ''^d-oed tern of Mineralogy," whli ^"''" "' Sana's "Svs "ud.es of Rammelsberg^hen 1 ,""""■"*» <"'-'»'i»al throw much Usht on thJ ■ P'^°P^'^^y interpreted will |-eaters»ph|ty::hett:;r,'t^^^^^ ''' ">'— ^- X have been omitted in diseussTnaTi^ "T ™" '° Table § 91. Related to the chlor "sTut^.T''.'"^ '''™'''''^- toe. IS the epichiorite of Ram„t\ """^ '" ^'™«- fibrous or columnar and ^''"""^'^erg, described as 4.;3--9:4,whioheo™sp'nVt:T^ *^,.'"°"'° -«ot "te. In this connection mav be '"'"■V^"'™»« Prochlo- nndescribed mineral wh^hfefouml """*""""' * '^"'erto «'e and it« accompanying caAo„l ™"t '" *''« ''"*''™- n-onth, Rhode Island. TWs sltn .' ""'"'^ *" ^'''-t^- • penetrating quartz, but in it" trf ;\'°"'«'-"es seen l-^yish-green mass, consis W of 'le ! f 'PP'""^ '«« a fibres, resembling chrv-oHU „ • .' ''"nsverse, flexible been confounded Tnor . "^ ."""""thus, with which it has • ef an inch wide gave me h!"" ^" ' ^'''' "'""■* "» eight! 21-80 : ferrous S;n6W mr=^"''?''^-««--''>»ina: ^ ' '"'•^"' "nagnesia, 8.96; lime, 2 01 • •^ iS n It 860 A NATURAL SYSTEM IN MINERALOGY. [vni. potash, 2.69 ; soda, 4.24 ; volatile, 9.30 = 102.90. A sub- sequent microscopic examination of the material analyzed showed the presence theriiin of interspersed films of pyrites, thereby vitiating to some extent the results of the analysis, which deserves to be repeated on a portion of the miner.'.l purified by the aid of bromine-water. Mak- ing allowance for some pyrites, the atomic ratios of this fibrous species are 4:4:6:3, being near to prochlorite, and to voigtite, which like it contains a little lime and soda. In this unnamed amianthoid mineral from Ports- mouth, possessing nearly the composition of a chlorite or a hydrous biotite, and in the epichlorite of Rammelsberg, we have apparently examples of a hydrospathoid form of these alumino-magnesian protopersilicates. With these should be noted the pilolite of Heddle, who has described under that name the substances hitherto known as mountain cork and mountain leather, which have a fibrous texture, are more or less flexible and tough, and occur in veins or fissures alike in crystalline lime- stone, in sandstones and shales, and also, as observed by the writer, as- a deposit upon quartz crystals in granitic veins. From several analyses by Heddle, pilolite is shown to be a highly hydrated silicate of alumina and magnesia with ferrous oxyd, and is nearly represented by (mg2.6feo.6al2sii6)o2o+12aq, a formula requiring: silica 51l7; alumina, 7.8; magnesia, 11.5; ferrous oxyd, 6.2; water, 24.8=100.00. More than one third of the combined water is expelled at a temperature of 100° Centigrade.* Tribe 10. Pinitoids. § 92. Corresponding with the ophitoids of the proto- silicates, we find in the present sub-order a tribe of hydrous silicates which, since the species known as pinite may be taken as the type, we have called Pinitoids. These bodies approach in composition and in density the * Mineralogical Magazine, 1870, ii., 206, cited in the Third Appendix to Dana's Mineralogy, p. 94. VIII.] ^ OlASBmcATIOK OP SILICATES. 00 urre„oeofthi,„,j,,i^,^;^a™ esewhere noted the ^^^^ ^— PimroiDs. ot other species. ?;„,>« • -^ — ■' a«d alumina, having" 1 1!!''^*^^"^ ^ «i««ate of potash *hus approaching cloTewT "^^''' «^ ^ •' « •' 12 : 3 and ;«"«eovite. Coss^aitewhLlT^^^^"^ '^ ^ ^dCs ^«' 'f not a pinitoid, aclmr . ? ' *^'' ^'•'^^^°« of 1 : 9 12 o musnnvif^ „- , . ' '^ compact narao.^,.,-^^ .. , • ^ • i^ . J, -- 'f not a Pinii ra'clr ?" "" ™«- of 1 9T""/ than ;. 'di:5 •• «• i-^ probably r.,lT"^['' "' ^"h the ratios, 1 . 12 Ji^T""""*. which resemM, "f" """^'^ 862 A NATURAL SYSTEM IN MINERALOGY. [VIII. I amorphous silicates with varying amounts of water, and the atomic ratios, 1:3:5, — the protoxyd-bases, as in joUyte, being chiefly ferrous oxyd and magnesia, with a little potash. Bravaisite, with the ratios, 1:3:9:4, and hygrophilite, 1:5:9:3, are similar species, the protoxyd- base of which, as in pinite, is chiefly potash. Near to this is sordavalite, a similar hydrous species, of which the prot- oxyd is magnesia and the peroxyd is in part ferric. With these pinitoids we have placed obsidian, pitchstone, tachylite, and palagonite, to which latter the atomic ratios, 1:2:4, excluding water, have been assigned. That its composition is not clearly fixed, or rather that more than one silicate may have been included under this title, does not detract from the interest which attaches to this curi- ous, unstable, hydrous colloid, so long since the object of studies by Bunsen, the importance of which I have else- where pointed out (^ante, page 129), and which are noted below, in § 109. § 93. The species of the sub-order of protopersilicates which approach the persilicates in composition, resist chemical agencies more than those species which contain larger amounts of protoxyd-bases. To this greater stabil- ity is due the fact that such species are often produced by the partial transformation, through aqueous influences, of silicates like the protoperspathoids. Such silicates, formed originally by igneous or by aqueous action, may thus con- tinue to lose protoxyd-bases, often with silica, until a con- dition of comparative fixity is rcaclied by the production of bodies having the chemical composition of pinite, of the muscovitic micas, and even of pyrophyllite or of kao- lin. Inasmuch as such compounds are in many cases the result of secondary processes like that just described, chemists have been disposed to assign a similar origin to them wherever found, not considering that where the proper chemical conditions unite, these compounds may be directly formed. That nephelite, for example, may, as is supposed, under favoring circumstances, be transformed vni.j ^ C,^«mOAT,OK OK »„.,e^,^. 363 'nto pmite, is aue to th. . r.t"-re^f;- rj':? r/r -«- ^^^^^^^^^^^^^^ ^n the inorganic vvorJd JW .^^ '""^'^''^ *'f the /ittesf '' conduce to tho »... i .' "^ *'i««o verv minf '^ epigenesiH .f / ^'^"ction of such st.I f ^'""' ^"^''^'J» J"-' mentioned 2 ?'"" ^""•"' »* W"" i h "'""■ pinite, or bed** nf »,• . " other word-^ tl.^ i i * STanftL • niica-sch St, lite tk. ""'^™' tue beds of gwnitic veins, have not be-n „ ! muscovitio mica, of SL: s^^^r'^ of sedi,::?,ts d ™: d°?"-"™^- oraers of minerals. Such nhZ?^' ^''■''ates, as in other ^flic'ed in fissures and opet ^f ^,,'>^™ been especiX gy of Canada. 1863, pp. 482-480. 864 A NATURAL SYSTEM IN MINKKALOGY. [VIII. W: f channels for waters of changing cuinposition and tempera- ture, during the long process which has filled these open- ings with mineral masses. In this way, crystals deposited at one stage are attacked at another, and are either more or less completely dissolved or transformed into insoluble products which are now found surrounding nuclei of the unchanged mineral, or in some cases penetrating its sub- stance. Examples of such actions are familiar to all who have studied attentively the history of granitic and related veinstones. Care should, however, always be taken in the study of pseudomorphs to keep in mind another and a different phenomenon, namely, that resulting from the power of a substance in the process of crystallization to cause other bodies to assume its own geometric form. Examples of these are seen in the cases of calcite, dolo- mite, and gypsum crystallizing in the midst of silicious sand, by which are generated such aggregates as the so- called crystallized sandstone of Fontainebleau, which, while having a crystalline shape belonging to calcite, includes from 50 to 63 per cent of quartz grains. A not less remarkable case is seen in staurolite, which, according to Lechartier, retains its crystalline form and general aspect even when by the inclusion of foreign matters, chiefly quartz, the proportion of silica is raised from the normal content of 28.0 to 50.0, and even 54.0 per cent, cor- responding to more than one part of quartz with two parts of staurolite, the mixture still retaining the crystalline form of the latter species. Thus a compound in crystal- lizing may give its geometric form to a large portion of some extraneous matter, which it compels, as it were, to assume its own crystalline shape. '5F i'i i' Tribe 13. Peradamantoid. § 95. It may be noticed that in the second sub-order the less protobasic silicates do not assume zeolitoid or spathoid forms. With the exception of sloanite and for- estite (both of which demand further study), we find no VlII.j ^ c^Asa,r,c.Tro» ok sa,0Ai.3. •P^O'-o^ in those tril ^ ™'* ««» p""toid5, through ;;ctf "''"'"»""'"'». I'l^yi w * r "^'--ted with thft of the ""■"'' ""» '"l^'oS t'." Species. | rT^ '^ T ^ ' D / y Oumortierite. •Andalusite. . ^'ibrolite. - . Topaz. . . . Pyanita . . ' BuchoJzite. . ' Xenolite. - . ' "^orthite. - . ' I-yncurite. - ' Malacone. - . . I Zircon. ... I Auerbachite. - - I -^thosidepite. (al,si,)o, (a^38i,)oj . ^*J»8J2)Oj - (aiasijoj . (a'3Si,)o, . (^hsish, - , /a^6Sij)o„ + iaq| (zri8i,)oj, . KsiiK + laqj (zi"iSi,)o, . ^fii8>sK + |aq • • • . exceptions, included in fi ^ ~~~ f '•e designated, as we L?' ^^^^^sponding tribes Th atnn.;!?!'" respecting it, to h. fT"' ^,^^«^er, from the details ^ ven "^°"^«^^'«"d, would seem h ^^^^^nite, ^tomic^rJlVf ^^«^i"^ it, to be a perz'eorr?::' ^^"^"^ ^^^« inic 7,-rn. • '-^^ b^smuthic oxv.l i ^ Peradaman- %'^'.^^J 866 A NATURAL SYSTEM IN MINERALOGY. [VIII and the related agricolite and bismutoferrite, will, from their chemical and physical characters, take a place in the tribe of perspathoids.] § 96. The adamantoid persilicates constitute a charac- teristic and remarkable group. Of the aluminic species, we fi'id dumortierite, andalusite, fibrolite, bucholzite, and worthite, with differing atomic ratios, and in one case hydrous, all presenting the same value for V ; besides the remarkable oxyfluorid, topaz, and the more highly con- densed kj'^anite and xenolite, the latter two having a smaller atomic volume than any other silicates known. A single ferric species, anthosiderite, appears, and more than one zirconic species. It is known that minerals having the crystalline form and the centesimal composition of zircon present variations in density from 4.86 to 4.02. The careful studies of A. H. Church, in 1875, confirmed the previous statements of others as to the differences in density among the minerals included under the name of zircon. Thus he found the hyacinth-red crystals from Expailly to have a specific gravity 4.863, which was not changed by ignition. Of a large number of zircons examined by him. not less than twelve varied from 4.60 to 4.70, while an opaque brown zircon from North Carolina had a density of 4.54, which was changed to 4.67 by long ignition, and a transparent brown zircon from Frederickvarn had its density by the same process raised from 4.48 to 4.G3. Another zircon, dark green in color, slightly opalescent, and flawless, had a pecific gravity of only 4.02, which was not changed by igjiition. It was, nevertheless, according to Church, a true zircon, giving by analysis the percentages of that species.* Auerbachite, an isomorphous zirconic species, has, with different atomic ratios, a specific gravity of 4.05, and agrees in volume with malacone, a hydrous zircon, which has a similar density, while other related hydrous * Church oil Densities of Precious Stones, Geological Magazine for 1875. (• [Vlli from ice in ;harac- ,pecie8, te, and le case des the ily con- iving a known, id more me form ariations of A. H. iments of , minerals :ound the a specific on. Of a less than [ue brown 54, which ransparent lity by the ker zircon, [wless, had hanged by [church, a es of that lie species, Lty of 4.05, ms zircon, d hydrous Magazine for VIII.] A CLASSIFICATION OF SILICATES. 367 zirconic silicateEi give tjpecific gravities of from 4.00 to 3.60. It would appear that the zirconic, like the aluiiiinic adamantoids, include species varying alike hi atomic ratio, in condensation, and in the presence and absence of water. An anhydrous zircon with the ratio 1 : 1, and a density of 4.86, has an atomic volume of 4.68 ; and one of 4.70, a volume of 4.84 ; while a zircon with the lower density of 4.02 has a volume of 5.65. This last we may distinguish by the trivial name of " lyncurite," from the lyncurion of Theophrastus,* while the denser zircon of Expailly will also, perhaps, require a distinctive name. [Oerstedite appears to be a less silicious zirconic apecies than those already named, and hydrous, like malacone. Its analysis affords ratios approaching those of the aluminic species, dumortierite ; but a small amount of protoxyd-bases serves to make its composition doubtful. If theee make an integral part of the species, it will take its place, with catapleiite, wohlerite, eudialyte, and astrophyllite among Protopersilicates.] Tribe 14. Perphylloid. § 97. As regards the phylloid tribe of the persilicates, an important chapter of their history is connected with pholerite and kaolinite. It was in 1825 that Guillemin described, under the name of pholerite, a hydrous silicate of alumina, micaceous in structure, to which he assigned the atomic ratio for alumina, silica, and water of 3 : 3 : 2. This was the same as that deduced from the analyses of Brongniart and Malaguti for ordinary kaolin, although Forchhammer had proposed for the latter the ratios 3:4:2. The uncertainty as to the composition of these silicates which prevailed thirty years since is reflected in the fourth edition of Dana's " System of Mineralogy," published in 1854, where the first-named ratio was ascribed to kaolin, with the remark that it occasionally presents the second ratio ; while of pholcite it was said " that it does not * Moore's Ancient Mineralogy, p. 145. I i >t 368 A NATURAL SYSTEM IN MINERALOGY. [VIIL differ much from kaolin in composition." Hence it was that when, in the same year, the writer found and ana- lyzed a crystalline micaceous kaolin giving for its compo- sition, silica 45.50-46.05, alumina 38.37, lime 0.61, magnesia 0.C3, water 13.90 = 99.56, the analyses of pholerite and of kaolin were discussed by hm, and the conclusion was reached that the first ratio mentioned might represent the ^ composition of both of these substances when free from Table XII. — Pebphylloids. Species. Formula. P D V X Pho:erite . . . (al3si3)o8-f-2aq . . . 14.25 2.51 5.67 0. Talcosite . . . (ali,si,)o„-f-laq . . . 15.b3 2.50 6.13 ? Kaolinite . . . (aljsi,)©; -|- 2aq . . . 14.33 2.63 6.44 0. Pyrophyllite . . (al2si5)o, + faq . . . 15.00 2.80 5.35 0. Pyrophyllite . . (al2si4)08 + §aq . 15.00 2.92 5.13 0. foreign matters, and consequently that " the mineral in its pure form is no other than a crystalline kaolin." In 1863, in accordance with this view, kaolin and pholerite were regarded as identical. Of pholerite it was then said that " it may be regarded as that substance [kaolin] in a crys- talline condition. From its foliated or micaceous struc- ture it may be considered as a hydrated Uxica."* It should here be added that the writer had in 1855 an opportunity of comparing the crystalline mineral ^rom Canada with the original pholerite, and of discussing the question of the minerak w'th Guillemin, in Paris. § 98. Jo was not until 1867 that this subject was again taken up, and this time by S. W. Johnson and J. M. Blake,+ who showed that, as regard.^, the composition of * See Hunt, Report of Geological Survey of Cana^'a, 1853-56, p. 386, and further, Geology of Canada, 1803, p. 495. t Anier. Jour. Science, xliii., iJol. hyd I'eta VIII.J fo™ which ar^T"^ "^ ^''"""irafX?;'''''"''""™ Jislieri ^u '^^wiout, however fiii,,^- ^ name of - se;eT'^-»*ie3ub^e""t,f ^ ""^ »™ pub! wiu serve to show whv ih^ ^"^^ ^"storieal qItIj- i, ^ng in 1854 and ^m^\^ P'""^"* ^"ter, whOe 1 ^ § 99- In the )ast!!^* ^^^ ^'^^''^'''«- designated Argiiloid m f "*" "^ "'"ys, we hZl position assigned o K . " "■'"^''''1 having thl T -.cnsoono,ud:fth!t4'nw,:^'"'--^^^^^ a^i Kaohns, it i« n^f • ! ^"■^^^o^d species anr^^o • samp f,-rv, . °* ifnprobabJe tlinf +k ^PPears in Mrated than half^l '"'''°^°' »''.oonseq„ent]y!Te| ■*i ir'J^ii 870 A NATURAL SYSTEM IK MINERALOGY. [VIII. have noticed in the other sub-orders the clo';e relations between the ophitoids and pinitoids and their correspond- ing phylloids. Beginning with the most basic of the clays, we have given in Table XIII. their atomic formulas, with the values of P and V, so far as these can be determined. § 100. The genesis of these persilicates, whether phyl- loid or colloid, here demands consideration. The sub- Table XIII, — Argilloids. Species. Formula. P D V Schrotterite. - - (al48ii)05 + 5aq- - 12.80 2.15 5.95 CoUyrite. - - - (al38i,)o4 + 4Jaq - 12.53 2.15 5.83 AUophane. - - (alsSijK + eaq- - 12.27 1.89 6.49 Samoite. - - - (aljBisK + Saq- - 12.81 1.89 6.66 Halloysite. - - (alsSiOoj + Saq- - 13.80 2.40 5.75 Kaolin. - - - - (aljsi jo, + 2aq - - 14.33 Keramite. - - - (al28i3)05 + 2aq- - 13.85 Wolchonskoite. - (crjSisK+Saq- - 15.33 2.30 6.66 MontniDrillonite. (alisi2)03 + 2aq - - 13.00 2.04 6.37 Chloropal. - - - (fiiSijK+lJaq - 15.51 2.10 7.38 Cimolite. - - - (alisi3)04 + laq- - 14.20 2.30 6.17 Smectite. - - - (aliSi4)o5 + 4aq- - 12.55 2.10 5.97 aerial decay of the aluminous spathoids, orthoclase and albite, is apparently the direct source of ordinary kaolin, for which the ratios of protoxyd, alumina, silica, and water are 0:3:4:2. The derivation of this from feldspars, having for the same elements the ratios 1 : 3 : 12 : 0, is due to the loss of all protoxyds and two thirds of the silica, and the hydration of the residue. The adamantoid and phylloid aluminous protopersilicates are not generally subject to such transformations, although Damour has described a silicate with the composition of kaolin, de- rived from the decay of beryl. It is important, in this connection, to study farther the sub-aerial decay of other obsc the yiek little § neces do vvi the ca ^hich VIII.J "*' SILICATES. formation t:^°;^^ regarding th,^^^ , ^ '^' ^* sometimes rpt; x, ^® ^^cay of a «. , ^^^^au in JfaoJin of ohT y *^^ Process of it? ' '^^'^ ^^^^e i„ it. """ina, silica a"d I ' """ "^ albitlhl-/^' <"■' « we l^uchs, not • 3 /'!?"; according to th^' » , "-"^Jting ""'•■' 85.0, „,f,; * • 2. but .. 2 :V- 2 /- "^ ^^« "^ ^of "amed clav mat 1 •®^- This distin.f ?'"=" *«-4' alu- ■« resulting ^^^l^/^a Jed "teramt ""! '^"^ '""""•'o un- ^ ^™«ar JcaolSf.' *' '''»««« for wS "" '" """Po- ' removal with °, '^"''o" of labradorif "'* ^ •• 3 : 6 -ft ™'f '^d feldspar ;,.°V"^* " *'aS™ " "'/''^ ^■"«''- earlier er,.3taC' ^'2' '"""ttiea 0?"^ 7 °^ *«" and *'«= soluble matt., ' "'''' ™Portant Z ■ ^'"^' 'n 'he «'«oates, both of tf •'^"°^^<' »d the ri .' '", '^'"''"n to - »n«derable nar. I?*' '" P^^t ages "'T'l"''' "'"""inons "« «*emS ™'^,^°''»i^ation of fe„tite ' '"'' '^"''• *J"= crystals ott ■/ "''^ «'»wn bt R ■'^"■'''iabie in «°«"a Monflna il'if/^d iaoChJe »r,*'"^ *''-t "'■'a silicate, hav „ 1^' ""^'^t of ab T"^ ^"""d at "''^^'vation the tS f" """'PositL o/t'T ^"''''-'"- *''o same locality ^'' '''" "oniirmerl , f """'"'to- This ■y'-^Wed hi^i^V°^''Wte,andIr;bv '^''''•y^tolfro,, tie case of cryste " ^ °"" "'"' "''Sillowt""' '"^ to ^ ^- VCrated than' „;:" P^o^ •1 {I. m m n;?: ■tfil * B '"I if 372 A NATURAL SYSTEM IN MINERALOGY. [VIII. The origin of the more and of the less aluminous species, like allophane and schrotterite, on the one hand, and like montmorillonite and cimolite, on the other, remains to be discovered. These seem, for the most part, to be, like halloysite, true colloids, and their separation from aqueous solution is apparent from the occurrence noted by Dau- br^e of an amorphous halloy site-like matter deposited by the thermal water of Plombieres, in France, which is prob- ably identical with the saponite of Nickles, a highly hydrated silicate, more silicious than halloysite, from the same thermal spring. These aluminous silicates, like other colloids, such as opal and bauxite, are probably capable of assuming a soluble modification, and have all been deposited from solu- tions. That such a dissolution and deposition take place on a large scale is apparent from the existence of the so- called indianaite. This name has been given to a material found in the coal-measures in the State of Indiana, where it has been mined and employed to a considerable extent in the manufacture of potter's ware. It occurs in irregular beds, often several feet in thickness, beneath a stratum of sandstone, and 13, associated with and overlies limonite. It is often translucent in aspect, with a conchoidal frac- ture, and has all the aspect of a colloid. In composition it is somewhat more basic than halloysite, the atomic ratios of alumina and silica being about 6 : 7, and may, perhaps, be regarded as this species with an admixture of allo- phane, translucent masses of which, in a pure state, are found imbedded therein. Indianaite contains about 23.0 hundredths of water, which, after a long exposure to a temperature of 100° Centigrade, is reduced to 14.5.* The species wolchonskoite and chloropal, which we have placed in the table near to keramite and montmorillonite, show that chromic and ferric silicates present similar con- ditions to the silicates of alumina. Hisingerite, according * Reports of the Geological Survey of Indiana, 1874, p. 15, and 1878, p. 154. VIII.j A CLASSIPrOATroX OP sarCATES. to "~' ^^8 S.1 cate, at the end rf t,l t," ''''' '^™P«=»' ^"0 of Per- classification to the natm^al s Ctl ""/f'*'" "'""'"^--^l to show how the three ub ni? ■ """* ''"™ endeavored onche„,ioal and genetic So„tl ,'"'" "'""'' *''^^« '"V, d.vided into Ave Lbe by Ph*' •;' .T'"'"'' ''' ^^^^ -ore or less completely in each , '';*™«=«^ repeated type is naturally separated f "."''/"''-"'■'ler. The spathoid hy-K'ated, oonstUuZ" tte , ! ir T-'.^ ""^highly rented in the first and^se ond of Sr °^ *"''«»• ^re je have called the pectoltoWs Ld te f'^'^f ''"' ^^ "*"* decomposed by acids, even i„ tl!. ''°'''°"J^. both readily ^iieions species. The ot W ! ^^ 5 *« ■»"« WgWy smaller portion of ombi d T!' "''^''™"'' "' ^"h a two hnndredths, inclJes th„ "'' ™'''"'» "™^ one or spathoid tribes/in tl: Tatt'oT^rrlf" "'"• P^^er' spee,es resist the action rf acidT t?'.""™ ^"'"'o-s and spathoids in the first tw\ ^1" ''y<''-°^Pathoids exceptions, which have been nl^T'''' ^"■■«' =* fe'- a larger volume than tl!! '^''^ ='S«c i" having tribes. ">'™ «'« species of the succeedinf The adamantoids in tl>» tu are distinguished by their (.rr" '"''-<»'^c™ of silicates condensation, and, save in tb' "■"* *^'' »<>lccu lar Protadamantoids, by their resLr f '"" """" '^'^^ proportion of water entei into S ""''''• ^ small ' of these species, not onl^ TV!^' «°™Position of many a«d beiyl, and even in ep do e and? "'■,^"' "' ^^'ase and in malacone. It i, to belted T'".f"'' '» '^^^'o Protoperadamantoids are not so ftr* ! ' "' ""' "'""-'"io apparent ej^ception of garnetW ",''™""' Cwith the " Igneous fusion ; but, 0^0^^''"*'? ''^ "°««"g from heat, even below their me tL °" f^' ''f ">^ "ottn of ^O-ge. shown by a dimi^'^ZT^^Ct^^^^^^^^^^^ 874 A NATURAL SYSTEM IN MINERALOGY. [VIII. mil ,1,-. 'Iir-M:' ceptibility to the action of acids (§ 86). The atomic vol- ume of the adamantoids, while iii certain cases not very far removed from that of spathoids, is ahvaj's less, and in the harder or more gem-like species indicates a great degree of condensation. These characters are especially marked in the peradamantoids, which include the silicates of the lowest known atomic volumes. The great phylloid or micaceous type is, like the ada- mantoid type, represented in each sub-order, and ap- proaches it in condensation. The phylloids in the second sub-order, where they are most largely developed, include both anhydrous and hydrated species, and in the less pro- tobasic forms exhibit much of the same chemical indiffer- ence to acids, and to atmospheric action, which distinguishes the aluminic adamantoids. The species of all these tribes are crystalline ; and the system of crystallization to which they belong has, wherever known, been given in the pre- ceding tables. § 103. The colloid forms of matter which appear in each sub-order, and which we have designated respectively ophitoids, pinitoids, and argilloids, are, as is well known, generated both by aqueous and igneous processes, and hence include, as might be expected, both hydrous and anhydrous species. Those formed either directly or indirectly by the igneous method are necessarily very indefinite in composition, being volcanic glasses or the results of their hydration. The tendency to chemical change exhibited by colloids was insisted on by Graham ; and, in view of this characteristic (as shown by Bunsen in his studies of palagonite, which is readily transformed by heat, in part, into a crystalline zeolite), the writer has elsewhere spoken of this hydrated protopercolloid as a mineral protoplasm (ante, page 188), a designation equally applicable to other colloidal silicates. Of these, serpentine gives rise to various crystalline species, often hydrated, such as chrysotile, marmolite, talc, enstatite, and chryso- lite, which are generated in its mass by aqueous action, r?\M VIIIJ ^ -^A^Sr^ATXOK OP saicATKS. While by iffneonQ f„ • . ™ solution r ' ''P"™'^ '"" "'e 00] oM """"'' ""^ *athU :/eraS.:^'' "^ ""^ ^eo ^ri:; *"« taoliriite anrl .. ^^^^spars, pass into nhvli.- 1 ^^"^^^ous ^Oa'u^il t SKf • <"• into adaS; ; ^ ".^^ change of rprfn- /^ cyanite, while ^,,^ ^^^^^^ lite -"^ of the -b.rde.Xt .S;''- 7^^--*. -ve to di.„e ^ It m W i 'i\\ 876 A NATURAL SYSTEM IN MINERALOGY. [VIII. tribes re-appear in the nonnsilicated oxyds, and serve for their classification. Reserving for another occasion the details of classification of this great order of Oxydatks, we may note that, while the Oxadainantoid tribe embraces such species as periclasite, chrysoberyl, the spinels, mag- netite, corundum, diaspore, hematite, quartz, rutile, cas- siterite, etc., the Oxyspathoids include cuprite, zincite» crednerite, pyrolusite, tridymite, and senarmontite, and the Hydroxyspathoids, gibbsite, gothite, and manganite. Among the Oxyphylloids are brucite, pyrochroite, massi- cot, minium, melaconite, hydrotalcite, and pyraurite ; while the OxycoUoids or Opaloids embrace bauxite, lirao- nite, opal, uran-gummite, and eliasite. The plan of the present essay does not embrace a dis- cussion of the species of this order ; but it will be advan- tageous, in connection with the history of the silicates, to notice some facts regarding the atomic volume of certain of these oxyds. The adamantoid tribe of the Oxydates includes a large number of species crystallizing alike in the isometric and the rhombohedral systems, which give for V a vp^ue approximating to that of the adamantoid silicates, chrysolite, pyroxene, garnet, epidote, beiyl, and tourmaline. Ch. Gerhardt, in 1847, published a note on "The Atomic Volume of some Oxyds of the Regular System," which was translated and given in English by the present writer in the same year.* Therein accepting the view held by Laurent (§ 27) of an indefinite or un- limited divisibility of the molecule, Gerhardt proposed to reduce to a common formula, MjO (maOj of our present notation), not only protoxyds like periclasite and proto- peroxyds like the spinels, magnetite, chromite, and frank- linite, but sesquioxyds like martite, hematite, and braunite, and titanates like perowskite and menaccanite, including thus not only isometric and rhombohedral but tetragonal * Ann. de Phannacie, 1847, xii., 381-385, and Amer. Jour. Science, 1847, Iv., 405-408; see a,\so ibid., 1852, xiii., 370-372, where the same sub- ject was further discussed by the present writer. Vlii.j value of V ;,?'■"'»' -ith o„.. et, ^ ' -'• l"'""''. divided "scribed bvr ' ^•^<' *" 5.70 t '""•S'voforthe conceived tliat t„,T "^ various oxvd, ^ ^'"S '"'" « oorreetly d 111"" f "'^^o ^PecioM 7"-"' "'"' ''« over, poinded ::'™»;,^ -<>""' be ti,e t™' f"",7'"->.e. ^"'cite and Sk v'?"' "'"' "''"cS „t '?"'''"' '<"■• ^d ""mpared by )„„'*' »"'■■'' or less conder..! . ""'''«- § loe- Near t , "'^^ "»'« °.''^'<'3 thus studied bycT'"" *« 'he „eat „ to titanic oxvd »?« ,'5'^^'Sf'>es V.S.K ■^' "'"'« the orthorhombfc h/^ ?'"' ^°'' "'« tetrac^onff' """? "«« Cwhich i"'l„ '''''"S™"! octohecMt! v' " ^'"^ ^r V of ""'ecnle t nn T / ^ ^' ^-S"- A stT , "''^^"e'-s to Vdroua Tdatnt;:'* "' '•»''' ^'-s V'" ,f '"^'ohedra, '"•ke its maxim? ^a^Pore, orthoriiomi • '' '"''"« ""= fo' V, 4.2T Tart r™" ^f^-flo gt;; '" °™. if we ,t.' 878 A NATUiiAL SYHTKM IN MINKHALOOY, I VI II. bucholzito, and zircon (V = 4.80-4.90), while casHiterito and (lUiirtz are near to the spinel ^roiip, and to chrysolite, pyroxene, garnet, epi(h)te, heryi, and lyncurite. Corini- dnni, dias[)ore, and clirysohery 1 stand apart from all of these as having a more cotulensed molecule than oven cyanite anil xenolite, the most highly condensed silicates known. § 107. While small ditl'erences in atondc volup.e nuiy, Jis (ierhardt insisted, he set down to impurities and errors in determination, a careful survey of many silicates* and oxyds leads to the conclusion that among these there are great groups which, essentially agreeing among then:selve8 in molecular condensation, differ in the value of V from other groups by (luantities less than thos? admitted as accidental variations in volume in the large series lirought together by (Jerhardt; which may thus very well be found to include two or moie distinct grou{)s with unlike vol- umes. At the same time, the comparisons which we have here made among the adanumtoid oxyds, not less than those among the various tribes of silicates, serve to strengthen the conviction that the accident of geometric form, liowever valuable as a means of diagnosis, is of alto- gether minor importance in investigating the general rela- tions of mineral species. § 108. The metals proper, together with the bodies of the sulphur and the arsenic series, and the various binary and ternary compounds of all these, make up the great natural order of Metallates, which include two sub- orders. Of these, the first, or Metallometallates, di&tin- guished by opacity and metallic lustre, is divided into six tribes, which are: 1. Metalloids, — native metals and metal-like elements ; 2. Galenoids, — argentite, galenite, bornite, chalcocite, metacinnabar, onofrite, stibnite, etc. ; 3. Pyritoids, — pyrite, linmeite, stannite, chalcopyrite, pyrrhotite, etc. ; 4. Smaltoids, — smaltite, niccolite, breit- hauptite, with other arsenids, antimonids, etc. ; 5. Ar- senopyritoids, — including arsenopyrite, cobaltite, etc. ; 6. Bournonoids, — enargite, bournonite, zinkenite, etc. "'■ "''"»n.nti„e i„ """"' ''l""-'i<-'» ".ore „r I. "''"""' Christ,, ,^i/„ »"'^''"'''". a,„l i„cl,„i„ '"'■""'■'*•- ''"no- typ.cal forms snht .f ''""""""tallates -^^ '"''-"'•athoinetallates. II. 2. OxYDATES. — 3. Silicates: a. Protosilicates ; h. Proto- persilicates; c. Persilicates. — 4. Titanates. — 5. Nio- bates. - -• (5. Tantalates. — 7. Tungstates. — 8. MOLYBDATES.— 9. CHKOMATES.— 10. VANADATES. — 11. Antimonates. — 12. Arsenates. — 13. Phos- phates. — 14. NiTKATES. — 15. Sulphates. — 16. BoBATES. — 17. Carbonates. — 18. Oxalates. III. 19. IIaloidates: a. Fluorids; 6. ChlonJs; c. Bromjds; d. lodids. IV. 20. Pyricaustates : a. Carbates; h. Carbhydrates. § 112. The conceptions of high molecular weights in mineral chemistry, and of the existence of compounds 4. Rhetinite (Resins); 5. Paraffine ("VV^xos). The silicates of heavy metals are placed in Class III., but with tu?se exceptions the Sklerite include the spathoids and adamantoids, the Phyllite the phylloids, and the Amorphite the colloids of our three sub-orders of Silicates, arranged in part crystallographically and in part chemically. The Oxydates are divided between the Kuphoxyde and the Metalloxyde, or those of the lighter and the heavier metals. In Halometallite are found the silicates of yttrium, zirconium, thorium, cerium, zinc, copper, iron, and manga- nese, together with niobates, tantalates, tungstates, chromates, as well as arsenates, phosphates, sulphates, carbonates, fluorids, and chlorids of the heavy metals; the corresponding compounds of the lighter metals coming under the order Apyritite of Class II. In a second edition of the Synop.sis, in 1884, the Halometallite are made an order in a new class, called Mctallollthe, or Metalstones; while the order Asphaltite is divided by separating from it the order Elaolte (Petroleums). In the order of Thiometalle, the family Pyrite includes alike the pyritoids, smaltoids, and arsenopyritoids ; Galenite, the galenoids and bournonoids, and Cinnabar- ite, the sphaleroids and proustoids. From the point of view chosen for our essentially chemical system it seems unnatural to place in two distinct classes analogous and closely I'elated oxyds, silicates, carbonates, etc. Again, the different degrees of condensation, as shown in atomic volume, and in the relations of this to hardness and susceptibility to chemical change, which underlie the dis- it'fi I'- Jji-iiii msmwm ""'^r species, most of ttr •,•'" ^''^''^'^ Bre^h^ ''' f''^"" '«'<->', S 6;f '»i'^ 384 A NATURAL SYSTEM IN MINEKALOGY. [vm. MJ h . ill vs him, in his paper published a few months later in 1853, on the Constitution and Ecjuivalent Volume of Mineral Spe- cies (§ 17-18), to suppose the existence of differences between the volumes of isometric, rhombohedral, and various prismatic species. This notion was, however, soon afterwards discarded, as may be seen from the citations from his paper of 1867, already given in § 12-14. § 113. In further illustration of the supposed relatiors of density and equivalent, the following additional passage is quoted from the paper last cited : — "There probably exists between the true equivalent weights of non-gaseous species and their densities, a rela- tion as simple as that between the equivalent weights of gaseous species and their specific gravities. The gas or vapor of a volatile body constitutes a species distinct from that same body in its liquid or solid state, the chemical formula of the latter being some multiple of the first ; and the liquid and solid species themselves often constitute two distinct species, of different equivalent weights. In the case of analogous volatile compounds, as the hydro- oarbons and their derivatives, the equivalent weights of the liquid or solid species approximate to a constant quan- tity, so that the densities of those species, in the case of homologous or related alcohols, acioirf, ethers, and glycerids, are subject to no great variation. These non-gaseous species are generated by the chemical union or identii'ica- tion of a number of volumes or equivalents of the gaseous species, which number varies inversely as the density of these species. It follows from this that the equivalent weights of the liquid and solid alcohols and fats mupt be so high as to be a common measure [multiple] of the vapor-equivalents of all the bodies belonging to these series. The empirical formula CimHuoOij, which is the lowest one representing the tri-stearic glycerid (ordinary stearine), is probably far from representing the true eijuivalent weight of this fat In its liquid or solid state ; and if it should hereafter be found that its density corre- i> """'' THE QUESTION OF MOLECnr . MOLECULAR WEIGHTS QCR sponds to six f § 114 In the papers of 18« !u*'"' ™1""- C.H,0 .." te7r.;;:::;ri^.^ those of fK*'^'^" = 2500, and C nrA ^^ '"^^PeC" ,,,?r^ "'oes which belong f ' "ff'ier (a„d as yet u„! untouched. These silicates "nd u""" ''""'«'' was le t ^"o«., incapable of a,rcheli r'*""-^P"^ are,so fct Pos..o„s but such as IffecHr ""'"'''"'''»"«»'■ Jecir ";ei.-moleeu,a"v2ht:T/r"'" "'» S^-t ptb^; whJe, ],oweve,, the^it'^^t"' T"''" »'■''""- ^an favor of high »olecu]afwe Xtf ""1 ^^ '"' '» l^Sat l-y discoveries which serve f„, ™ '"*™ strengthened "' "'" ' '"""-» '^-Tal^Crtfr V' "P'^'™'^" ■om the chemistry of tlie . I] ' N 386 A NATURAL SYSTEM IN MINERALOGY. [VIU. h m carbon series. First among them may be noticed the various artificial crystallized cobalt comi)uiinds, such as the potassio-cobaltous nitrate, to which wirs assigned a formula with 2Co, 6K, and 12N, and a unit-%v^eight of 968. More remarkable still are the ammonio-cobalt bases, stud- ied by Fr'«idered a," '"^''''''''gstates til > ^* "<'»' rec f ^;nff term " bZg^'X" T"""^^''^ ^Z 'nk l'"^"''"- Pliowc acid in ti ! ^ ■""•>' also take thl f'"'^P'""-ous «rge„ in „,1'^^« "ompiex salts anV''" P'!''=« of phos- -" phenyl! appelttrt"" ™*«e ZhW "^'^"'' *h^ oompound^f^ '° '" "-^ capable of Z,^!^'""'*''^' ^ "7- Tiie sii;„ . "^ ""» the the subject of fa tt S'*«tes of Ma„Vnac h '"und that the L'jf"', ^^^ralization by G»r '"'° ''e™ by the oxyds J^ !"'" "f ««ea therein ^^' '""^ " is ™°'jbde„„rGTh, '! *' »".po.,nds of : """^ »"'" "f;.fr<"» vana lic^S t *■""'"' -vctl "Iv T^^'^" «"" which he r.„. " phosphoric m- , . '^''nes made j"«% remarks-T" '^ ''" such complex , • 388 A NATURAL SYSTEM IN MINERALOGY. [VIII. .1 > h ! i 1 3 t ■ ;» as in the case of most inorganic compounds, entirely- unknown." He adds that the progress of bcience " tends constantly to show that the structure of inorganic mole- cules is more complex than we formerly supposed," * and illustrates the great complexity in these compounds by a phosphotungstate including vanadium and barium, repre- sented by the formula 6OWO3. 3PA. 2V2O,. VO^. 18BaO + I44H2O, and having in his opinion "the highest molecular weight yet observed, 20,058." He describes another simi- lar compound, 6OWO3. 2,^,0,.Y,0,N0z. 18BaO+150HA of which he says, " It is almost certainly a double or triple salt, but it still shows how five different oxyds may exist in a single well defined and beoutifully crystallized com- pound." Besides these soluble and hydrous species, all produced in the moist way, is the curious gold-colored insoluble anhydrous crystalline body discovered by Wohler, which is formed at a red heat, and is generally described as a tungstate of tungstous oxyd and soda. This, Gibbs suggests, may possibly be represented by I6WO3. 4WO2. 7NaO, which corresponds to a unit-weight of 5002. § 119. The researches of Gibbs upon these complex inorganic acids, resuming, extending, and generalizing those of other laborers in the same field, are of much * Wolcott Gibbs on Complex Inorganic Acids; Amer. Jour. Science, 1877, xiv., 61; also, in abstract, Proc. Brit. Assoc. Adv. Science, Montreal, 1884, p. 667, and more fully in Amer. Jour. Chemistry, 1879-188.3, i., 1, 217; ii., 217, 281 ; iii., 317, 402; iv., 377 : v., 361, 391. [See farther ibid., vii., 392-417, wherein are noticed the pyrophosphates of Wallroth, such as Cam. Nai,. (P207)9, etc. It would appeal*, says Gibbs, that complex molecules similar to these "enter directly into combination with twenty-two molecules of tungstic oxyd " to form the still more complex pyrophosphotung- states. He also notes the molybdates described by Struve, in 1854, in which aluminic, chromic, ferric, and manganic oxyds are united with molybdic oxyd, and give such salts as AI2O3. i2M02. 6K20-f- 20Aq; and MnoOa. I2MO3. 5H2.0-l-21Aq; and, farther, the salts of Parmentier, AI2O3. IOMO3. 2K20+15Aq, etc., as additional examples of salts having necessarily high molecular weights.] VIII.] THE QUESTION OP .xOLEcrLV,.^ ^*.CCJLAR WEIGHTS. 889 «iffnifica„co to the h ■ ^^^ "■organic compou„dr •! "^'''"' ^*™'=ture of so.,I ^ would notonh-.ll . '""^ affirmed that th ^"""""^d b'-t would "be fo f '° ' """'"'t >»ine " otf ' , " ^"''' "'■enn-ea. tie .ter'i/" f -'^e and ."; rlT,'^'»>: o^pnic chemistry which 890 A NATURAL SYSTEM IN MINEUALOOY. [Vlir. ( » M 'U I t 5 '^ '»( needs not fear comparison with tlie order wliicii reigns in the organic! brancli of onr science." lie ad-.is, ''It is well to be reminded that complexity of constitution is not the solo prerogative of the carbon compounds, and that before this systematization of inorganic chemistry can be effected we shall have to come to terms with many comi)ounds concerning whose composition we are at present wholly in ignorance," and by way of illustration refers to the com- plex inorganic acids of Gibbs.* § 121. Ilecognizing from the beginning of this inquiry, in 1853, that the molecular weights of mineral species, while far exceeding those of hydrocarbonaceous or so- called organic liquids and solids, are equally unknown, we have sought, nevertheless, to show the comparative condensation in different mineral spe '.ies, and at the same time the existence of homologous series among them, by the use of atomic formulas. In these the results of chemi- cal analysis are reduced to their simplest term, and are presented independent c»f all hypotheses as to the struc- ture or the molecular weight of the species. These for- mulas suggest to the chemist something more than the elemental atoms represented by the symbols employed. While he admits in the simplest mineral silicate or oxyd the existence of oxygen, silicon, and one or more metals, all being chemical elements physically dissimilar to each other, and to the species before him, the tendency of the mind is to conceive this as made up of identical or of similar units or individuals. The justification of this mental process appears in the fact that it is in the com- parison of such individuals or chemical units that we find the chief data for the intelligent study of the chemical species. Such a conception of units underlies the doc- trine of polymerism, and that of homologous or progres- sive series, and enables us to compare silicates, oxj'^ds, and carbon-spars in a manner the correctness of which is * Sir H. E. iloscoe, Address to the Chemical Section of the British Assoc. Adv. Science, Montreal, Aug., 1884, Report, p. G63. (i': \* VIII.] T«K QTJESTION OF Mr»r 881 ^"■■"i"'! by the nl '"'-'""■'•■''• 8 "f ".-e ' If"' '"■'•*■■-' ' e C r "■ "- »t,.,,y „ °^«en. M , u ".">".. "Ht, o.juiv,.^":"';' "' " ■"""■ "■'"to! ntonte ''""■• T"'""' "•Zo.^'lrT"""'' divide tI,evoTr7' J' ''"^ «»' »l«o,l " "'"■'• ''"^ ■ P't """ivMuul, ,S """."""ience sugL ts /l ■ ""• '" «>■* tl,e sim2,f ''*'"' ^^ "" mi. eru r' '"'S'^' ;"*oated, the ^t /:r'.'^'""« "-uW u,"^'"'^ ""O for eiich specific 7 ^'^P ■" our innuirv „ . * "" "l^ove designated p.rf '^'S'"; tbis mean^w f/' ™''"™d ■■elation of th, oh^'"^' '»'"' to be nT^l* ''"^ ^-'en "« .'exus be tet ";r' ,"'"' 'o space a Sf''^"'^ '"e ^i««Mo gravity „?'r;*"S^ »« mean „t' f ^T^ '• '''"d '^ f. .ted by D^ "!r ''" species Ovate,- be ,t ^'^'" ''^ "'« designated as th ^ '""""'y ti,.« ,,? ? ""'W' ''epre- hypotJ^est T "''"""-n- But til T'""'"^ '" "'hioh attained bv . • ""' «o far „. ,, ' ,"" e^l'ression is sponds to tit !!"';"*■' =■« "uitv tte n ,'""''™' o«ly be •^ stui be some ■f.l"'' '■' (55 89:i A NATURAL HYHTEM IN MINKUALOdV. [vm. lil! II h \ m m 1 ■ < ■ multii)lo of this qiiiiiitity, luid will, iit the siime time, be the coinnion imilti[)le of all the utoinie volimies tletluced from various chemiL'al units. § 123. Ill ai)i)roac'liiug the consideration of this molec- ular volume, it may he noted that while in salts of the same ty|)e the specilic gravity sometimes rises with the molecular weight of the base, as when zinc replaces magnesium, or lead replaces strontium, in carbon-sj)ars, the specific gravity of double or triple salts is essentially the same as tliat of the corresi)onding salts with a single base, as may be seen by comparing the densities of sim|)Ie and double hydrous sulphates, orthophos[)hates, and tartrates; so that the value of P, deduced from the more complex salts, considered as chemical units, will be essen- tially the same as that of the apparently simple salts of the type. Taking, then, of the anunonia-cobalt salts (§ 115), not the simple chlorids, but Braun's complex phosphate of luteo-cobalt, with a unit-weight of 2540 (of which the specific gravity is undetermined), we have, if we assume for it a density of 1.701 (which is that of the chlorid of the same base), a unit-volume of not less than 1493. § 124. The complex tungstates give still higher, vol- umes. The golden anhy ^rous tungstico-tungstate of so- dium has a specific gravity of 6.617, while for two allied tungstic compounds, the one potassic and the other sodic, are given the numbers 7.60 and 7.28, showing that these are similar in condensation to the anhydrous calcic and ferrous tungstates, scheelite and wolfram. For the solu- ble and hydrated polytungstates, Scheibler found the specific gravity of 4W*O3.Na2O+10Aq to be 3.987, while that of the corresponding barium salt, with 9Aq, is 4.298, and that of I4WO3. 6Na,0+32Aq is 3.846.* The density of the complex hydrous phospho-vanadio-tungstate of barium described by Gibbs (§ 114), with a unit-weight of 20,058, is unknown, but, if we assume for it the number * See Constants of Nature, by F. W. Clarke, Tart i., 83. vni.j '"" ■•'••"lily uttm- , 1 "•■'«'"« ■'« timt 1, ' ' <="">«'>rml mineral spel ' "' '"o ''"o ..mue.l ' '""'""•'^l'- 8«itcd bv , . ■"'^'' '^"t their vo n """"»»"> molec- '>•!* a eorri" , '""'" ■""""'« units 1 "' "'°'^«"!e three nf fi. formula for fi,„ "• ^"t as "^^e manner th^ 1 ^^o^^astonite hv OQa/ '''^^ V by '-^"^^ albite 'nn '.'"""^ ^"'"^«e of the L? ^^^'^'^^'^o^- J" '?*'?'"""• •' "• ""=""■ ==.'"5 -tt Will be evifipnf- +1, x «- ui "' that attempts lifce these at , , ^nese at moleeujar 394 A NATURAL SYSTEM IN MINERALOGY. [VIII. Sill li i: foi-nuilas are of value only so far as they serve for illustra- tion, since the unit-volumes assigned to the various spe- cies are but approximations, and the molecular volume, 4666, wliich has been assumed, is based on a supposed specitic gravity, and can only be conjectured to be not far from the truth. A series of careful studies of the specific gravities of various salts of the complex inorganic acids may furnish us with more trustworthy data for sindlar calculations. Meanwhile, it is to be repeated that the formulas here given for pyroxene, wollastonite, and the feldspars, are of value only as they serve to illustrate our conception of the complex constitution of these silicates. For the purposes of comparison, and for the elucidation of polymerism and homologies, the unit-volumes which we have calculated for the preceding tables of species of the different tribes of silicates, serve every purpose, and show in a simple manner the relative condensation of tlie mole- cule in the various species. Attempts to devise structui-al formulas for these very complex silicates appear, in the present state of our knowledge of their constitution, to be premature, and, at the same time, unnecessary. § 126. We have seen in our studies of the volume of mineral species two cases ; the first being that in which, in analogous compounds, the density varies with the unit- weight, so that the species compared have identical unit- volumes ; and the second, that in which species, otherwise analogous, have such densities as give very unlike unit- volumes, — a fact showing the existence of progressive or homologous series of polymerides, as illustrated in the case of many silicates and carbon-spars. Examples of these differences are seen in the chlorids of potassium and sodium, the latter of which itself presents remarkable differences in density. Thus, while the numerous deter- minations for potassium-chlorid do not vary very much from a specific gravity of 1.99, the careful observations of different experimenters with sodium-cMorid show varia- tions from 2.011 by Playfair and Joule to 2.15-2.16 by f . \k ■V .^«^.r-.>*.,^^,M^.)..^,f|^ (jy-gjp,,.^,,.,- VIJI.] THE QUESTION OF MOLECULAR WEIGHTS. 395 I ?!' Stolba, 2.195-2.204 by Deville, and 2.24-2.26 by Mohs and Filhol.* With these various determinations of density before us,, says Henry Wurtz, "we are forced to infer the existence of four modifications of sodium-chlorid," while he adds, "common salt is far from being alone among saline combinations in its passage into livers modifications or allotropes. On the contrary, the circumstance is almost universal among salts, throughout the whole range of chemistry." It would seem, in fact, that such variations in specific gravity in a homogeneous solid (like those in the specific gravity of a vapor or gas at constant pressure and temperature) can have but one meaning ; which is that these sodium-chlorids of different densities are so many distinct species, related to each otlier as fibrolite to cyauite, as lyncurite to zircon, and as tridyniite to quartz. § 127. All such allotropic variations in compound spe- cies, which are marked not only by differences of density, but in many cases, if not in all, by differences in hardness and in chemical relations, are by Henry Wurtz conceived to be "dependent on a variability through a certain (sometimes not very narrow) range of diameters, of one element, always the basylic or electropositive of a group, — in salts, therefore, always the metallic base." Accord- ing to him, the volumes of elemental molecules, that of oxygen excepted, " are expressed by quantities having, at the temperature of ice-fusion, the relations of even cubes of a series of whole numbers, of which series the number pertaining to the molecule of ice at this temperature is 27." This, he adds, is "a standard volume in nature, to which the volumes of all liquid and solid bodies may be compared when at the same temperature." The cube roots of these numbers are by Wurtz designated as " molecular diameters," and the variations in specific gravity in the different forms of sodium-chiu::id are explained by supi)os- ing the diameters of one or more of the sodium molecules in a complex group including 4NaCl to vary from 23 to 24 * Constants of Nature, by F. W. Clarke, Part i., 30. ! 4 396 A NATURAL SYSTEM IN MIKEKALOGY. [VIII. /■!; h '. t and 26, the diameter of the chlorine molecules remaining invariable.* This method enables him, by admitting more or less complex groups, in which the similar elemental moleciiles have varying diameters, to approximate closely to the densities of liquid and solid species. § 128. To such a scheme it must be objected that it involves the notion of existing elements or groups of elements, dissimilar to each other as well as to the spe- cies under examination. The conception that the chemical elements enter as such into combination, and there retain their volumes, is, it is believed, inadmissible in chemi- cal philosophy. The view which I have constantly main- tained, and have set forth i,i the present essay, it^ that differences in density, such as we have just considered, are not dependent on variations in the hypothetical units adopted for convenience in calculation, but belong to the species as an integer, and correspond to a greater or less condensation of its mass, — that is to say, to the identifi- cation in a constant volume of a greater or less number of chemical units. The very terms of atom and molecule, which we apply to these imaginary units and to the mass, are concessions to a popular terminology borrowed from physics, and are not only inadequate but to a certain ex- tent misleading when applied to chemical operations. I venture in this connection to reprint the words employed in 1874 1 in the discussion of this same question : — "The phenomena of chemistry lie on a plane above those of physics, and, to my apprehension, the processes with which the latter science makes us acquainted can afford, at best, only imperfect analogies when applied to the explanation of chemical phenomena, to the elucida- tion of which they are wholly inadequate. In chemical change, the uniting bodies come to occupy the same space * Geometrical Chemistry, by Henry Wurtz, page 72, 1876; reprinted from the American Chemist, March, 1876. t A Century's Progress in Theoretical Chemistry: being an address at the grave of Priestley, July 31, 1874, reprinted from the American Chem- ist for August and September, 1874. See, also, ante, pp. 13-15. ^"^•^ A KATITBAL SYSTEM X^ ,„^,, a'^ the same tinie an.1 fi • ^^^ seen to be no loi,p-Ir ^ '^ ""Penetrabilitv nf "masses is confS V' ^"'^ ' ^^^« volume o/i, '"^^^^^" ^« cal characte tS^^ ^"^ "^^ ^'- Pi^yX:/.^;^ .t'"^^"-^ : «i* bids t:d7„\:t ^"T' "' ••"- an'r:!?, °^ '-ien we give a e<^.o; 1% P'"'" "''^""''''i affinreif"'^^/ nature, and „f t "f^-'^^ of i,^ ^ tl^.'T"""' and faet.' " '^'^ '» distinguish betlel '^'"■^ "^ § 129. We her« * ™njeoture "nd, as a fecessl""""^^ ""^ P'an of ehl'^'T'''- ^'^"« N^*"ai Systr Xr" ,"'^' '°'^- ' "fot Te'"K".''''«=;-" set forth ,-r. ^''"iieraloffv W^ i ® '^''^«is of ;. o^yds, which '!; ,*'"* '""'•» briefly to i, P""«'I''es over, g,-,:: ° ' 5;'^er of MetalUtel^te"! 0/^*"e. " fication whi<.h o ; ' "^ an outline of „ ! ^^''' '"o'e- and already el? ^ Tf "«^ 'anguage "!"'?"• ^' '^ ^ »"ed, .n § 18. that alf eh^''^' '•" 186V. unoal species really 1 : i,M ' i: 398 A NATTJllAL SYSTEM IN MINERALOGY. [Vin. \'l Hi u * \ lit belong to the mineral kingdom, and that, "in this extended sense, mineralogy takes in not only the few metals, oxyds, sulphids, silicates, and other salts which are found in nature, but also all those which are the products of the chemist's skill. It embraces, not only the few native resins ar:l hydrocarbons, but all the bodies of the carbon series made known by the researches of modern chemis- try." A Manual of Mineralogy, based on the principles here set forth, such as we hope to prepare, would, however, be limited to the consideration of natural species. § 130. In conclusion, we give three synoptical tables, in which are resumed, under their respective tribes, the principal species of the three sub-orders into which we have divided the order of Silicates. In these tables the dominant atomic ratios are given in the left-hand columns, while more rarely recurring ratios, as in danalite, schorlo- raite, sloanite, etc., r re placed in parentheses after the names of the species, which are in their appropriate posi- tions in their respective columns. In the case of Tribe 4, the exigencies of construction have caused its displace- ment in the table, and hence the atomic ratios of its included species . are there also appended. The calcu- lated values of V are given with the respective tribes. These tables are necessarily much abridged, and should be studied in connection vdth the systematic grouping of sub-orders, tribes, and spesies, to be found undor § 55 of the present essay. For the colloid tribes the reader is referred to pages 374-375, where the very variable composition of the vit- reous products of igneous fusion is insisted upon. As regards th'> limits of species in such cases, the question which arises is similar to that presented by the various intermediate feldspars and scapolites, already discussed, and by the intermediate carbon-spars, and is one inti- mately connected with the high molecular weights which must be assigned to mineral species. il I'i VIII.J ^ NATURAL SYSTEM IN MiNEltALOGY. 399 'I '■, 400 A NATURAL SYSTEM IN MINERALOGY. [VUI. 'ii;;;*^-::;- B I w I id ■I CO « 1 « n a •saiiaoiH^ •III II I a Cfi A a, P 9 5 M o ^ N »3 ft S"^ 3 a o 5 0.2 •SVOl 1!L II I 8 p5 S s i3 1$ e B V 13 Ho I O 10 'S. «5 5 '3 I saNnvKHao' a •C I §•■ |2 Ml 9 a i I* •a S' a oQ ■ •c 0^ 9 •II •g DO H Q 2 OQ 3 o 1 I a e I ■Si o a B m M i*^y :vm. 1 vm.j ^ ^^"™^^ ^^-^^i^M I.V ,„K.„ liALOGY. I ! » .MW J- 1 I ! ' t IX. THE HISTORY OF PRE-CAMBRIAN ROCKS. The following suuitnary is in good part a condensation from the account given in my volume on " Azoic Hocks," to wliich, for details previous to 1878, tlio reader is rtiiorred. His intended to servo as an introduction to the two succeeding essays in whicli certain parts of the history of these roclts are discussed. The diief portion of the text wiis published in the Anierloan dournal of Science for -May, 18S0. There are, however, many later a■ as : ;:* ™ ^-^ Je..e; h 'i ;t "■"^^^ southern Nevv y ""'"'"'' " S-'oa* part o T'^ "^""^ "^ class he refenwl .11 ., """"■!''''« >ocks T„ *i '''^■ -hat he had dose; ed^a« fti '"'" "'""'^'- ^ S'^ pttoT n his Metamorjihie elass "t7 ™f "°' '° ^' '""'"ded Jiitchell and of r„„i. , "'« subsequent i,v, " f'e views of V " '''''^^' lioivever Pl„n,r *"^ "^ nr»n . ? ^anuxeniand Jf„,t,- ' ™ai'ly established inown tn R- T'"'"' S^"'"™c series in P . Me'a„,o,,,hio ,.athe,. tl!::? 1'°,^;.' "A":^' ' I. .:•'* i •I' 'I 3 / 404 THE HISTORY OF PRE-CAMURIAN HOCKS. [IX. described by Logiin, in 1847, as consisting of a lower group of hornblendic gneisses, without limestones, and an upper group jf similar gneisses, distinguished by inter- stratified crystalline limestones. These rocks were found by Logan and by Murray to be overlaid, both on the north shore of Lake Superior and in the valley of the upper Ottawa, by a series consisting of chloritic and epidotic schists, witii bedded greenstones, and with conglomerates holding pebbles derived from the ancient gneiss below. The same overlying series had, as early as 1824, been described by Bigsby on Lake Superior, and by him distinguished from the Primary and classed with Transition rocks. § 4. Labradoritic and hypersthenic rocks, like those previously described by Emmons in the Primary region of northern New York, were, in 1853 and 1854, discov- ered and carefully studied in the Laurentide hills to the north of Montreal, when they were described as being gneissoid in structure, and as interstratified with true gneisses and with crystalline limestones. In 1854, the writer, in concert with Logan, proposed for the ancient crystalline rocks of the Laurentides, including the lower and upper gneissic groups already mentioned, and the succeeding labradoritic rocks (but excluding the chloritic and greenstone series), the name of Laurentian. In the same year he wrote that, "in position and lithological Qharacters, the Laurentian series appears to correspond with the old gneiss formation of Lapland, Finland, and Scandinavia." * Subsequently, in an essay published in 1856, these gneisses of Scandinavia, together with the oldest gneisses of Scotland, were, on lithological and on stratigraphical grounds, referred to the Laurentian series, and, at the same time, the name of Huronian was pro- posed for the chloritic and greenstone series, which had been shown to overlie unconformably the Laurentian series in Canada. * Amer. Jour. Science, 1854, xix., 195. 1 1' rx.] ROCKS IN NOIITH AMERICA. I';;;' ''oscribc-,1 ,,,,0 ,,„„,:;;;;,';;, ."f • '''"^ter a„,l Whitney ^"l;e.-.or „s co„»fit„ti,Kr „ ' V "'"■"'""" """ks of L,ke «-!„, with g,,.„ia,«, „' " ;;J37"^ »f.' ™' of Meta,„o..„l, c ^f'""? fo a single mC ,yT ' '■"'^'' "' "'» «gion "^^o.v„u,„„ „f Kin,I„, "• ;7^'' " e.-.;.n tic nucleus. The Credne,-, of Uvooks ,.nd i"„ I,'""' ""= '"'<"• studies of of the P,™„.,, ,,,;" I^J^™-^ as also f„,,,»;„ .;■; seen, supposed to be iLf,: "" ™ '"'ve alieadv ?;ooks in New E„g,a„d flT; , ''"'^"^''^ »t,-atu. Th ^ • ■'«« ""d limestones of al tI '''°'""°" "f "'<' 'eissic system stone, which he regarded '" *'"" ">" Se-'litlmi Id' •'r • While he sSs d :L"" "'"'-'"«" of the Po t -"ed by him AzotTo^tTorelt^dTf "^^"•■'^'^ connected stratigraphically 400 THK IIISTOKY OK I'UE-CAMUUIAN UOUK8. [IX. IMii. -m with the base of tlio Piileozoic Hcries, ho ncvertlioless assigned them to a positif.i below the base of the Now Yolk system ; thus reco^'ni/iiig in Pennsylvania, beneath this hori/on, two nneonl'orniable [groups ol" erystallinc roeks. Tiie existence amonj;' these newer crystalline schists of Pennsylvania, of a series tiistinet from the lluronian, and representing the White Mountain or Montalban rocks (the IMiiladelphia and Manhattan gneissic group), had not then been recognized. A farther discussion in some detail of these rocks in Pennsylvania, will be found farther on. Essay XI., §§ 37-42. § 8. 'J'he views of II. D. Rogers with regard to tlie crystalline schists of the Atlantic belt were thus, in effect, if not in terms, a return to those lield by Eaton and by Ennnons, but were in direct oi)p()sition to that maintained by Mather, which had been adopted at that time by Logan and by the present writer. The belt of chloritic and cpi(k)tic schists with greenstones, serpen- tines, and steatites, the extension of a i)art of the Azoic of Rogers, which, through western New England, is traced into Canada (wliere it has been known as the Green Mountain range), was, previous to 18G2, called by the geological survey of Canada, " Altered Hudson River group." It was subsequently referred to th Upper Taconic of Ennnons, to which Logan, at that date, gave the name of the Quebec group, assigning it, as had long before (in 184G) been done by Einmons, to a Cambrian horizon between the Potsdam and the Trenton of tlie New York system. Henceforth the crystalline schists in question were by Logan designated the " Altered. Quebec group." § 9. In 1862 and 1863 ai:)peared, independently, two imjiortant papers bearing on the question of the ■ ''^""o h. called U»cl.iefo,.. The c,.„c.|„ i„ ' A ' ? '," "'' ''"'•^''^W'"- ■n connection with t|,e vie " k ".f "'""" '"-'•■» ""'ice,l, ■ Norway, in •• The Ceoi, J „ e" ,tr ''•'' ,"" "'""' '■"^^'' " eo'^raa^on, between the New E 't' '," '«"3, with fa.the,- "■■'1 the II„,,.„i,„, i,,,^ „„f ; ^ «'"'"! ^OxaUline »el,i,s,„ -.o,nt,,eview»oa,,e.i.^-/---X-^^^^ -'o^-;:'St;s::.:l7'-"'-'^-e,.ie, S'-u,,, whiel, ineln,lecl u lowe .nd'",',','" "' "'? *'"'<"'™"k " Jomt report of Matthew ™' "'''"•"■ '""»'""• In ^"■■vey of Canada, i„ ISot h" o r^ I '^ '" '''" ^""'"Si-'l overhud uneonfonnahly hy I r>.l^f 'V''''"''"'''' '° be """le known a Lower cl.t^anM '" ""'='' "'"" ''"^ ;vere compared with tiTllZJ "7""'^ *""""■ ""d lower division of the Col 1„.„ i '" "^ ^""'«'»- The - ;..cludi„g a larg am™, of S'"; T """ ^'"-"'^ ^ ond of bluish and reddish nor, I,v ''''""""'= 1""''^"« report was described, nnde tl 7 , T'' '" «'« same f cup, a series lithoW,, "'!;,"";"" "* "'' Blooras^ury but apparently resting ?, ' fe Men"'- '" "" ^""'"■"t^ fo>is,life,.o„s Upper Devon »n , "=™"'' ""d °™ri'dd bj^ apposed togradnate TirWoom'^ ""° "'"»" '' -«» ore regarded as altered Uppe" Dev"'^- ^'■°"" ™^ "■«■■«- 'anty to the pre-Cambria,. Coldb^ooT'""'' "'"' ''' '""'- supposrng both groups to consist ,nt """ '"P'"'"'" ''>' rocks. ^ ^ "*'^' "> '"rge part of volcanic I (! 1 • ' t ! , . n 408 THE HISTORY OF PRE-CAMBRIAN ROCKS. [IX. § 11. In 1869 and 1870, however, the writer, in com- pany with the gentlemen just named, devoted many- weeks to a careful study of these rocks in southern New Brunswick, when it was made apparent that the Blooms- bury group was but a repetition of the Coldbrook, on the opposite side of a closely folded synclinal holding Mene- vian sediments. These two areas of pre-Cambrian rocks were accordingly described by Messrs. Matthews and Bailey in their report, in 1871, as Huronian, in which were also included the similar crystalline rocks belonging to two other areas, which had been previously described by ibe same observers under the names of the Kingston and Coastal groups, and by them regarded as respectively altered Silurian and Devonian. § 12. After studying the Huronian rocks m southern New Brunswick, and their continuation along the eastern coast of New England, especially in Massachusetts (where, also, they are overlaid by Meneviau sediments), the writer, in 1870, announced his conclusion that the crys- talline schists of these regions are all of them pre-Cam- brian, and lithologically and stratigraphically equivalent to those of the Green Mountain range of western New England and eastern Canada. These he further declared, in 1871, to be a prolongation of the newer crystalline or Azoic schists of Rogers in Pennsylvania, and the equiva- lents of the Huronian of the Northwest. The pre-Cam- brian age of these crystalline schists in eastern Canada has now been clearly proved by the presence of their fragments in the fossiliferous Cambrian strata in many localities along the northwestern border of the Green Mountain belt, and farther by the recent stratigraphical studies of the geological survey of Canada. § 13. In close association with these Huronian strata in eastern Massachusetts is found a great development of petrosilex rocks, generally either jaspery or porphyritic in character, and sometimes fissile, whicli by Edward Hitchcock were regarded as igneous. These were now 'i5 W trata it of h'itic kvard I now IX.] PEE-CAMUKIA"!^ HOCKS IN NORTH AMEUICA. 409 found to be identical with tlie rocks designated, by Matthews and Bailey, feldspathic quartzites and silicious and porphyritic slates, which form the chief part of the Lower Coldbrook or inferior division of the Huronian series in New Brunswick. The petrosilexes of Massachu- setts were, after careful examinations by the writer, described by him, in 1870 and 1871, as indigenous strati- fied rocks forming a part of the Huronian series. He subsequently, in 1871, studied the similar rocks in south- eastern Missouri, and, in 1872, on the north shore of Lake Superior, but was unable to find them in the Green Mountain belt, or in its southward continuation, until, in 1875, he detected them occupying a considerable area in the South Mountain range, in southern Pennsylvania. The stratified petrosilex rocks of all these regions were described in a communication to the American Associa- tion for the Advancement of Science, in 1876, as appar- ently corresponding to the halleflinta rocks of Sweden, and, having in view their stratigraphical position, both in that country and in New Brunswick, they were then " pro- visionally referred " to " a position near the base of the Huronian series." Their absence in the Huronian belt in western New England, and in the province of Quebec, as well as at several observed points of contact between the Laurentian and the well defined Huronian in the North- west, led to tlie suspicion that these rocks might belong to an intermediate group (since named Arvonian). They may be briefly described as a series of stratified rocks, composed essentially of petrosilex, often passing into a quartziferous porphyry. There are found with it strata of vitreous quartzite, and thin layers of soft micaceous schists, besides great beds of hematite, and, more rarely, layers of crystalline limestone. § 14. C. H. Hitchcock has pointed out that the char- acteristic Huronian rocks do not form the higher parts of the Green Mountain range in Vermont, which he con- ceives to belong to an older gneissic series. He, however. .i ' '. 410 THE HISTORY OF PRE-CAMBRIAN ROCKS. [IX in Ids fiiuil report on the geology of New llampsliire, in 1877, adopts the name of Huronian for the cuystalline rocks of the Altered Quebec group of Logan, which make up the chief part of the Green Mountain range in Quebec, are largely developed along it in Vermont, and appear in a parallel range farther east, which extends southward into New Hampshire. In his tabular view of tlie geoguostic groups in this State, Hitchcock assigns to these rocks a thickness of over 12,000 feet, with the name of Upper Huronian ; while he designates as Lower Huro- nian the petrosilex series of eastern Massachusetts, already noticed, where these rocks are of great, though unde- termined, thickness. The similar petrosilex or halleflinta rocks in Wisconsin, where they have lately been described by Irving as Huronian, have, according to this observer, a thickness, in a single section, ( f 3200 feet. They here sometimes become schistose, are interbedded with unctu- ous schists, and rest in apparent conformity upon a great mass of vitreous quartzite. The writer has since exam- ined these rocks as seen on ibe Baraboo River, and else- where, in Wisconsin, and has satisfied himself of their identity with the similar rocks previously studied by him on the Atlantic coast, in Pennsylvania, in Missouri, and on Lake Superior. Besides the details respecting these petrosilex rocks to be found in the writer's volume on "Azoic Rocks," pp. 189-195, and, again, pp. 231-232, the reader is referred to Essay XL in this present vol- ume, §§ 37-42, for a fartlier account of their occurrence in Pennsylvania. The general high inclination both of this series and of the typical Huronian, renders tlie determina- tion of their thickness difficult. The maximum thick- ness of the Huronian (excluding the petrosilex or Arvonian series) to the south of Lake Superior, may, according to Brooks, exceed 12,000 feet, while the esti- mates of Credner and Murray, respectively, for this region, and for the north shore of Lake Huron, are 20,000 and 18,000 feet. ^ il i 8 ,^^ IX.J PRE-CAAIBEIAN liOCKS m NOKTH AMvr ISra. l.y the observutiou o ' ^.2? T '=°"'''™<'''' "' "-hei-e l,e found between the I '^ ^'* Biunxwick, -d the typical Hu.-onira trk rihT -"17''^^ ^'-'i' « indicated by a stratigraphtal di^^'"' '"'"'''"• '^''"^ presence, in the lower part o tie ,a r'r'™' "'"' ''^ "'« oonglonierates made up from te ,, JT""' "^ '=«»» pr petiosilex division TirH • "^ "'" «" Je,lyin„ - "-"y places, ; bbles afd T"' """"'" ^'""''"- P'eisses, a character commortoH^*""';"' "' "■« "Wer orystalline series, wiS led V "'"^ "" ^«" >'™"Ser America to class ;il o te e w Tt """''' S-°^'^g^^t.\ The Huronian contain? „ f"'""'""" ''""'''s- epidote, hornblende "and pyro?"'"""'^ i""i-'«™ of varieties of diabasic rocks oft "''„""'' '-^ """■'^'l by -e truly stratified, but are' not "Jt ^^ n"""'"^' ^''-h the nontes of the Norian «. '"' """founded with g*ro is „3o frequtt,;;™:' Th"'"^" "^ --** •noreover, includes imperfect i.' • ""'''""a" series, »'te,. serpentine, and Ste Tf'T', '"""•'^'"^«' dolo- chlorrtic, micaceouc, and ari M,. '"''' '"'«* "■""'"■ts of to be identical with the S-:::7 ^"^ists. It appears oi the Alps, which is thfre found °'' S'-^^nstone group the ancient gneisses beW «^ , """'^ '^'"■*'' •'"tweei gneisses and miea-schists tL ^ i" ^"""^'^ '""'^^ of g'ven at length in Essay x'"''^- °' "" ^^'"<''> « %i'lst.t'i:!;'^''>^Ap-i:;."' "'--^"^ '^« -ies in North°Amert: ifnt? hf "'"? " *'«■"»"'» ' eontanrs fine-grained wirte l5 ' ""'^ ^' '^^<^ that it mn it^tf tf > ^S 412 THE HISTOUY OF PllE-CAMBRIAN ROCKS. [IX. other. It also includes hornblendic gneisses and black hornblende-schists, together with serpentine, chrysolite- rocks, diciiroite-gneiss, and crystalline limestones. The mica-schists of the series often contain garnets, staurolite, andalusite, fibrolite, and cyanite, while in the granitic veins which traverse the series are found tourmaline, beryl, and cassiterite. The total thickness of the Montal- ban is apparently much greater than that assigned to the Huronian, upon which it sometii^es rests unconformably, or, in the absence of the Huronian, as is often the case, directly upon the Laurentian. § 17. As we are here following not the stratigraphical succession but the historic development of our knowledge of the American pre-Cambrian rocks, we return to a con- sideration of the more ancient gneisses. We distinguish at the base of the Eozoic system a massive and essentially granitoid gneiss, with little or no mica. To this funda- mental rock, sometimes called the Ottawa gneiss, and of unknown thickness, succeeds what has been named in Canada the Grenville gneissic series, made up in great part of a gneiss somewhat similar to that last mentioned, with intercalations of hornblendic gneiss, of quartzite, of pyroxenite, of serpentine, of magnetite, and of crystalline limestones, the latter often magnesian, occasionally graph- itic, and sometimes attaining thicknesses of a thousand feet or more. The Grenville series, the strata of which are generally highly inclined, has an aggregate volume of not less than 15,000 or 20,000 feet, and appears to rest unconformably upon the fundamental or Ottawa gneiss. This gneissic series, with its intercalated limestones, some of which contain Eozoon Canadense^ was the typical Laurentian of Logan and Hunt, named by them in 1854, with which they included, at that time, however, not only the underlying fundamental gneiss, but an upper granitoid and gneissoid series, composed in large part of plagioclase feldspars, chiefly labradorite. § 18. These three divisions of the Eozoic system were ' > ) IX.] PRE-CAMimiAN llOCKS IN NORTH AMKUICA. 413 I were thus coMfouncled under the common name of Laurentian until, in 1862, tlie hist was separated, under the pro- visu)nal name of Upper Laurentian, the two other divis- ions united being called Lower Laurentian. The syno- nym of Labradorian was subsequently, for a time, employed by Logan to designate the upper division, until 1870, when the present writer proposed for it the name of Norian, retaining that of Laurentian for the two lower divisions. It will probably be found desirable to separate the typical Laurentian or Grenville series, as studied and mapped by Logan, Hunt, and Dawson, from the less known fundamental or Ottawa gneiss, and to make of this latter a distinct group. The name of Middle Laurentian, sometimes given to the typical Laurentian, loses its signi- ficance with the disappearance of that of Upper Lauren- tian, now replaced by Norian. The Norian series is made up in great part of granitoid or gneissoid rocks, composed essentially of plagioclase feldspars, without quartz, but with a little pyroxene or hypersthene, often with titanic iron-ore, and apparently identical with the norites of Norway. With these rocks are, however, found alternations of gneiss and of quartz- ite, and also crystalline limestones, scarcely different from those of the Laurentian. We therein find also a grani- toid rock made up of pink orthoclase, quartz, and bluish labradorite. This Norian series is found in many places covering considerable areas, and apparently resting in discordant stratification upon the typical Laurentian. Its thickness has been estimated at over 10,000 feet. § 19. Passing now above the younger or Montalban gneisses and mica-schists, we come to a series composed essentially of quartzites, limestones, and micaceous and argillaceous schists. The quartzites, occasionally con- glomerate, are sometimes vitreous, sometimes granular, and often micaceous, graduating into mica-schists very distinct from those of the Montalban. The mica is often damourite or sericite, and gives rise to unctuous glossy # l'^: ti I* 414 THK HISTORY OF PRE-CAMBRIAN ROCKS. [IX. |M I' r t^ * f I i ) I schists, passing into argillites, which sometimes contain a feldspathic admixture. The limestones of this series, often magnesian, are crystalline, and include statuary marbles and cipolins. We find in the schists, which are intercalated alike among the quartzites and the lime- stones of this series, masses of serpentine and of ophical- cite, and occasionally of chloritic and hornblendic min- erals, as well as siderite, magnetite, and hematite, the iron-oxyds being often mingled with the quartzites. These last are sometimes flexible and elastic, and the whole series much resembles the Itacolumitic gi'oup of Brazil. It has a thickness in different parts of North America of from 4000 to 10,000 feet, and is seen lying unconformably alike upon the Laurentian, the Huronian, and the Alontalban. There are found in the quartzites of this series tlie imi)ressions of ScoUthtis, and in the lime- stones other undetermined forms. This is the Lower Taconic series of the late Dr. Emmons, which we dis- tinguish by the name of Taconian. Some recent writers have by mistake confounced it with the Upper Taconic of the same author, a distinct group, which Emmons declared to be the equivalent of the Primordial (Cambrian) of Barrande, and which is the original or unaltered Quebec group of Logan. § 20. The Taconian series is widely spread over east- ern North America, to the eastward of the great paleozoic basin, from the Gulf of St. Lawrence to Alabama. It is also found in an area to the north of Lake Ontario, in Hustings County, where it was described by the Canadian geological survey under the provisional name of the Hastings series, and is represented around Lake Superior by what has been called the Animikie series. This, though early recognized as Taconian in northern Michi- gan, and early separated by Logan from the Huronian on the north shore of the lake, has since been confounded with it. Much that in the northern peninsula of Michi- gan, as elsewhere, has been called Huronian, is Taconian. IX.] PRE-CAMBRIAN ROCKS IN NOl TH AMEIIICA. 415 Imau. The latter, the writer has elsewhere compared with a great seri^ of similar schists and quartzites, including serpentire, anhydrite, dolorute, and marbles, greatly developed in nortliern Italy, wliere it overlies the younger gneisses and mica-schists, and has been by various observers successively referred to the mesozoic, the pale- ozoic, and the eozoic periods. A full account of tlie Taconian series, its stratigraphical relations, and its dis- tribution in North America and elsewhere, will- be found farther on, in Essay XL § 21. The Taconian on the north shore of Lake Superior was by Logan made the lower division of his Upper Copper-bearing series, which, as a whole, was by him, after 1862, described as a modification of wliat lie then called the Quebec group. The upper divi^non of this Copper-bearing series, rer.:arkable for its native copper, had been previously, for a time, confounded by Logan with the Huronian itself, while by otliers it was referred to the Potsdam period, or even conjectured to be of mesozoic age. The geological distinctness of tliis great series of more than 20,000 feet of strara was, however, finally asserted by the present writer in 1873, when he called it the Keweenaw group, a name subsequently changed by him to Keweenian. It has since been shown by various observers that the fossiliferous sandstones which rest in horizontal layers upon the inclined strata of the Keweenian, belong to the Cambrian, and hold the fauna of the Potsdam. The conglomerates of the Ke- weenian cupriferous series contain portions alike of Lau- rentian, Arvonian, Huronian, and Montalban rocks, and overlie the schists which we have referred to the Taco- nian. The sandstones and argillites of the Keweenian, which are interstratified Avith great masses of melaphyre, are uncrystalline. It remains to be determined whether the intermediate Keweenian series has greater affinities with the Taconian than with the Cambrian, from both of which it is distinct. 416 THE HISTORY OP PltE-CAMBRIAN UOCKS. [TX. We have thus sought to include, provisionally, the wliole vast system of Primitive and Transition crystalline rocks, from the fundamental granitoid gneiss upward, under the names of Laurentian, Norian, Arvonian, Iluronian, Mont- alban, and Taconian. Certain considerations regarding the distribution and the stratigra[)hical relations of these have already been set forth, on i)age 184, to vv^hich the reader is referred. ♦ t II. — PRE-CAMBRIAN ROCKS IN EUROPE. § 22. In an address before the American Association for the Advancement of Science, in 1871, in which the writer maintained the Huronian age of a portion of the crystalline schists of New England and Quebec, he ex- pressed the opinion, based in part upon his examinations at Holyhead in 1867, and in part upon the study of col- lections in London, that certain orystalline schists in North Wales would be found to belong to the Huronian series. The rocks in question were by Sedgwick, in 1838, separated from the base of the Cambrian, as belonging to an older series, but were subsequently, by De la Beche, Murchison, and Ramsay, described and mapped as altered Cambrian strata with associated intrusive syenites and feldspar-porphyries. § 23. In South Wales, at St. David's in Pembrokeshire, is another area of crystalline rocks, which the geological survey of Great Britain had mapped as intrusive syenite, granite, and felstone (petrosilex-porphyry), having Cam- brian stratr, converted into crystalline schists on one side, and unaltered fossiliferous Cambrian beds on the other. So long ago as 1864, Messrs. Hicks and Salter were led to regard ti.cse granitoid and porphyritic rocks as pre-Cam- brian, and in 1866 concluded that they were not eruptive but stratified crystalline or metamorphic rocks. After farther study. Hicks, in connection with Harkness, pub- lished, in 1867, additional proofs of the bedded character of these ancient crystalline rocks, and in 1877 the K I IX.] PRE-CAMBUIAN ROCKS IN EUIIOPE. first-named observer announced the conclusion tluit they belong to two distinct and unconformable series. Of these, the older consisted of the granitoid and por- phyritic felstone rocks, and the younger of greenish crys- talline schists, the so-called Altered Cambrian of the official geologists ; both of these being overlaid by the undoubted Lower Cambrian (Harlech and Menevian) of the region, which holds their ruins in its conglomerates. To the lower of these pre-Cambrian groups. Hicks gave the name of Dimetian, and to the up[)er that of Pebidian. The last, with a measured thickness of 8000 feet, he sup- posed to be the equivalent of the Huronian, and com- pared the Dimetian with the Upper Laurentian of Logan. The Dimetian, including the granitoid and gnei '.c rocks of both Norti) and South Wales, so far as seen oy the writer in the limited outcro[)S, resembles the Lauren- tian of North America. It was by a misconception that Hicks provisionally referred the Dimetian to the Upper Laurentian, — a name at one time used by the geological survey of Canada to designate the Norian series. Hicks, at the same time, designated as Lower Laurentian the gneiss of the Hebrides (Lewisian of Murchison), which he believed to be distinct from and older than the Dime- tian. These two may correspond to the Ottawa and Grenville divisions of the proper Laurentian in Canada. § 24. The similar crystalline rocks of North Wales, already noticed, were now studied by Prof. T. McKenna Hughes, of Cambridge, who described them in 1878. These include in Carnarvonshire and Anglesey the green- ish crystalline schists which the writer, in 1871, referred to the Huronian (pre-Cambrian of Sedgwick, and Altered Cambrian of the geological survey), certain gi-anitoid rocks formerly described as intrusive syenite, and also a reddish feldspar-porphyry which forms two great ridges in Carnarvonshire. This latter was by Professor Sedg- wick regarded as intrusive, and is, moreover, mapped as such by the geological survey, though described in Ram- 418 THE Ul«TOUV OF PBB-(;AMBRIAN ROCKS. PX. say's meiucir on the geology of North Wales as probably the result of an extreme metamorphism of the lower beds of the Cambrian. The pre-Cambrian age of all tlieso rocks was clearly shown by Hughes, who, however, con- sidered that tlie whole might belong to one great strati- lijd series; wliile Hicks, from an examination of the same region, regarded them as identical with the Dimetian and Pebidian of South Wales. § 25. Dr. Hicks continued his studies in both of these regions in 1878, — being at times accompanied by Dr. Toroll of Sweden, Professor Hughes and Mr. Tawney of Cambridge, and the writer, —and was led to conclude tl'.at, besides the chloritic schists and the greenstones of the Pebidian series, and the older granitoid and gneissic rocks, there exists, both in North and South Wales, an- other independent and Intermediate series, to which belongs the stratified potrosilex or quartziferous por- phyry already noticed. Tliis is sometimes wanting at the base of the Pebidian, and at other times forms masses some thousands of feet in thickness. At one locality, near St. David's?', a great body of breccia or conglomerate, consisting of frrgments of the petrosilex united by a crystalline dioritic cement, forms the base of the Pebid- ian. For this intermediate series, which constitutes the quartziferous porphyry-ridges of Carnarvonshire, Hicks and his friends then proposed the name of Arvonian, from Arvonia, the Roman name of the region. § 26. This important conclusion was announced by Dr. Hicks at the meeting of the British Association for the Advancement of Science, at Dublin, in August, 1878. The writer, previous to attending this meeting, had the good fortune to examine these various pre-Cambrian rocks in parts of Carnarvonshire and Anglesey with Messrs. Hicks, Torell, and Tawney. He subsequently, in company with Hicks, visited the region in South Wales where these older rocks had been studied, and was enabled to satisfy himself of the correctness both of I?:. J rillMIAMniMAX KOCKS I .V KUUOl'E. 410 by for 878. the the observations and conclusions of Tlickd, and of the complete parallelism in 8tratigra[)liy and in mineral com- position between these pre-Cambrian rocks on the two sides of tlie Atlantic. It may hero be mentioned that Torell, who, during his visit to America in 187G, had an opl)ortunity of studying, with the writer, the potrosUexes of New England and Peinisylvania, — which he regarded as identical with the hiilleflinta of Sweden, — at once recognized them in the Arvonian series of North Wales. Of the many areas of these various prc-Cambrian rocks which the writer was enabled to examine in company ■with Hicks, may be mentioned the granitoid mass of Twt Hill in the town of Carnarvon, and the succeeding Arvonian to Port Dinorwic, followed, across the ^Nlenai Strait, by the Pebidian on the island of Anglesey, near the Menai bridge. Farther on, the Pebidian was again met with, near the railway station of Ty Croes, in the southwest part of the island, succeeded by a large body of Arvonian petrosilex, and a ridge of gronitoid gneiss, fragments of which make up a breccia at the base of the Arvonian series. The Pebidian is again well displayed at Holyhead. § 27. In South "Wales, the similar rocks -were exam- ined by him at St. David's, where three small bands or veins of an impure, coarsely crystalline limestone are included in the Dimetian granitoid rock, which is here often exceedingly quartzose. It may be remarked that the Dimetian, as originally defined at this, its first recog- nized locality, included a great mass of Arvonian petro- silex, the two forming a ridge which extends for some miles in a northeast direction, flanked by Pebidian rocks, which are sometimes in contact with the one and some- times with the other series. At Clegyr Bridge was seen the base of the Pebidian, already mentioned as consisting of a conglomerate of Arvonian fragments. Another belt of the same crystalline rocks was also visited, a few miles to the eastward of the last, and not far from Haverford- 420 THi: IIISTOUY OF rUE-CAMimiAN HOOKS. lix. I !i- west, forming, according to Hicks, a ridge several miles in lenjith and about a mile wide. Where seen at Ilocii Castle, it was found to consist of Arvonian petrosilex, with some gianitoid rock near by. The ridge is Hanked on the northwest side by Pebidian and Cambrian, and on the soutiieast by Silurian strata, let down by a fault. § 28. On the shore of Llyn Padarn, near the foot of Snowdon, in North Wales, the jiorphyritic petrosilex of the Arvonian is again well displayed, while in contact with it, and at the base of the Llanberis (Lower Cambrian) slates, is a conglomerate made up almost wholly of the petro- silex. This locality was supposed by Kamsay and others to show that the petrosilex is the result of a metamor- phosis of the lower portion of the Cambrian, the conglom- erates being regarded as beds of passage. The writer, after a careful examination of the locality, agrees with Messrs. Hicks, Hughes, and Bonney that there is no ground for such an opinion, but that the conglomerate marks the base of the Cambrian, which here reposes on Arvonian rocks, and is chiefly made up of their ruins. In like manner, according to Hughes, the Cambrian in othet parts of this region includes beds made of the dShris of adjacent granitoid rocks. § 29. These petrosilex conglomerates of Llyn Padarn are indistinguishable from those found at Marblehead and other localities near Boston, Massachusetts, which have been in like manner interpreted as evidences of the sec- ondary origin of the adjacent petrosilex beds, into whicn they have been supposed to graduate. Tlie writer has, however, always held, in opposition to this view, that these conglomerates are really newer rocks, made up of the ruins of the ancient petrosilex. He has found simi- lar petrosilex conglomerates at various points on the Atlantic coast of New Brunswick, of Lower Cambrian, Silurian, and Lower Carboniferous ages, all of which have, in their turn, been by others regarded as formed by the; :,-i ni IX.] rUE-CAMnUIAX HOCKS IN EUUOT'K. 421 of of the alteration of strata of thcso geological periods. Tho cvideijco now furnislicd in South WuIoh of still older (Iluronian) beds of i)otro.siU;x conglonierato sliould be noted by students of North American geology. From observations near Doston, made by one of my former students, I liavo for some time suspected the exist- ence of petrosilcx conglomerates of pre-Cambrian age. § 30. To the eastward of the localities already men- tioned in Wales, are some other small ureas of crystalline rocks, including those of the Malverns, and tho Wrekin and other hills in Shropshire, all of which appear as islands among Cambrian strata ; also those of Charnwood Forest, in Leicestershire, which rise in like manner among Triassic rocks. Tho Wrekin, regarded by ]Mur- chison as a post-Cambrian intrusion, has been shown by Callaway to bo unconformably overlaid by Lower Cam- brian strata, and consists in part of bedded greenstones, and in part of banded reddish i^etrosilex-porphyries, closely resembling tho Arvonian of North AVales and the corresponding rocks of North America. Tho geology of Charnwood has within tin past two years been carefully studied by Messrs. Hill and Bouncy. The ancient rocks of this region are in part crystalline schists (embracing, in the opinion of Hicks and of the writer, — who have seen collections of them, — representatives both oi! the Pebidian and the Arvonian of Wales) and in part erup- tive masses, including the granitic rocks of Mount Sorrel. § 31. The crystalline schists of Charnwood offer, as was pointed out by Messrs. Hill and Bonney, many resem- blances with parts of the Ardennian series of Dumont in France and Belgium. These, which have been in turn regarded as altered Devonian, Silurian, and Lower Cam- brian, were, as shown by Gosselet, islands of crystalline rock in the Devonian sea, and in one part include argil- lites with impressions of Oldhamia and an undetermined graptolite. These rocks have lately been described in >4i! n 'm ii 422 THE HISTOPwY OF niE-CAMBRIAN ROCKS. [IX. I 5 ^^l .i (' ■M iM<.' detail in the admirable memoir of De la Valine Poussin and Renard. The writer had the good fortune, in 1878, to visit this region, and, in company with Gosselet and Renard, to examine the section along the valley of the Meuse. The crystalline rocks here displayed greatly resemble those of the American Huronian, in which may be found most of the types described by the authors of the memoir just mentioned.* § 32. Hicks, in a paper on the Classification of the British Pre-Cambrian Rocks, which is published in the Geological Magazine for October, 1879, concludes that the Pebidian is " a group of enormous thickness, which is largely distributed over Great Britain, where it has a pre- vailing strike of N.N.E. and S.S.W., or from this to N.E. and S.W." In addition to the localities which we have * The following is a partial list of recent publications regarding these rocks, as discussed in §§ 23-32, to the close of IST'J. For their later history, see farther on. Essay XI., §§ 189-193. In the Quar. Jour. Geol. Sec. of London are the following papers on these rocks in Wales: Hicks, May, 1877, p. 230; Il'cks & Davies, February, 1878, p. 147, and May, 1878, p. 153; Iluglies & Bonney, February, 1878, p. 137; Hicks & Davies, May, 1879, p. 285; Ilicks & Bonney, ihid., p. 205; Bonney, ibid., p. 309; Bonney & Houghton, ibid., p. 821; Hughes, November, 1879, p. 082; Maw, August, 1878, p. 704. Also Hicks, Rocks of Ross-shire, November, 1878, p. 811; Tawney, Older Rocks pf St. David's: Proc. Bristol Naturalists' Society, vol. ii., part 2, p. 110. On these rocks in Shropshire, Quar. Joui'. Geol. Soc, Allport, August, 1877, p. 449; Callaway, November, 1877, p. 653, and August, 1878, p. 754; Callaway & Bonney, November, 1879, p. 043. On these rocks in Cham- wood Forest, in the same journal, Hill & Bonney, November, 1877, p. 753, and May, 1878, p. 199. See farther. Hunt, Chemical and Geological Essays, pp. 34, 209, 270, 272, 278, 283; also his Azoic Rocks (Second Geol. Survey of Penn., 1878), pp. 187, 188. For the rocks of the Ardennes, see Mgmoire sur los Roches dites Plu- toniques, etc. (4to, pp. 264), by De la Valine Poussin and Ifenard, from Memoiresdel'Acad. Royalede la Belgique for 1870 : Mt'moire surlaComp. Min^ralogique du Coticule, by Renard, from the same for 1S77; and The Mineralogical and Microscopical Characters of the Belgian Whetstones, by Renaid, Monthly Microscopical Journal for 1877. vol. xvii., p. 209. Also Gosselet and Malaise, Terrain Silurien des Ardennes, Bull. Acad. Roy. de la Belgique (2), No. 7, 1808; Dewalque, Terrain Cambrieii des Ardennes, Ann. Soc. G^ol. de la Belgique, torn. I., p. 03; and farther, Hunt, Chem. and Geol. Essays, p. 270. IX.l PRE-CAMBRIAN ROCKS IN EUROPE. 423 already mentioned in Great Britain, he notes its occur- rence in Shropshire and in Charnwood Forest, and also in the nortliwest of Scotland, where, as elsewhere, it enters largely into the Lower Cambrian conglomerptes. The group is concisely described by him as consisting, "for the most part, of chloritic, talcose, feldspathic, and micaceous schistose rocks, alternating with slaty and massive green- stones, dolomitic limestones, serpentines, lava-flows, por- cellanites, breccias, and conglomerates. It is also traversed frequently by dikes of granite, dolerite, etc." § 33. There is not, so far as yet known, in any of the British localities especially mentioned, any representative either of the Taconian or the Montalban series. The presence of rocks having the characters of the Huronian was, however, indicated as having been observed by me in various parts of Perthshire and Argyleshire, and also on Lough Foyle, in L*eland, where I have observed the Montalban well displayed in the Dublin and Wicklow Hills, and pointed out the probable presence of both Huronian and Montalban in specimens of rocks from Donegal. To the latter series I also referred, from similar evidence, in 1871, certain crystalline schists from the Scot- tish Highlands, where the typical Pebidian of Hicks, previously designated by me as Huronian, is also largely displayed. Hicks has since found there a series of crystalline strata which succeed the Pebidian, and which he has called Upper Pebidian. These, as they are the predominant rocks in the Grampian Hills, he proposed to name the Grampian series. The}'' consist in great part of tender gneisses or granulites, with mica-schists, and, as I have elsewhei-e pointed out, have all the characters of the Mont- alban or younger gneiss series, as seen alike in North America and in the Alps. The conclusion from all the observation of Hicks and Callaway up to 1882, as then stated by me, was that " the crystalline strata of the Scot- tish Highlands, regarded by the geological survey of Great 424 THE HlSTOllY OF PllE-OAMIUlIAN HOCKS. [IX. 1^ f' ,r * ,1 Britain as altered paleozoic; strata, include representatives of various pre-Cambrian groups, including Montalban (Grampian), Huronian (Pebidian), and Arvonian, to which group Hicks refers the petrosilex st ies found in Glencoe." * § 34. For an account of more recent investigations in these crystalline rocks of the Scottish Highlands, the reader is referred to Essay XI., §§ 190-193. In this essay, which discusses the Taconian, and its relations alike to newer and to older oeries, will be found much on the various divisions of Eozoic rocks. Therein is noticed the recent attempt of Dana to resuscitate the long abandoned views of Nuttall and Mather as to the paleozoic age of a great area of crystalline rocks, principally Laurentian, in Westchester County, New York. The veinstones of these ancient crystalline rocks, and especially of the Laurentian series, have been described at length, ante, pages 223-238. An account of the pre-Cambrian rocks of the Alps and the Apennines will be found farther on, in part iv. of Essay X. § 35. In closing this historical sketch, mention should be made of the stateniv^ents put forth, in 1879, by A. R. C. Selwyn, director of the geological survey of Canada, in his report of progress for 1877-78 (p. 14 A), in which he sought to set aside the whole of the preceding classifica- tion of Eozoic rocks. He then proposed to unite in one group, under the name of Huronian, not only the rocks around Lakes Superior and Huron, to which this name was originally give , and the crystalline belt in the prov- ince of Quebec (which I had already, in 1871, called Huronian), but the whole o*^ the Upper Copper-bearing series of Lake Superior, thus embracing both the Taco- nian, or lower 'division of the latter, and the Kewee- nian. Not content with this, he would farther include in the Huronian the "Templeton, Buckingham, Grenville, and llawdon crystalline limestone series," which is the * Progress of Geology, Smithsoniau Report for 1S82. /. IX] PEE-CAMBRIAN KOCKS IN EUROPE. 425 Grenville gneissic series of Logan and myself, and also the " Upper Laurentian or Norian." He moreover added to these the Hastings limestone series, which he supposed might be an equivalent of the Grenville series. The Mont- alban and the Arvonian were overlooked in his scheme, but with these exceptions the Huronian, as imagined by Selwyn, was made to include every known grouD of strata, whether crystalline or uncrystalline, from the base of the fossiliferous Cambrian to what he designated as "those clearly lower unconformable granitoid or syenitic gneisses " which contain no bands of calcareous or other extraneous rock; and which may be supposed to corre- spond to that basal division described above as the Ottawa gneiss. The whole of the vast succeeding series of 16,000 feet of granitoid gneisses, with quartzites, crj'stalline lime- stones, etc., which had been studied during thirty years by Logan, Murray, and myself, and constituted, with the Norian, the Laurentian system as originally defined and named in 1854, was thus to be confounded, under one name, with the widely different series to which the desig- nation of Huronian had been given in 1855, and with the entire Upper Copper-bearing series of Logan, embracing alike the Taconian and the Keweenian. It may be pre- sumed that longer study and larger opportunities of observation will lead Mr. Selwyn to conclusions more in harmony with those of his predecessors. Appendix. The whetstone or coticule of the Ardennes, named on page 422, consists, according to the chemical and microscopical studies of Kenard, of rounded grains or minute ciystals of manganese-alumina garnet (spessartine), with others of green tourmaline and probably of chryso- beryl, included in a damourite-like mica, sometimes with pyrophyllite in fissures, and with intersecting veins of quartz. Layers of this aggregate from one to ten centimetres thick, pale yellow in color, conchoidal in fracture, and with density 3.22, are interstratified with, and graduate into, a fine-grained schist or phyllade, itself with transverse cleavage, made up chiefly of similar micas, but contalnin^i besides the garnets, etc., plates of heuiatite, and carbonaceous grains. The evidence of contempo- raneous formation of these various species is clear, and illustrates well the crenltic process. 4 M :ii# ;^^^ m hi )' I i,'i , ! i I X. THE GEOLOGICAL. HISTORY OF SERPENTINES, WITH STUDIES OF PRE-CAMBRIAN ROCKS. This essay wat presented to tlie Royal Society of Canada, May 23, 1883, and printed under its present title in the first volume of thvi Transactions of the Society. I. — HISTORICAL INTRODUCTION. § 1. Few questions in geology are involved in greater obscurity or more contradiction than the history of ser- pentine. As a preliminary to a discussion of certain observations by myself and others thereon, it seems, there- fore, desirable to recall some passages in this Instory, which nivay serve to show the differences of opinion now existing, and, it is hoped, prepare the way for their recon- ciliation. These differences may be considered under two heads, namely : the geognosy of serpentine, or its relation to the other rosks of the earth's crust ; and the geogeny, or the origin and mode of formation of serpentine, the mineralogical relations of which are discussed on page 333. Setting aside for the moment the question of the occur- rence of serpentine as an accidental mineral disseminated in calcareous rocks, and considering only its occurrence in rock-masses, either pure or mingled with other silicates, the first question which presents itself is whether such massive serpentines are contemporaneous with the enclos- ing rocks, or whether they have been subsequently in- truded among these, — in other words, whether serpentines are indigenous or exotic rocks. § 2. We find at the beginning of our century that the most competent ol>servers were agreed in regarding ser- pentines as stratified contemporaneous deposits in the so- called primary rocks. Patrin described those of Mont 426 ^■] THE GEOLOGICAL HISTORY OF SERPENTINES. 427 Rose and of the Rothhorn as interstratified with calcare- ous and micaceous schists, while Saussure found those of Mont Cerviu to present similar conditions, and described certain serpentines, found near Genoa, as alternating with bands of calcareous, quartzose, and micaceous schists or argillites. Humboldt, in like manner, noticed the strati- fied character of the serpentines near Bareith, and Jameson found those of Rothsay, in Scotland, to be interstratified with micaceous and talcose schists, and with crystallii;e limestone, in repeated alternations, of which he gives a diagram, mentioning, l.owever, as an opinion held by some, that the masses both of serpentine and of limestone " form great veius rather than vertical sheets." He else- where describes serpentine as a primitive stratified rock, contemporaneous, and alternating with crystalline schists.* § 3. A little later we find, in 1826, jNIacculloch, in his "Geological Classification of Rocks," separating the prim- itive rocks into two groups, stratified and unstratified, the latter consisting of granite and serpentine. He as- signed as a reason for placing serpentine in the latter class that it does not appear to be decidedly stratified, but, at the same time, remarks that, unlike other unstrati- fied rocks, as granite or t'-ap, he had not found serpen- tines to present ramifying- veins. Subsequent studies in the Shetland Isles led him .o make what he calls " an im- portant correction " in its history, in the Apiiendix to the volu'^'ie just named, where he ainiounces his conclusion that tiie serpentines are stratified rocks, like gneiss or mica-schists, adding a revised tabular view, in which they are included with these in the stratified division of the primitive vocks, granite alone being retained in the un- stratified division.! § 4. Boase, in his " Primary Geology," in 1834, de- scribes the serpentines of Cornwall as associated with tal- * See, for the text of the above references, the quotations in .' nker- ton's Petral'jgy, 1821, i., 334-343 ; ii., COS-612. t Macculloch, loc. cit., pp. 78, 243, 052-655. •US «'r 428 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. cose and cliloritic and actinolite-schists, and what had been " called hornblende-slate," to which the serpentine seemed in some instances subordinated. He farther com- pares these associations and modes of occurrences with those described by MaccuUoch.* De la Beche, in like manner, in his " Geology of Cornwall and Devon," notes the seeming passage of the serpentine into the hornblende- . tte in man}' places, but also its apparent "intrusion an id the latter with force"; a seeming contradiction, which he recognizes, but endeavors to explain.f § 5. Unlike MaccuUoch and Boase, De la Beche re- garded serpentine as nn eruptive rock of posterior origin to the associated schists, agreeing in this with Brongniart, who had placed serpentine among pliitonic rocks. A similar view was held b}^ Elie de Beaumont | and by Savi, and, without entering into farther details, we may notice that they have been followed by Sismondl, Lory, and others, who maintain the plutonic origin of the Alpine serpentines, while, on the other hand, Scipion Gras, Gastaldi, Favie, and Stapff regard them as of aqueous and sedimentary origin. The views of the present school of Italian geologists, as well as Dieulefait and Lotti, will be noticed in part vi. of this essay. § 6. In the United States, we find Edward Hitchcock, in 1841, reviewing the opinions of MaccuUoch, Brongniart, De la Beche, and others, and deciding that the serpentines of Massachusetts are to be regarded as stratified rocks. § Emmons, in 1842, after noticing the conclusions of Hitch- cock as to serpentine, regarded it, nevertheless, as an un- stratified rock, but distinguished it from trappean rocks, * Boase, loc. cit., p. 46. t De la Beche, loc. cit., pp. .35, 09. ^ t After discussing the question with Elie de Beaumont, in 1855, I asked his eminent colleague, De Senarmont, as to the eruptive origin of serpentines. He replied that his own extended studies of the serpentines of Europe had led him to reject as wholly untenable the theory of their plutonic character. § Geology of Massachusetts, II., 616, X.] THE GEOLOGICAL HISTORY OF SERPENTINES. 429 inasmuch as, according to him, it is never found in injected veins or dikes.* Later, however, in 1855, he separated it from so-called pyroplastic rocks, like " basalt, trap, and greenstone," and included it in lioth divisions of his pyro- crystalline class: that is to say, (1) as laminated serpen- tine, with gneiss, micaceous, talcose, and hornblendic slates and limestone ; and (2) as massive serpentine, with granite, syenite, etc.f § 7. J. D. Whitney, in 1851, included hornblende and serpentine rocks, together with the magnetic and specular oxyds of iron, under the title of " Igneous," and the sub- title of " Trappean and Volcanic Rocks." J Henry D. Rogers, in 1858, described the steatite belt on the Schuyl- kill River, in Pennsylvania, as formed from the mica- schists of the region through impregnation from " the dike of serpentine ^yhicll everywhere adjoins it," thus implying the posterior origin and eruptive character of the latter. Elsewhere he describes the cr^-stalline rocks of the same region as including "true injected serpentines." He, however, looked on veins of quartz and epidote, and even of carbonate of lime, as also of eruptive o'' gin.§ Lieber, at the same time, in his report on the geology of South Carolina, regarded not only the serpentines of that region, but the associated steatite and actinolite rocks, as erup- tive masses. § 8. In opposition to these plutonic views, the geologi- cal survey of Canada, from an early date (1848), insisted upon the stratified character of the serpentines found in the northern extension of the Green Mountain range in eastern Canada. They were shown to be accompanied by hornblendic, steatitic, dioritic, and other schistose rocks, as well as by dolomites and magnesites. The writer, in dis- * Geology of Kew York, Northern District, pp. 67-70. t American Geology, I., 43. t Geology of Lake Superior, II., 2. § Geology of Pennsylvania, vol. I., passim. See also the author, in Azoic Bocks, pp. 15-19. 411 430 THE GEOLOGICAL HISTORY OP SERPENTINES. [X. cussing the relations of tliese in 1863, announced " the conclusion that the whole series of rocks . . . from diorites, diallages, and serpentines, to talcs, chlorites, and epidosites, have been formed under similar conditions," and were aqueous deposits.* § 9. Here it will be seen that we approach the second question mentioned in § 1, namely, that of the origin and mode of formation of serpentines, which, in the view of t^ 3 wlio maintain its indigenous character, is, of course, «; sely connected with the problem of the origin of its .1, lated crystalline rocks. The notions of the earlier ;;^fcolo,"' ' s with regard to this latter problem were, in most cases, \tiry vague, some of them holding the view, still taught in our own day by H(3bert, that these rocks, includ- ing gneisses, micaceous, chloritic, and hornblendic schists, were all formed by some unexplained process during the cooling of the globe, without the intervention of water.f With few exceptions, however, they admitted, with Wer- ner, the aqueous origin of tliese, whether holding, with De la Beche and with Daubree, that they were deposited successively from the highly heated waters of a primeval sea,J or the more commonly received view, that the sedi- ments were laid down under conditions of temperature not unlike those of the present time, and were afterwards the subject of internal change (diagenesii), or of indefinite replacement and substitution (metasomatosis). § 10. The latter doctrine, which, in the hands of some of its disciples, has found an extension limited only by their imaginations, was at once applied to explain the origin of serpentine. Silicated rocks destitute of magne- sia, and carbonated rocks destitute of silica could alike, it was maintained, be converted into serpentine, which was * Geology of Canada, p. 612. See also the author's Contrihutions to the History of Ophiolites, 1858, Amer. Jour. Sci., xxv., 217-226, and xxvi., 234-240. t Bull. Soc. Geol. de France, 1882, xi., 30, and ante, p. 85. t Chemical and Geological Essays, p. 301, and ante, pp. 104-106. ! ^ 11^^ [X. X.1 THE GEOLOGICAL HISTORY OF SERPENTINES. 431 held to be tlie last term in the metasoniatic changes of a vast number of mineral species. Hence, it was no longer necessary to supi)ose the direct deposition of a magnesir<'> sediment, or an eruption of an igneous magnesiun rock, tc explain the presence of contemporaneous or of inject serpentines. The legitimate outcome of this hypothesis is found in tlie teaching of Delesse, in 1858 (when he yet held the eruptive nature of serpentine, which he classed with other " trappean rocks"), — namely, tliat "granitic and- trappean rocks " may, in certain cases, be changed into a magnesian silicate, which may be serpentine, talc, chlorite, or saponite.* § 11. I have elsewhere shown ^ ow Delesse, three years later, abandoned alike the met. on ic hypothesis and the notion of the eruptive ori 'in r ^he serpentines, in favor of that view which I ] \0 puc forth in 1859 and 1860, that the serpentines wc'e ■' imdoubtedly indigenous rocks, resulting from the aUera.xOn of silico-magnesian sediments." At the same ' e, is a concession to those who maintained the occurrence of eruptive serpentines, it was said that " the final result of heat, aided by water, on silicated rocks would be their softening, and in certain cases their extravasation as plutonic rocks," which were to be regarded as " in all cases altered and displaced sedi- ments." t Later, in re-stating this point in 1880, it was said, " The eruptive rocks, or, at least, a large portion of them, are softened and displaced portions of ancient nep- tunian rocks, of which they retain many of the mineralogi- cal and lithological characters." $ The proviso contained in the last sentence is explained by the view since main- tained at length, in Essays V. and VI. in this volume, that the rocks of the basaltic and doleritic tyjjcs are portions of an original igneous mass, which antedated the appear- ance of liquid water at the surface of the globe. * Ann. des Mines (5), xii., 509, and xiiL, 393, 415. t Chem. and Geol. Essays, pp. 316-318. t Amer. Jour. Science, xix., 270, and ante, p. 126. t : fit] :i r ■ 'Vi. m \'''.i 432 THE GEOLOlilCAL IIISTOUY OF SERPENTINES. PC- § 12. After careful stiuliea, alike in the field and in the laboratory, I was led, in 18(50, to maintain that the origin of serpentine and related niagnesian rooks was to be found in deposits of liydrous silicates, like the magnesian marls of the Paris basin ; and in 1861 we not only find Delesse teaching this doctrine of the origin of these rocks from the alteration, or so-called metamorphism, of such magne- sian precipitates, but declaring, in the spirit of my teach- ing, as above, that " the plutonic rocks are formed from the metamorphic rocks and represent the maximum of intensity, or tiie extreme term of general metamorphism."* The history of the abandonment by Delesse of his former view of the plutonic for that of the neptunian origin of serpentines, and his acceptance at the same time of the liyi)othesis of an aqueous origin of plutonic rooks, is sig- nificant as a recognition of the new ideas for which I had contended, and which constitute a new departure in theo- retical geogeny. See, farther, on this point, a note to § 116, by Dieulefait.f § 13. In farther explanation of this source of magne- sian silicates, it was shown by the writer, in a series of experiments the results of which were published in 1865, that whenever the comparatively soluble silicates of alkalies or of lime (which are set free by the decay of crystalline silicates, and are found in many natural waters) are brought in contact with solutions, like sea- * Delesse, fitudes sur le Metamorpliisme, 1861, p. 87. t The testimony of Sclpion Gras, in 1854, in his learned memoir, " Sur le Terrain Anthraxiffere des Alpes" (Ann. des Mines ['>], v., 473-602), against the igneous hypothesis, should here be recorded. Of the seipen- tlnes, euphotides, variolites, and so-called spilites, of the Alps, having said that they are either eruptive or rocks altered in place, he adds: " We have long since adopted this latter hypothesis, which alone appears to us to be 'n accordance with observation. It is not uncommon to see these so-called plutonic rocks of the Alps offer a distinct stratification, the appearances of which are exactly like those of adjacent sediments. This is especially true for the spilites and the serpentines, the epigenir; origin of which isevident." (Loc. cit., p. 601.) This epigenic hypothesis involved a metamorphosis of sediments by heat and moisture, apparently not iinlike that mentioned farther on, in § 111. I'-r i. :^ ; X.] THE GEOLOGICAL HISTORY OF SEIIPENTINES. 433 water, lioldiiig iiingnesian sulphnte or clilorul, cloublo decomposition takes place, with the sei)aration of a very insoliibki gelatinous silicate of magnesia; and farther, that preciiiiaited silicate of lime is decomposed by diges- tion with such nnignesian solutions, its lime becoming partially or wholly replaced by magnesia. Tlup process, it was pointed out, is the reverse of tliat which happens when carbonates of alkalies or of mag- nesia come in contact with sea-water, in which case the comparative insolubility of carbonate of lime causes ilie decomposition of the soluble calcium-salts present. "In the one case, the lime is separated as carbonate, the mag- nesia remaining in solution ; while in the other, by the action of silicate of soda (or of lime), the magnesia is removed, and the lime remains. Ilcnce carbonate of lime and silicates of magnesia are found abundantly in nature, while carbonate of magnesia and silicates of lime are produced only under local and exceptional conditions. It is evident that the production from the waters of the early seas of beds oP sepiolite, talc, serpentine, and other rocks, in which a magnesian silicate abounds, must, in closed basins, have given rise to waters in which chlorid of calcium would predominate." * The generation of magnesian silicates in aqueous sediments was thus shown to be the result of a natural process as simple as that giving rise to carbonate of lime. § 14. There are many questions connected with this theory of the source of serpentine and related rocks, such as the probable variations in the composition of the origi- nal silicates ; their admixture with other silicates and carbonates ; the changes wrought in these by subsequent chemical reactions, resulting in the genesis of talc, ser- pentine, enstatite, and olivine, and, in certain cases, the subsequent changes of these anhydrous species ; the pres- ence, in these magnesian minerals, of ferrous silicate, which is so abundant in many serpentines, and its relations to * Amer. Jour. Sci. (2), xl., 49 ; also, Chem. and Geol. Essays, p. 123. nm ! ■ M ! !■: i : J t 434 THE GEOLOGICAL HISTOnY OF SEItPENTINES. PC the problem t)f tlio ori<^lii of glaiiconitc, itself sometimes a more or less magnesiiiii siliciite; finally, the notable fact of the prcseiico in most of these magnesian rocks of small portions of the rarer metals, such as nickel and chromium, which is to be considered in connection with the similar metallic inii)rognation of certain mineral waters that may well liave intervened in the production of these magnesian silicates. All of these are important points, which must be reserved for future discussion. [Some of these have .since been considered, in connection with the question of glauconite, om pages 19G-198.] § 15. One great object in geology is to discover by what natural processes the different chemical elements liave been segregated and combined during successive ages in the forms in which we now find ihem in the earth's crust ; in other words, how from a once homogeneous mass have been separated quartz, corundum, bauxite, car- bonates of calcium and magnesium, as well as carbonates, oxyds, and sulphids of manganese, iron, zinc, copper, and other metals. Not less important is the problem of the genesis of the corresponding protoxyd-silicates, and espe- cially of those of calcium, magnesium, and iron, Avhich form, often with little or no admixture, considerable masses in the earth's crust. Of these, it is unnecessary to say, the magnesian rocks under consideration constitute an important part, and all analogies lead to the conclu- sion that their constituent elements have been brought together by aqueous processes, such as we have already indicated. n. — SERPENTINES IN NORTH AJFERICA. § 16. It is evident that if we once come to regard ser- pentine as a rock formed from aqueous sediments of chemical origin, there is no reason, a priori, why it may not be found, like limestone, dolomite, or gypsum, inter- calated in stratified deposits at different geological hori- zons, and with different lithological associations. Several X.1 SERrENTINEa IN NOUTIT AMERICA. 485 such horizons of serpentine have been observed in North America, which will bo noticed in ascending order. Included in the ancient gneissic series to which the name of Laurentian has been given, serpentine is fre- quently met with, associated alike with crystalline lime- stone and with dolomite. In these, Mie serpentine is often disseminated in grains or small irregular masses, giving rise to varieties of so-called ophicalcite. These imbedded masses of serpentine are sometimes concretion- ary in aspect, and may have a nucleus of white granular pyroxene. They often recall, in their arrangen.ent, im- bedded chert or flint, and, like it, sometimes attain large dimensions. These berpentines occasionally include the calcareous skeletons of Eozodn Canadense, the silicate replacing the soft parts of the organism, as described by Dawson and Carpenter. Occasionally, the serpentines of this horizon form beds of considerable size, either pure or mingled only with small portions of calcite or dolomite. Of these, many instances are seen with the limestone^ of the Laurentifin in Canada, and a remarkable example occurs at New Rochelle, on Long Island Sound, near New Yoi'k city, where massive bedded serpentine, highly inclined, and interstratified with crystalline limestone, often itself mingled with serpentine, occupies a breadth of about 400 feet across the strike, the whole being con- formably interstratified with massive gneisses and black hornblendic rocks with red garnet.* The general charac- ters of the serpentines found with the Laurentian lime- stones have been elsewhere described by the present writer.f Their lower specific gravity, and generally paler colors, together with a larger proportion of combined * For an account of this locality, see Mather, Geo) F-irst District of New York ( 1842), p. 462; also J, D. Dana, Amer. Jour. Sci. (3), xx., 30-32. t For descriptions and analyses, by the t uthor, of Laurentian serpen- tines, see Geol. Canada, 1863, pp. 471, 591 ; also Contributions to the History of Ophiolites (1858), Amer. Jour. Sci. (2), xxvi., pp. 234-2.S6, 239. Much of this so-called serpentine belongs to the species retinalite, ante, p. 332. .itL. 436 THE GEOLOGICAL HISTOEY OF SERPENTINES. CX. water, serve, in some cases at least, to distinguish the serpentines of this horizon from those to be mentioned as occurring in the Huronian series. To this may be added a smaller amount of combined iron-oxyd, and, in most cases, the absence of compounds of nickel and chrome, which are almost invariably present in the latter. This distinction is probably not absolute, since chromite is said to occur in the serpentine of New Rochelle, and a chroraiferous garnet has been found in the Laurentian rocks in Canada. § 17. The serpentines next to be noticed occur in very different lithological associations from the last, and in a group of rocks which has been described under the name of Huronian. These may be defined as in large part greenish hornblendic schistose rocks, passing, on the one hand, into massive greenstones, diorites, or euphotides, and, on the other hand, into steatitic, chloritic, and hy- dromicaceous, or so-called tulcose or nacreous schists, some varieties of which resemble ordinary argillites, with quartzose layers, ofton with epidote, and with associated beds of ferriferous dolomite and magnesite. In this litho- logical group (already referred to, in § 8), which is now known to mark a definite geological horizon, the serpen- tines are found interbedded, sometimes mingled with car- bonate of lime or of magnesia, but seldom or never presenting varieties like the granular ophicalcite of the Laurentian. To this horizon belong the serpentines of eastern Canada, found in the continuation of the Green Mountain range, as well as those of Newport, Rhode Island, and apparently those of Cornwall, Anglesey, and Ayrshire, in Great Britain. The serpentines of this series are darker colored than the last, and generally contain small portions of chrome and nickel in combination, the former in part as chromite.* \^ <' \ * For an account of these serpertires, see Geology of Canada, 1863, pp. 472, ''()«~612 ; also Contributions to the History of Ophiolites (1858), Amer. Jour. Sci. (2), xxv., 217-226. ;m the ned be I, in and tter. mite ,nd a itian very 1 in a name ! part le one Dtides, id hy- ichists, s, with Dciated litho- is now serpen- ith car- • never of the tines o£ ) Green Island, yrsiiire, e darker portions in part nada, 1863, lites (1858), X.) SERPENTINES IN NOETH AMERICA. 437 § 18. Serpentines are also met with in eastern North America in somewhat different associations from the two foregoing groups, and apparently belonging to a third geological horizon. The determination of the precise stratigraphical relations of the serpentines in question presents, however, certain difficulties, ai'ising from consid- erations which will be made apparent in the sequel. Ser- pentine, though not exempt from sub-aerial decay, resistif this process better than hornblendic, feldspathic, and cal- careous rocks. Hence it happens that in regions Avhere these are decomposed and disintegrated to considerable depths, associated masses of serpentine may be found rising out of the soil, without any evidences of the pre- cise nature of the rocks which once enclosed them. Illus- trations of this condition of things are found in the vicinity of Westchester and of Media, in Chester County, Pennsylvania. The underlying rocks in this region are known to be chiefly gneisses, with hornblendic and mica- schists, and include what are believed to belong to two distinct series, both of which are well displayed in the section seen on the Schuylkill River, below Norristown. Here the older Laurentian gneiss, such as it appears in the South Mountain and the Welsh Mountain, comes up in Buck Ridge, while the newer gneiss and mica-schist series is seen succeeding it to the southward, at Mana- yunk and Chestnut Hill, at which latter locality it also appears on the north side of the narrow Laurentian belt. In this section, as it is exposed on the Schuylkill, a belt of serpentine, Avith steatitic and chloritic rocks, appears between the two series, but elsewhere it is wanting along the outcrop of the older gneiss. In tho localities farther west in Chester County, alreadj^ mentioned, at West- chester and Media, where the rocks adjacent to the ser- pentine are disintegrated, and have disappeared from decay, it cannot be determined whether these serpentine masses belong to the older or the newer series — which latter appears to be similar to that including "the serpen- fl -^TT mim 438 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. tine and chrysolite rocks of Mitchell County, North Caro- lina. (§ 123.) § 19. The serpentine of Brinton's quarry, near West- chester, Fenns^ivania, is distinctly bedded, granular, and often finely laminated, with disseminated scales of a mica- ceous mineral, giving it a gneissoid structure and aspect. A black schistose hornbleudic rock, with red garnet, is said to have been found in an excavation adjoining the serpentine, pnd fragments gathered in the vicinity showed thin interlaininations of black hornblende with greenish serpentine. TIk dip of the strata, of which several hun- dred fee' ;iie here exposed, s to the northwest, at a high angle, approaching the vertical. They are traversed, nearly at right angl'^s, by a vertical granitic vein, which has been traced for many hundred feet in a northwest course. This vein, wliich is generally from three to six feet in breadth, is white in color, and in parts may be described as a fine-grained binary granite, the feldspar of which is superficially kaoliuized. In other parts, it be- comes very coarse-grained, presenting large cleavage- forms of orthoclase. A banded or zoned structure, parallel to the well defined A\'alls, is observed in some parts, and in one case a lenticular mass of white vitreous quartz occupies tlie centre. This veinstone, which carries black tourmaline, and is said to have afforded beryl, has all the characters of the ordinary endogenous granitic veins found in the gneissic rocks of the Appalachians, which veins I have elsewhere described in detail.* § 20. The rocks in the vicinity of the serpentine near Westchester are, as already said, deeply decayed, but wherever seen in cuttings are found to be mica-schist and micaceous gneiss. Such rocks, with a northwest dip, appear to underlie, at no gi'cat distance, the mass of ser- pentine exposed at Stroud's Mill. Similar rocks are also found on the railroad between Westchester and Media, * Amer. Jour. Science (3), i., 182-187, and Chem. and Geol. Essays, pp. 192-200, also ante, p. 223, X.] SEIIPENTINES IN NOllTH AMERICA. 439 Geol. Easays, where tliey are exposed in a cutting nea^* the latter station, about a mile from which is found a great outcrop of distinctly stratified serpentine, resembling that of Brinton's quarry, and with a steep northwest dip. It includes an iuterstratified mass, about twenty feet thick, of a fine-grained reddish gneissoid rock, approaching leptynite or granulite in character, divided into distinct beds, generally from four to eight inches in thickness, between which are sometimes found layers of a few inches of a soft serpentine, and, in one case, of a broadly fcliated green chloritic mineral. Considerable differences in texture and aspect were observed between tlie serpen- tine beds below and those above this quartzo-feldspatliic mass, which is indigenous, tand not to be confounded witli the endogenous transversal mass described at Brinton's quarry. In the study of these rocks near Westchester, I was much aided by Dr. Persifor Frazer, who kindly accompanied me, and, from his previous labors in the geological survey of the district, was familiar with its details. § 21. Serpentine rocks also occur on Manhattan Island, in the city of New York, where they are still exposed between Fifty-seventh and Sixtieth Streets, west of Tenth Avenue, and are directly iuterstratified in gneissic and micaceous rocks, which may either belong to the older gneiss series of the Highlands, or to a newer group. Associated with the massive serpentine of this locality are found small quantities of a granular ophicalcite, and near it is a mass of anthopyllite rock. This locality was long since described by Dr. Gale, when the rocks were more fully exposed than at present.* § 22. Serpentine masses are also found in the vicinity of the last, on Staten Island, and at Iloboken, in both of which localities the encasing gneisses, seen in New York city, are wanting, and the serpentine api)ears along the eastern margin of the triassic belt of the region. The * Mather, Geology of the Southern District of New York, p. 401. ^ 1 i:i ^ti r » ill r^ ' a II Ml '5 I ^ * I I, 1 '. 440 THE GEOLOGICAL HISTORY OF SERPEINTINES. [X. serpentine of Staten Island is of much interest, as it presents many features which would seem at first sight to lend support to the view of its igneous origin. The serpentine rocks here occupy an area of a little over thir- teen square miles in the northern half of the island, and form a ridge, presenting a succession of rounded hills, from a mile and a half to two miles or more in width, fcxtending in a northeast and southwest course, with an average height of 200 feet, but rising in one part to 420 feet above the sea. Along the western base cf this ridge lie the red sandstones of the trias, but the contact of these with the serpentine is concealed beneath the soil. A long ridge of diabase rock, similar to tliut which pene- trates the trias on the west bank of the Hudson, runs through the sandstones for a length of nearly six miles, nearly parallel to the serpentine belt, and ai a distance of from half a mile to a mile. Along tlie sontiieru and east- ern borders of tiie serpentine are spread horizontal cre- taceous clays, partially overlaid by ilrift, while on the north side of the island, where tlip serj)eiitine hills rise abruptly at a little distance from the shore, arc the only known outcrops of other rocks; one, a ledge of anthophyl- lite rock like that .."conjianying the serpentine in New York c'ty, and jt.oIIu ■ , a few hundi'cd feet distant from the latter, and fr^ui tue serpentine, consisting of a coarse pegmatite, having all the aspect of an ordinary concre- tionary granitic vein, and containing besides ^^'ystals of orthoclase, sometimes twelve inches in length, small por- tions of a white triclinic felds])ar, and rare cr3'3tals of red garnet. A second, smaller outcrop of a similar kind is found near by. These granitic and anthophyllite rocks appear from beneath the water and the sands of the beach. § 23. Such an occurrence of serpentine, rising from out of the nearly horizontal and low-lying mesozoic strata of tlie island, was well calculated to sustain the notion of the e aptive nature of this rock which was put forth by [X. of by X.] SERPENTINES IN NORTH AMERICA. 441 Mather in liis description of this locality. He, in his report, above cited, included the serpentine in his " Trap- pean Division," in the same category with the adjacent eruptive mesozoic diabase, regarding the serpentine " as due to the action of the same general causes, modified in a manner unknown to us." * The history of this area of serpentine becomes intelli- gible when studied in the light of the facts already men- tioned above. It was apparently, in triassic time, a range of hills left by the disintegration of the adjacent gneiss, the lower-lying surfaces of whicl; are concealed beneath the ne '.'ji' sediments of the region. Since that time, as I have elsewhere pointed out,t the serpentine itself has undergone a process of sub-aerial change, as is evident by the layer of deca^-ed ma:ter, with included masses of limonite, which, in tliose portions that have escaped ero- sion, still covers the serpentine to the depth of tei. or twelve feet (ante^ p. 2G8). For many of the above 'letailw of this region, I have availed myself of a description of its geology, witli map and sections, published in 1880, by Dr. N. L. Britton,$ of the School of Mines, Coiumbin. College, New York, with whom I had, in 1883, the aiivni- tage of visiting this interesting locality, and to w- ,m I desire to make my grateful acknowledgments for valuable information respecting it. § 24. The serpentine rock which seen at Castle Ilill, Hoboken, on the west bank of the ' ,dson, opposite New York city, is believed by Dr. Brittoii to be a continuation of that of Staten Island, and, like ,i, lies on the eastern border of the trias ; while the soriientine outcrop on the west side of New York city h x strike which v/ould carry it to the east of Staten Island, and probably corre- sponds to a repetition of the same belt. Gneissic rocks are * Loc. cit., p. 283. [For a farther notice of this serpentine, seepos^, Essay XL, § 178.] t Amer. Jour. Science (3), xxvi., 206, X Tlie Geology of Kiclunoml County (Staten Island), N. Y., Ann. New York Academy of Sciences, Vol. II., N , 'i. liili 1 r'\M n iis 'rm 1 ' I r'-P n ' V i) 'I ^4 442 THE GEOLOGICAL HISTOllY OF SERPENTINES. [X. met with in a boring near the serpentine at Hoboken, and are found in the small islands between Manhattan and Staten Islands, so that there can be no reasonable doubt that the serpentines of Staten Island and of Hoboken be- long, like that of New York city, to the gneissic series of the region. The determination of the precise relations of these gneissic rocks to those accompanying the serpen- tines of eastern Pennsylvania, already described, remains for farther inquiry. See Essay XL, § 187. § 25. We have next to notice the occurrence, in Penn- sylvania, of serpentine in the Lower Taconic rocks of Emmons, the Primal slates of Rogers, which he supposed to belong to the horizon of the Potsdam of the New York series. In accordance with this view, we find that in a report by Genth on the mineralogy of Pennsylvania, in 1875, the occurrence of serpentine is mentioned, though without any details, in the "Potsdam sandstone" near Bethlehem, at the iron mines of Cornwall, and also in the township of Warwick, Chester County.* This statement is, however, misleading, inasmuch as the serpentine is not found in the sandstone which has been conjectured to be the equivalent of the New York Potsdam, but in certain schists and limestones, which have been referred to that geological horizon, — namely, the so-called Primal slates. The history of these is given at length in Essay XL § 26. I have had an opportunity of observing the occurrence of serpentine at Cornwall, where it forms small, irregular masses disseminated in a bed of crystalline limestone, itself subordinate to the great mass of crystal- line schists which include tiio magnetite largely mined at this locality. Serpentia^, generally with limestone, is found at many other looaUties associated with iron-ores at the same geological horizon, as at Fritz's Island and else- where near Reading, at Boyerstown, and at the Jones iron mine, near to Warwick, where it is found in small, lenticular masses imbedded directly in the crystalline * Second Geological Survey of Penn., Report B, p. 115. -' 'fe'-i'-i^ aw SEIUENTINES IN NORTH AMERICA. 443 schists, whicli, as at Cornwall, include the cupriferous rjagaetites of tlie region. These schists include hydrous mieacecus minerals, among which are chlorite, and the greenish foliated silicate of copper, magnesium, and alu- minium, to which I have given the name of venerite (ante^ p. 357). The manner in which lenticular masses of pure serpentine, sometimes only a few ounces in weight, are found imbedded in these schists, not less than the mode of their occurrence in the limestones at this horizon, is such as to suggest very forcibly the notion that they have been formed under conditions not unlike those which have given rise to chert or to iron-stone nodules. No large masses of serpentine have, so far as known, been found at this horizon, yet they may be expected. § 27. We have next to notice the existence of a bed of serpentine at Syracuse, New York, which was, in 1839, examined and described by Professor Vanuxem, then en- gaged in the geological survey of the State. The locality, "on the Fort-Street road, to tlie east of Syracuse," or, according to Dr. Lewis Beck, "on the hill, a short distance east of the mansion of Major Burnet, at Syracuse," has long since been concealed by the growth of the city, and we have, so far as I am aware, no other description than those given by Vanuxem, in the years 1839 and 1842,* of which, on account of the interest and significance of this curious occurrence of serpentine, I make the following summary : The rocks of the region, as is well known, belong to the Onondaga salt group of the New York series, and occupy a position near the summit of the Silurian, being overlaid by the Lower Helderberg, and resting upon the Niagara division. The strata are, as elsewhere throughout this region, undisturbed and nearly horizontal, the inclination at Syracuse, as measured by Vanuxem, * Vanuxem, Third Annual Report on the Geology of the Thu-d Dis- trict of New York, pp. 260 and 283; also Final Report on the Geology of the Third District, pp. 108 and 110, and Beck's Mineralogy of New York, p. 275. 'i!--'m 444 THE GEOLOGICAL HISTORY OF SEHrENTINES. [X. * Ii ! 'i 1] J I < being less than thirty feet to the mile, in a southwest direction. The thickness of the Onondaga salt group is subject to gi-eat variations, and at this point, not far from its eastern limit, it is thinner than farther west. It is described by our author as here cojisisting, in its lower portion, of a mass of red shales, " .«rying from 100 to 500 feet in thickness, passing upward into a body of greenish shales including more or less gypsum, and followed by a third division, in which are found masses of gypsum of economic value. § 28. These occur on two horizons, one at the base and the other at the summit of the division, in the form of lenticular masses included in soft shales or marls, which are often marked by hopper-shaped cavities, doubt- less formed through the removal, by solution, of imbedded crystals of sea-salt. Interposed in these marls is found a peculiar porous dolomite, generally drab or buff in color. The cavities in this are very irregular in form, and in most cases communicate with one another. They are sometimes spherical, and contain crystalline crusts, besides some pulverulent carbonate .of lime. They also vary greatly in size, in some portions attaining a diameter of half an inch, and giving the rock a vesicular aspect. Our author remarks, " The cavities of these porous rocks have no analogy whatever with those derived from organic remains." As seen in one locality, "the cells show that parts of the rock are disposed to separate into very thin layers which project into the cells, an effect wholly at variance with aeriform cavities, but evidently the result of the simultaneous forming of the rock and of soluble minerals, whose removal caused the cells in question " ; a condition of things which Vanuxem considers analogous to that shown by the hopper-shaped cavities in the asso- ciated marls. § 29. The distribution of this porous dolomite in the third division of the Onondaga group near Syracuse is somewhat irregular. Besides a well defined stratum ex- l-im • -I X.] SERPENTINES IN NOIITII AMERICA. 445 [X. iwest up is from It is lower to 500 •eenisl^ d by a sum of le base he form • maris, 5, dovibt- nbedded , found a in color. , and in fbey are s, besides ilso vary uneter of ect. Our ocks have n organic show that very thin wholly at the result of sohible [^l^uestion " ; . analogous II the asso- uite in the Syracuse is stratum ex- tending over a large part of the gypsum-bearing region, and from three to four feet in thickness, Vanuxem noticed other "masses, limited in extent, without fixed positions, appearing to have been deposited at irregular intervals in the marls " ; while in some places, as at the serpentine locality about to be described, there is a lower mass, with smaller pores than that above, sometimes attaining a thick- •^-^s of twenty feet. The interval between the upper and jiuwer gypsum-horizons, from various sections noticed by Vanuxem, would appear to be from forty to fifty feet. The marls found in this interval contain more or less dis- seminated gypsum, and in some cases small grains or crys- talline masses of sulphur, and more rarely crystalline plates of specular iron in druses in the dolomite, as ob- served and shown me by Dr. Goessmann. The marls are described as yellowish or brownish in color, and generally soft and shaly, with harder masses included. Above this gypsiferous division, is a fourth, consisting of a compact magnesian limestone, marked by the presence of numer- ous small needle-shaped cavities, which forms the summit of the Onondaga group. § 30. It is, as already stated, between the two masses of porous dolomite near Syracuse that the bed of the ser- pentine was observed. Its thickness is not stated, but it was said to extend northward " for many rods." Accord- ing to the original notes of Vanuxem, there was seen, in ascending the hill, after passing twenty feet of the lower porous dolomite, and an interval concealed by soil, " first, a marly shale, then mixtures with more carbonate of lime, some compact, some crystalline, some confusedly aggre- gated, presenting cavities lined with crystals of that min- eral, and containing also sulphate of strontian in the mass and in the cavities. With these, and above these, are other aggregates like serpentine, marble, etc., with pur- plish shale or slate, which are followed by a green and blackish trap-like rock, as to appearance, but too soft for that rock." After this, — that is, above it, — is a mass ' . » a Wt'ia 446 THE GEOLOOICAL HISTORY OF SERrENTI.NES. [X. i'fi' * i' :;^ 1^ V. :^h:- I i which resembles the material overlying the lower beds of gypsum, and this last is covered b}'^ the upper porous dolomite. § 31. In a supplement to the report of 1839, above quoted, it is added, "The green and trap-like rocks ob- served near the top of the hill to the east of Syracuse, have been examined so far as time would admit. They are all 8eri)entines, more or less impure, and of various shades of bottle-green, black, gray, etc. They all pro- duce sulphate of magnesia with oil of vitriol. . . . Some have a peculiar appearance, like bronze, owing to small gold-like particles, with a lamellar structure, resembling bronzite or metalloidal diallage ; also other particles, highly translucent, like precious serpentine, with fre- quently small nuclei, resembling devitrifications or porcel- lanites, colored white, yellow, blood-red, variegated, etc. The grain of this is like common serpentine. In other kinds, the mass seems to be made of small globuliform concretions, varying in size, being centres of aggregation. Some are of dark vitreous serpentine, others of the com- pact kind, the enveloping part of a light color." Van- uxem's farther notes, in his final report, add some impor- tant details to the above. He says: "The great mass of entirely altered rock is a well characterized serpentine, especially when examined by the microscope." He men- tions, moreover, the occurrence of mica, both white or light-colored and black, besides accretions which he com- pares to granite, and others in which a hornblende takes the place of mica, forming aggregates resembling syenite. He also describes granular carbonate of lime, like marble in texture, which " existed as accretions or nodules envel- oped in the serpentine." § 32. I endeavored many years since to obtain speci- mens of these rocks, and, through the kindness of Prof. James Hall, secured a single mass of the ser])entine, which contained small plates of a copper-colored bastite or bron- zite. Neither mica, hornblende, nor an}' other crystalline X] RKIIPENTINES IN NORTH AMERICA. 447 silicate, was, however, present in the mass, which was n well defined serpentine, with some iidniixture of carhon- ates. It agrees closely with the description given by Vanuxem, being an aggregate of grains and rounded niassr' .f serpentine, with others of a fine-grained carbon- ate of lime, imbedded in a greenish-gray calcareous base. The colors of tiio serpentine vary froi , blackish-green to greenish-white; it is often translucent, a ikI takes a high polish. An average portion of this rock gave to acetic acid, 34.4P) parts of carbonate of lime, and 2.73 of carbon- ate of magnesia, with 0.34 of iron-oxyd and alumina, leav- ing a residue of 62.50 of insoluble silicate. This was a nearly pure serpentine, as shown by its analysis. It was completely decomrosed by sulphuric acid, and gave silica, 40.67; magnesia, 32.61; ferrous oxyd, 8.12; alumina, 6.13; water, 12.77=99.30. No traces of either chrome or nickel could be detected. One of the small imbedded calcareous masses or concretions found in this serpentine was finely granular, greenish in color, and was nearly pure carbonate of lime.* § 33. The associated shales and limestones of this gyp- sum division are, however, generally, if not always, highly magnesian. Beck found twenty per cent of magnesia in the limestone overlying the lower range of gypsum-beds, and the precisely similar rocks associated with the gypsum at the same horizon in Ontario are dolomitic, the porous or vesicular beds being nearly pure dolomite, and other specimens of the limestones and shales consisting of dolo- mite with an argillaceous mixture, the latter sometimes predominating.f § 34. From a study of the facts before us, it is apparent that we have here evidences of the formation by aqueous deposition of a bed of concretionary silicate of magnesia, taking the form of serpentine, with a little associated bas- * For details of this serpentine and its analysis, see Amer. Jour. Sci- ence (2), xxvi., 203, and Geology of Canada, 1S03, p. 635. t Geology of Canada, 1868, pp. 347, 625. IMAGE EVALUATION TEST TARGET (MT-3) t. '% 4 ^**% .AV.4. A Ua u.. 1.0 I.I •jrlM IIM 4.0 12.0 2.2 1.8 IL25 ill 1.4 IIIIII.6 m/ e 7a 7 'c>l ^; ¥ Photographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716)873-4503 m 1 1 448 THE GEOLOGICAL HISTORY OF SERPENTINES. PC. tite or bronzite, and probably some otlier crystalline sili- cates. The intimate association of silicate of magnesia with carbonate of lime is significant when it is considered that the magnesia which abounds in the accompanying strata is in the form of dolomite, and serves to illustrate the views set forth in § 13, as to the relation between the car- bonates and silicates of these two bases. It seems probable that we have in this deposit the results of some spring bring- ing to the surface, in this locality, waters holding in solu- tion calcareous or alkaline silicates, which have given rise to a silicate of magnesia, in accordance with the reactions already explained. It is to be hoped that farther re- searches at this geological horizon may disclose other localities of magnesian silicates similar to that of Syracuse. § 35. We may recall in this connection some facts about the occurrence of magnesian silicates in other geo- logical periods more recent than that of Syracuse. De- posits of sepiolite, a hydrous silicate approaching to steatite in composition, are well known in the tertiary strata of the Paris basin, in Spain, and elsewhere along the Mediterranean. I have long since described some of these deposits, and have discussed at length their chemi- cal and geological relations.* Mention should here be made of the talc found with the anhydrous sulphate of lime (karstenite) in the schists at the Mont Cenis tunnel, to be mentioned farther on (§ 62), and also of the associ- ation of gypsum and serpentine in the crystalline schists of Fahlun, in Sweden.f Freiesleben, and, after him, Fra- polli, has described the occurrence of a magnesian silicate which occurs frequently in the mesozoic gypsums of Thuringia, in nodular imbedded masses resembling flints in their aspect and mode of occurrence, but composed essentially of a soft magnesian silicate, near to talc in composition, and colored bcown with bituminous matter.^ • Amer. Jour. Science [2], xxix., 284; and xxx., 286. t See the author's Chem. and Geol. Essays, p. 336. ' t Bull. Soc. Geol. de France, 1847 [2], iv., 837. »• X.] SERPENTINES IN EUROPE. 449 ni. — SERPENTINES IN EUROPE. § 36. Having thus passed in review some of the princi- pal facts Icnown with regard to the occurrence of serpen- tines in North America, we proceed to the consideration of the same rocks in different parts of Europe, where, as shown in the opening sections of this essay, thoy have long been objects of study, and have been alternately regarded as indigenous and as exotic in character. The hypothesis of the igneous and eruptive origin of serpentine is well illustrated in the paper by Professor Bonney on the serpentines of Cornwall, England, in the "Quarterly treological Journal," for November, 1877, supplemented by his later observations on the geology of that region, communicated to the Geological Society of London in November, 1882, and published iu abstract in the "Geological Magazine," .for December, 1882; in which connection should also be consulted his paper on Ligurian and Tuscan serpentines, in the same magazine, for August, 1879. § 37. Bonney at first accepted the then generally re- ceived opinion that the crystalline schists in which the serpentines of Cornwall are included, are altered paleozoic strata, but in his latest studies of the region he announces the conclusion that the}' are not paleozoic, but eozoic (archaean), and consist of a great series, divided into three groups. The lower one, of greenish micaceous and horn- blendic schists, he compares with those of Holyhead, An- glesey, and the adjacent shores of the Menai Strait, in Wales. The rocks of these localities, belonging to the Pebidian series of Hicks, have been examined b}'^ the present writer, and by him compared with the Huronian of North America.* § 38. Above these greenish schists in Cornwall, accord- ing to Bonney, is a black hornblendic ^roup, and a still higher gvanulitic group with granitic bands ; the charac- * Amer. Joui. Science, 1880, vol. xix., pp. 276, 281 ; and ante, pp. 416-419. 450 THE GEOLOGICAL HISTOKY OF SERPENTINES. [X. !i il ters of tliese two recalling portions of the Montalban or upper gneissic series of North America and of the Alps. It is in the lowest of these three divisions, consisting chiefly of micaceous and hornblendic schists, that the Cornish serpentines appear, accompanied by so-called gab- bros or greenstones. Bonney finds, with Boase and with De la Beche, examples of apparent interstratification and passage between these rocks and the schists, but con- cludes, nevertheless, that there is evidence that the ser- pentine was introduced after the crystallization of these, and that its eruption was followed by that of gabbros of two dates, aiid subsequently by that of granitic and dark- colored trappean rocks. He throws doubt upon the an- cient hj'-pothesis of the conversion of hornblendic and pyroxenic rocks into serpentine, and supposes this mineral species to have resulted from the hydration of an olivine- rock, such as Iherzolite, which consiiits essentially of oli- vine with enstatite ; grains of both of which species may be detected by the microscope in thin sections of some of the Cornish serpentines. According to John Arthur Phillips, some of the so-called greenstones of Cornwall are eruptive, while others are undoubtedly indigenous, and graduate into the crystalline schists of the region. Respecting these, the writer said, in 1878, " These bedded greenstones, with their associated crystalline schists, ap- pear to have strong resemblances to the rocks of the Huronian series, to which farther study will probably show them to belong." * § 39. Bonney has also extended his observations to the serpentines and associated rocks in Italy, which we have included uader the general title of ophiolites. This name, and the kindred one of ophites (Greek, dcpirr/g')^ alluding to their greenish color, resembling that of the skins of some serpents, has been extended so as to include both true serpentine and the frequently associated rocks which present some analogies with it in color. In fact, we * Harpers' Annual Record for 1878, p. 308. ..^ ,0E» X.] SERPENTINES IN EUROPE. 461 3an or Alps, sisting at the ;d gab- id witli on and at con- :he ser- f these, )bros of id dark- the an- idic and i mineral 1 olivine- ly of oli- icies may I some of Arthur Cornwall digenous, ,e region. ie bedded ihists, ap- tcs of the probably rations to [which we Ites. This lat of the [to include ited rocks lln fact, we pass from pure serpentine, and admixtures of this with car- bonates, to serpentinic rocks including more or less of dial- lage, bronzite, or bastite, and thence to aggregates in which an admixture of these with a feldspathic element marks a transition to the great group of rocks essentially made up of an anorthic feldspar with a pyroxenic element (hornblende, pyroxene, enstatite, etc.), including the so- called "greenstones," — diorites, diabases, and eupho tides, — which are the frequent associates of serpentines. All of these rocks were embraced by Savi under the conve- nient name of the ophiolitic group. § 40. The name of gabbro (from an Italian locality of these rocks, near Leghorn) was adopted and extended by Tozzetti, in the last century, in a similar sense. His numerous species of gabbro embraced alike serpentine and the various diallagic, hornblendic, and feldspathic rocks already noticed, of which the red gabbro, or gabbro rosso, seems but a locally discolored and partially decayed form. The name of gabbro has come, with many lithologists, to mean a diabase ; but it is employed in such a very indefi- nite manner that it would be well if it were dropped alto- gether from use.* It is often made to include the granitone of the Tuscan stone-workers, the so-called euphotide, in which, as we are told, the feldspathic element is replaced by saussurite. Although this latter term is often given to a compact variety of triclinic feldspar, the true saussurite is, as I have elsewhere shown, a compact zoisite, distin- guished from feldspar by its much greater density and hardness. ' The two minerals are, however, intimately associated in the euphotides alike of the Alps and the Apennines, as seen in specimens which I have examined both from Monte Rosa f and from Monteferrato, where I found saussurite in 1881. . * See, in this connection, Cocchi, Bull. Soc. G^ol. de France (1856), xiil., 261; also his valuable memoir on the Igneous and Sedimentary Rocks of Tuscany, ihid., 1861, pp. 227-300. Cocchi was a pupil of Savi. t Contributions to the History of Euphotide and Saussurite, Amer. Jour. Science, 1858, xxv., 437. " ' 452 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. § 41. The results of Bonney's studies are given in a paper on Ligurian and Tuscan serpentines in the " Geo- logical Magazine," for August, 1879. He therein records his observations in different localities in these regions, which, for reasons to be made apparent farther on, we arrange in three geographical groups. First, ophiolites on the sea-coast west of Genoa, where Bonney describes the serpentines as occurring with dark-colored schists and gabbros, instancing among the mineral species found with them pyroxene, amphibole, glaucophane, chlorite, and saussurite. He states that the serpentines of this region are so like those of Cornwall that he feels justified in claiming for them a similar origin. In a second group, he notices the serpentines of a region immediately eastward of the first, between Genoa and Spezzia, which he de- scribes as very similar to these. Bonney rejects for all of these serpentines, as for those of Cornwall, the notion that they have been formed by metasomatosis from diorite, diabase, or hornblendic rocks, a hypothesis which he con- ceives to have been founded on hasty and imperfect generalizations, and regards them as generated by the hydration of intruded olivine rocks. In the third geo- graphical group of the ophiolites described by Bonney, he places those of Monteferrato in Prato, near Florence. In each of these districts he notices the close resemblances between the ophiolitic rocks and those met with in the similar areas in Great Britain, and supposes an intrusion of serpentine, or rather of olivine rock, among crystalline schists, followed by a later intrusion of gabbro. He has no hesitation in assigning to the serpentines of these three districts similar conditions and origin to those in Corn- wall, North Wales, and Scotland, remarking that, notwith- standing the fact that the Italian serpentines are, in part, at least, assigned to the cenozoic period, " they are practi- cally identical " with the serpentines and gabbros of more ancient times. § 42. Bonney further calls attention to the breccias of X.] SERPENTINES IN EUROPE. 453 serpentine with a calcite cement, found at various points with these Italian serpentines, and concludes that the ser- pentines have been brecciated in situ, so that it is possible to trace, in a short distance, the passage from unbroken or slightly fissured blocks to completely crushed and recemented fragments, and even to mixtures of finely broken serpentine cemented by carbonate of lime, in which he notes, here and there, filmy patches of a serpentinous material, as if it had been redissolved and again deposited. He believes that the crushing took place after the rock became a serpentine. The correctness of these views of Bonney, as to the breccias, I can confirm from my own observations in the same regions, and also from my studies of similar breccias, accompanying the ophiolites of eastern Canada. Gastaldi, in this connection, has made an impor- tant observation of a breccia in the valley of Trebbia, resting upon a diallagic serpentine, and consisting of cemented fragments of silicious and argillaceous slate with limestone (alberese), the paste being traversed in various directions by veins of chrysotile.* § 43. Bonney 's observations thus bring us face to face with the views of those Italian geologists who regard certain of these serpentines as of tertiary age, and speak of them as having had an eruptive origin, although, as we shall see, their views of the genesis of these rocks differ as widely as possible from those of Professor Bonney. In anticipation of the International Geological Congress at Bologna, in Se])tember, 1881, the Italian geologists had, under the direction of the Royal Geological Commission (R. Comitato Geologico), made extraordinary preparations for the study and the full discussion of the problems offered by the serpentines of Italy. A map, prepared for the occasion, was published, showing the localities of the ophiolitic masses for the whole kingdom on a scale of l-l,lll,lllth ; besides separate maps of particular regions on a scale of l-10,000th, as that of Mazzuoli and Issel for * Studll geologic! sulle Alpi occideutali, parte II., p. 61. I II i II .I'?!! ilpll 454 THE GEOLOGICAL HISTORY OP SERPENTIITES. PC the Riviera di Levants in Liguria, and that of Capaoci for Moiiteferrato in Prato, in Tuscany ; with especial memoirs on these districts, also published by the R. Comitato Geo- logico, in 1881. Ophiolitic rocks are met with in greater or smaller outcrops in many localities from the Alps, throughout the Apennines, and as far as Calabria. To these, the studies of Taramelli, Lovisato, De Giorgi, and De Stefani, among others, in addition to those previously named, have contributed a great body of information. A collection of ophiolitic rocks from various localities was also made, and submitted to chemical and microscopical study by Cossa of Turin, aided by Mattirolo, the results of which occui^y about 200 pages, illustrated with many plates, in the fine quarto volume recently published on Italian lithology.* § 44. During the International Geological Congress, a special meeting was held for the discussion of the question of serpentines, on Sept. 30, 1881, in which took part Tara- melli, Capacci, Zacagna, Sella, Szabo, Daubrde, De Chan- courtois, and the writer, who presided on that occasion. A detailed report of the proceedings at this meeting is published in the first fasciculus of the Bulletin of the new Geological Society of Italy, pages 14-31, followed by an address on the general subject of serpentines by the pres- ent writer, pages 32-38, by notes on the same subject by De Chancourtois, pages 39-44, and finally by the extended studies of Taramelli on the Italian serpentines, pages 80- 128. It is impossible to speak too highly of the zeal, the industry, and the scientific spirit exhibited by the Italian geologists in these researches undertaken for the solution of the question of the ophiolites, which may well be held up as an example to be followed by other nations in similar circumstances. § 45. Mention should also be made of the brief memoir, of thirteen pages, in the French language, by Pellati, * Eicerche Chimiche e Microscopiche sui Roccie e Mineral! d'ltalia, Torino, 1881. 9 SERPENTINES IN EUROPE. 455 prepared for the Geological Congress, entitled " fitudes 8ur les Formations Ophiolitiques de I'ltalio," in which are set forth, with great conciseness, the principal facts with regard to the geography and the geo ogy of these ophiolitic masses, and the theoretical views entertained with regard to them by various Italian geologists. According to De Stefani, whose discussion is confined to the ophiolites of the Apennines, these rocks belong to three distinct hori- zons: — 1. upper eocene; 2. upper trias; 3. paleozoic; none of them pertaining to a more ancient period. These ophiolitic rocks form zones and regular beds in the midst of the sedimentary rocks, and in no case plutonic dikes. The different varieties of serpentine, and of the non-sedi- mentary rocivs which accompany it, are themselves found in regular alternating bands.* The conception of this observer as to the mode of eruption of these rocks appears to be essentially the same as that of Issel, Mattirolo, and Capacci, to t splained farther on (§§ 91-93 and 100). . § 46. The more recent studies of the R. Comitato Geo- logico, as announced in 1881, lead them to reject the views of De Stefani as to the age of the ophiolites, and to refer the whole of these rocks in Italy to two geological periods. They distinguish ancient serpentines, probably pre-paleozoic, and younger serpentines, referred to the ter- tiary. The older serpentines appear in large masses to the west of Genoa, between the valleys of the Polcevera and the Teiro, and from thence are traced to Monviso, from which point the ophiolitic group passes north-north- east to Monte Rosa, and thence, by the canton of Ticino, to the Vjiltelline. To the same ancient series are also referred the serpentines of the north of Corsica, those of Elba in part, and those of northern Calabria. These ancient serpentines, according to Pellati, follow the con- tour of the great zone of old gneissic and grnnitic rocks, which passes along the Alps, through Corsica and the Tuscan archipelago, and "re-appears in Calabria. The * Boll. Soc. Geologica Italiana, i., pp. 20-33. ! I 45G THE GEOLOGICAL HISTORY OP SEUPENTINES. [X. older geologists, Collcgno, Parcto, and Sisniondi, regarded the serpentines of the areas thus defined (in common with the others yet to be mentioned), as having been erui)ted, like granites, jjorphyries, and basalts, at various geological ages. Gastaldi, however, as early as 1871, assigned the Alpine serpentines to a distinct pro-paleozoic horizon, which, from the association of the serpentines with vari- ous rocks known as greenstones, or pictre venli, he desig- nated as the pietre-verdi zone, and compared with the Iluronian of North America, of which he supposed it to occupy the horizon. § 47. The conclusions of Gastaldi as to the Ali)ine ser- pentines have, according to Pellati, been confirmed by Baretti, and by Taramelli, the latter of whom clearly shows that the view held by many, that the rocks of the pietre verdi are carboniferous or triassic, is inadmissible, and that they belong, as maintained by Gastaldi, to pre- paleozoic or eozoic time. All of the ophiolitic maswes west of the meridian of Genoa, as well as those of north- ern Calabria, are by Pellati included in this class. To the east of this meridian, according to Pellati, we find the newer or tertiary serpentines, including, first, those of eastern Liguria, which have their greatest devel- opment along a line ruiniing north-northwest from Spezzia, and second, those of the Bolognese Apennines, consisting of a great number of small masses scattered between Florence and Reggio, in Emilia. A third group includes the masses of serpentine found between Grosseto and San Miniato, in addition to which tertiary serpentines are indicated in Elba and in the upper part of the valley of the Tiber. Fai fher south, others are met with at Lago- negro in the Basilicate, from which point to Neopoli a remarkable development of serpentines is found along the upper part of the valley of the Sinni. The areas of ser- pentines thus indicated by Pellati are, according to him, generally found in the midst of the limestones, argillites, and sandstones of tjie eocene, except in the case of those X.] ROCKS OP THE ALPS. 457 b'^tween Grosseto and San Miniato, tho oiitorops of which are often Been rising out of pliocene chiyu and sands. IV. — GEOLOGY OF THE ALPS AND THE APENNINES. § 48. Before proceeding farther in tlie discussion of tlie Italian serpentines, it will be well to get a view of the present state of our knowledge of Alpine geology, and especially of the conclusions and generalizations of Gas- taldi. These, so far as the Alpine serpentines are con- cerned, are, as we have seen, accepted by the Comitato Geologico, and, this conceded, it is diflicult to escape his wider generalization which brings the whole of the so- called tertiary serpentines of Italy into tlie same eozoic horizon with those of the Alps. If we go backward to the early history of Alpine geol- ogy we shall therein find the origin of the well known hypothesis that the crystalline stratified rocks are but portions of paleozoic. or more recent sediments which, in certain parts of their distribution, have undergone a pro- cess of alteration or so-called metamorphism. The iufn - position of the uncrystalline to the crystalline rocks in Mont Blanc, first noticed by De Saussure, was thus ex- plained by Bertrand, who suggested that these crystalline schists were altered rocks of a more recent date than the uncrystalline mesozoic strata of Chamonix. This notion was adopted without critical study by Keferstein, INIurchi- son, Lyell, Studer, Sismondi, and Elie de Beaumont, among others, till it was generally believed that the crys- talline rocks of the Alps are wholly or in great part of mesozoic and cenozoic age. It is hardly necessary to say that this hypothesis in the Alps, as elsewhere, was based upon false stratigraphy. I have elsewhere discussed it in its relations to Alpine geology, in a review of the great work of Alphonse Favre,* whose life-long studies in the Alps of Savoy have shown for all that region the fallacy * Araer. Jour. Science, 1872, vol. iii., pp. 0-10, and Chem. and Geol. Essays, pp. 337-330. . :i!l ' 1 I In 458 THE GEOLOGICAL IIISTOUY OP lEilPENTINES. [X. of the nietaniovpliic hypothesis. The fnrther studies of Gerlach, of Fr. Von Huuer, of Biirctti, and especially of Gastaldi, have now fully established the great anti(iuity of the crystalline rocks in question, and have enabled us to compare them with the pre-Cambrian rocks of other regions. It is not here, however, the time nor the place to discuss this (question, except so far as is necessary to the understanding of the geological relations of the Italian serpentines. § 41). The work of Gastaldi, interrupted by his death in 1878, was unfortunately left incomplete. We have, however, valuable records of it in a memoir in two parts, published in 1871 and 1874, entitled, "Studii geologici sulle Alpi occidental! " ; in a letter to De Mortillet in 1872 ; in one to Zezi in 1876, ^nd finally in one to Sella in 1877, with a postscript in 1878.* These various papers are illustrated with numerous maps, plans, and diagrams. In attempting to gather from these sources a brief state- ment of Gastaldi's conclusions as to the geology of the Alps and the Italian peninsula, I feel that I am both ren- dering a veritable service to science and paying a tribute to the memory of my honored friend and correspondent of many years. § 50. The "Studii," etc., contain, besides Gastaldi's own descriptions and sections, many important historical details and extracts from the literature of the subject. In the second part will also be found reproduced two en- graved sections, the one by Gerlach, from Monte Rosa, by Varallo and the Lago di Orta to Arona on Lago Maggiore, * Studii geologici sulle Alpl occldentall ; meraorle del Kegio Comitato Geologico, vols I. and II.; Deux mots sur la g^ologle des Alpes cotti" nes; lettre i\ M. de Mortillet, Comptes Rendus de I'Acad. des Sciences de Turin, vol. vll., 28 avril, 1872; Lettere del Prof. B. Gastaldi all' Ingeg- nere P. Zezl, Boll, del R. Com. Geologico, 1876; Sul rllevaraente geolo- gici fattl nelle Alpl Plemontesl durante la carapagna del 1877, lettere del Prof. Gastaldi al Presidente Qulntlno Sella; Reale Accademla del lilncel, memorle della classe dl sclenze flslche, ecc. anno CCLXXV. (1877-78). See, also, the writer, in Azoic Rocks, p. 245, and Cbem. and Geol. Essays, pp. 336,347. . jii ..111 ■ X.] EOL'KS OF THE ALl S. 459 and the other by CaiU) Neii, from the same point, in a course more to the soutlieiistwartl, by Valsesia, to Monte Fenera, and beyond.* A comparison of these sectiona with those described by Gustaldi, will be found of much value for the elucidation of the (luestiona before us. Starting frcm the granitic gneiss of Monte Rosa (the cen- tral gneiss of Von Iluuer, and the ancient gneiss o( Ger- lach and Gastaldi) we find in Neri's section a breadth of not less than seven kilometres included in the zone of the pietre verdi, and described as a stratilied series of "ser- pentines, talc-schists, etc.," followed by seven kilometres additional, designated as diorites ; the two being classed together as a "protozoic terrane." To this succeeds a breadth of not less than fourteen kilometres occu^)ied by what is described as a more recent crystalline terrane, conjecturally referred to the paleozoic period, and consist- ing of calcareous schists and quartzites, with mica-schists, and a great mass of intruded granite. Succeeding this is a great breadth described as porphyry or porphyritic con- glomerate, followed by limestones and dolomites, all of which are referred to the trias, and appear in Monte Fenera, succeeded by fossiliferous liassic and tertiary strata. The section by Gerlach, from Monte Rosa to Arona, shows above the ancient central gneiss a great breadth described simply as diorite, having at its base a thin belt of micaceous schists, and above it, between Varallo and the lake of Orta, a wide extent of recent gneiss and granite, followed, to the east of the lake, by gneissic mica-schists, succeeded by porphyry, until we reach the dolomitic limestone at Arona. § 51. Coming now to Gastaldi's own sections, we have one from Turin passing westward to the Fre:ich frontier, and crossing a broad mass of the ce.i.tri>l gneiss; co the * The section by Gerlach is probably from his Karte der Penninischen Alpen; Nouv. M^tn. de la Soc. Helvet. do Sci. Nat., 18G9. That by Neri is from the Boll, del Club Alphio, vol. viil., No. 22, Torhio, 1874. !] mm :i:'M 460 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. west of which, in a distance of forty kilometres, we have, first, three and a half kilometres of euphotide and serpen- tine, followed by about the same breadth of mica-schists, calcareous schists, and diorites, and finally by a great extent of calcareous schists, with numerous intercalations of serpentine and, towards the summit, gypsum and dolo- mite. The less complete section, to the eastward of the central gneiss, shows also the serpentinic and dioritic rocks overlaid by mica-schists, and the same story is repeated in other sections. Subsequently, in his letter to Zezi, Gastaldi describes and figures a section from Monte Bracco through Monviso and Monte Pelvo, along the upper part of the valley of the Po, and the valley of Varaita, to the frontier. His conclusions from the study of all these sections may be thus summed up : The crystalline rocks of the Western Alps are classed in two great groups, the lower of which (the central gneiss of Von Hauer) was described by Gas- taldi as the ancient gneiss, and by him compared with the Laurentian of North America. It consists chiefly of a highly feldspathic granitic gneiss, sometimes porphyritic or glandular, and includes bands and lenticular masses of quartzite and crystalline limestone, with white steatite, and graphite. Reposing upon the ancient gneiss, is a great and complex group, designated by Gastaldi as the "newer crystalline series," wliich, from the frequent pres- ence therein of serpentines, diorites, diabases, and related rocks of a greenish color, is also called by him the zone of the greenstones, or the pietre verdi. § 61. In a generalized diagrammatic section which accompanies Gastaldi's last published statement (his letter to Sella, in 1878), the first division of the newer crystalline series is described as a great mass of serpen- tine, followed by a second division consisting of eupho- tide, succeeded, after an interval of crystalline schists, limestones, and gneissic rocks, by a series made up of many alternations of epidotic, dioritic, and variolitic \m X.] ROCKS OF THE ALPS. 461 schists, with green steatite. In some localities Jire found great beds of Iherzolite and of amphibolite, with varieties of diorite, and rocks in \*'hich a triclinic feldspar prevails, together with schists more or less calcareous, and crystal- line limestones. The serpentines and their associated ophiolitic rocks, which constitute the lower members of the newer crystalline series, are described by Gastaldi as resting in some cases in nearly horizon lal stratification upon the ancient gneisses, and, elsewhere, ae overlying the limestones of this older series, from which their uncon- formable superposition may be inferred. § 52. The group of newer crystalline rocks, as given in Gastaldi's section of 1878, includes also what he desig- nates as recent gneiss and granite, besides various unde- scribod schists, with crystalline limestones, followed by a second horizon of serpentines, to which succeed gypsum and dolomites. All of these, as is shown in the section from Turin to the frontier, are intercalated with quai tzite, in a vast series of schists which a^e placed above the recent gneiss and graniie. Finally, the whole series is overlaid by the uncrystallino sediments cf the anthracitic group, of carboniferous age. § 53. The lithological characters of the lower part of this vast series of newer crystalline schists are sufficiently well defined in the various sections already noticed. As regards those which immediately succeed the serpentinic, chloritic, and talc-schist zone. — the group of "mica-schists" of Neri, the "recent gneiss and granite" of Gerlach and Gastaldi, we get additional light from various passages in the writings of the latter. They are spoken of in one place as gneissic mica-schists, more or less rich in horn- blende, in which, at Traversella, are also included serpen- tines. Elsewhere, the rocks of the same area are succes- sively called mica-schist, recent gneiss and mica-schists, gneissic mica-schists, and also, a very micaceoas gneiss, often passing into mica-pchist and sometimes hornblendic. With these, or with the lower portions of the series, are ti I 1^^ I i 462 THE GEOLOGICAL HISTORY OF SERPENTINES. l^- associated granitic and syenitic rocks which, in the opinion of Gastaldi, are not eiaptive, but the result of local modi- fications of the surrounding gneiss.. From my own obser- vations, I conclude that, while these recent gneisses in the Alps, as in North America, assume a highly granitic aspect in certain beds, they are not to be confounded with veritable intrusive rocks which penetrate them. ' § 54. Gastaldi has described in detail and figured a section in the Biellese, a region carefully mapped by Qiiintino Sella and G. Berutti, and studied both by Ger- lach and Gastaldi.* Here, in the section as given by the latter, the granite or granitic gneiss is bounded to the northwest by serpentine, diallagic rooks, "and ui/her greenstones," followed by a band of diorite. To this suc- ceeds a great breadth of the n< wer gneisses, in which is included a large dike of melaphyre, evidently of eruptive and posterior origin, and, farther to the westward, a mass of syenite, which is extensively quarried, and has been studied with great care and described by Cossa in his work already mentioned. I had the good fortune to visit this well known region in 1881, in company with Signor Quintino Sella. The granitic rock of the eastern part of the section appeared to be a part of the ancient gne 3sic series so largely developed elsewhere near Blella, and con- sisting of reddish granitoid gneisses, sometimes horn- blendic, but scarcely micaceous, often thinly banded, highly contorted, and indistinguishable from much of the gneiss of the Laurentides, or of the South Mountain in Pennsylvania, east of Schuylkill. Interstratified with it, near Bie'la, are beds of coarsely crystalline impure lime- stone, holding graphite, mica, and hornblende, and resem- bling closely some Laurentian limestones. Elsewhere in the Alps, it may be noted, similar gneisses include serpen- tinic limestoPujj, as for example the pale green ophicalcite found by Favre in the gneiss of Mattenbach, near Lauter- brunnen, which is indistinguishable from that of the Lau- * Gastaldi, Studii, etc., part I., pp. 3 and 26. X.] EOCKS OF THE ALPS. 463 re' tian of Canada, and like it contains Eozoon Canadense.* It is well known that similar sevpentinic aggregates are often found with the limestones in the ancient gneisses of Scandinavia and Finland, as well as in North America. § 55. This ancient gneissic series in the Biellese is directly overlaid by the ophiolitic and dioriti: belt (pietre verdi), and this is followed to the west by the newer gneisses and mica-schists, which cannot be distinguished from those found in the vicinity ci Philadelphia, or in the White Mountains of New Hampshire, which I have called Montalba.i. The intruded mass of syenite, made up of reddish orthoclase with some albite, hornblende, and a little sphene, presents, in the extensive quarries which I visited, the massive character and the comparative homo- geneousness which belong to a plutonic rock. The usu- ally great breadth of ophiolitic rocks met with in this part of the Alps is here, as pointed out to me by Signor Sella, rapidly reduced, to the southward, by the encroachment of the newer gn isses on the westward side, and, where the crysti'Uine rocks sink beneath the alluvial plain, does not exceed a kilometre. These relations suggest a trans- verse superpop'tion of the newer gneiss series alike upon the ophiolitic group and the older gneiss, of which we shall find evidence elsewhere. § 56. It has been seen that the designation oi pietre verdi was by Neri restricted to the ophiolitic group beneath the newer gneisses, which he referred to a later and distinct geological period. Gastaldi, on the other hand, extended the term so as to include not only the newer gneisses and mica-schists, but the vast mass of crys- talline strata between these and the anthracitic series, with their included gypsums and dolomites. The grounds of this extension are these : Serpentines are not confined to the lower ophiolitic zone, but also occur alike among the newer gneisses and the succeeding crystalline schists. * Favre, Recherches g^ologiques dans la Savoie, etc., iii., 320, and also Chem. and Geol. Essays, p. 342. 464 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. It is, says Gastaldi, " in contact with the gypsums and dolomites that we find the last limit of the serpentinous rocks which, for us, characterize the zone of the pietre verdi." This was in 1872, in his letter to De Mortillet, at which time Gastaldi was disposed to place in a separate group the crystalline schists above the horizon of the upper serpentines. He, however, subsequently included the whole of these schists in the zone of the pietre verdi. § 57. As will be made apparent, the schists for a great distance below this horizon are not to be separated from those above. We have in them, in fact, a third great crys- talline group, overlying the younger gneisses, but by Gas- taldi included with these and the lower ophiolitic group under the common name of the pietre-verdi zone. At other times Gastaldi used the term of pietre verdi in the more restricted sense in which it was employed by Neri. He speaks, in 1874, of "the pietre verdi properly so called," and in this sense he declares it to be comprised between " the ancient porphyroid and fundamental gneiss" and "the recent gneiss, which latter is finer grained and more quartzose than the older." He says farther, " I will not assert that when specimens of this newer gneiss are confusedly mixed with those of the more ancient, it would always be practicable to distinguish them petrographi- cally ; but I do not hesitate to affirm that, on the ground, the distinction is not difficult, on account of the frequent alternation of the younger gneiss with the other charac- teristic rocks of the upper series ; while the older gneiss, however wide its extent, is generally unmixed with other rocks." * § 58. The newest crystalline group, mentioned as over- lying the younger gneiss and mica-schist series, is that of the argillo-talcose schists of Favre, the gray lustrous schists of Lory (jglamschiefer)^ with their included serpen- tine, gypsum, karstenite, dolomite, micaceous limestone, banded and statuary marbles, and quartzites ; a group very * Studii, part II., p. 31. [X. s and tiuou3 pietre rtillet, sparate of tlie .eluded ) verdi. a great 3d from at erys- by Gas- 3 group ne. At i in the by Neri. )erly so )mprised L gneiss" ned and " I will neiss are would rographi- ground, requent charac- r gneiss, th other as over- is that of lustrous id serpeu- ■imestone, roup very X.] EOCKS OF THE ALPS. 465 conspicuous in Alpine geology. These rocks are well seen in the section from Turin to the French frontier, and are traversed in the Mont Cenis tunnel. (See also §§ 62-66.) § 59. The vast thickness assigned by various observers to this eutire series of newer crystalline schists, counting from the ancient gneiss below, is a remarkable fact in their history. We have seen the great breadth ascribed to the successive zones or groups in the sections already noticed. Gastaldi, in 1876, estimated the real thickness of the pietre-verdi zone, including the upper lustrous schists, at 24,000 metres, of which 8000 metres, or one- third, was assigned to the lower ophiolitic group, or proper pietre verdi, apparently without including the younger gneisses and mica-schists, which make up the middle group. To the upper group, as seen in the Mont Cenis tunnel, Sismondi and filie de Beaumont assigned a vertical thickness of not less than 7000 metres, and Rene- vier finds for it elsewhere an apparent thickness of 6000 metres. § 60. We ha^'e hitherto spoken of the Western Alps, and the sections as yet noticed do not extend to the east- ward of Lago Maggioic. The map by Von Hauer, of the Lombard and Venetian Alps, published in 1866-68, em- braces the region from this meridian eastward, and shows the same order of succession as that laid down by Gerlach in the west.* The various groups, as indicated by Von Hauer, are as follows : 1. The ancient gneiss and granitic rocks, designated by him as the " Central gneiss " ; 2. Greenish schistose rocks, described as hornblendic, dioritic, and euphotidic, with serpentines, chloritic and talc-schists ; 3. Saccharoidal limestones, more or less mica- ceous, with talc-schists ; 4. Serpentines, euphotide, diorite, and talcose and chloritic schists, as before ; 5. A fine- * Gastaldi, Studii, part I., p. 18, and Fr. Von Hauer, Geologische tJbersiclitskartfc der Osterreichlsch-Ungarischen Monarchic, fol. v., West- Alpen, u. fol. vi., Ost-Alpen; Wien, 1866-68. 1" , 46G THE GEOLOGICAL HISTORY OP SERPENTINES. [X. grained gneiss, designated as " Recent gneiss " ; and 6. Mica-schist, with hornblendic and feldspathic varieties. We have evidently here the same great pietre-verdi zone as in the west, comprised between the older gneiss and the younger gneiss with its attendant mica-schists. There appears, however, a considerable development of crystal- line limestones in the midst of the pietre verdi. § 61. Further light is thrown upon the question of these crystalline rocks of the Alps by the observations of Renevier, Heim, and Lory, especially as embodied in an essay by the latter on tlie Western Alps, published in 1878, and in a study of the geology of the Simplon, by Renevier, in the same year.* According to Lory, the ancient crystalline rocks, designated by him as the " Primi- tive schists," as seen in the Simplon, and elsewhere in this region, include three groups, in ascending order: 1. The stage of the Gneiss, properly so called, including varieties from the highly feldspathic and massive grani- toid gneisses to others less feldspathic and more distinctly laminated. 2. The stage of the Mica-schists, often gar- netiferous, which embraces, however, alternating beds of gneiss, the two rocks passing insensibly the one into the other. These mica-schists, tender, and gray in color, are often more or less impregnated with carbonate of lime, and contain bands of limestone and marble. 3. The st?ge of the Talc-schists, a term which, as Lory explains, he uses in a very general sense, to include not only stea- tites, but talcose, chloritic, and hornblendic schists, the latter sometimes without visible feldspar, but often more or less feldspathic, and thus passing into varieties desig- nated by him as talcose, chloritic, or hornblendic gneiss. The so-called protogine of the Alps, according to Lory, is but a granitoid variety of talcose or chloritic gneiss, sub- * Essai sur I'orographie des Alpes occidentales, par Charles Lory, p. 76 ; Paris and Orenoble, 1878. Also, Structure geologique du massif du Simplon, etc., par E. Renevier, Bull, de la see. vaudoise des sciences naturelles, vol. xv., No. 79. p> X.] BOOKS OP THE ALPS. 407 ind 6. •ieties. i zone S8 and There jrystal- tion of Aons of 1 in an shed in plon, by ory, the " Primi- vhere in g order: including xe grani- distinctly )ften gar- beds of one into in color, honate of . 3. The explains, only stea- jhists, the iften more [ties desig- lUc gneiss. |to Lory, is ;nei8S, sub- bhavles Lory, lue du massif des sciences ordinate to the talc-schist stage, and passing insensibly into the talcose and chloritic schists, with which it alter- nates. It is not, therefore, as some have supposed, tlio fundamental rock of the Alps, but belongs to an upper portion of the Priznitive schists. The lower gneiss of the Simplon, described by Gerlach as the gneiss of Antigorio, to which this distinction apparently belongs, is further noticed b\ Kenevier, who assigns to it a great thickness, and regards it as the basal rock of the Alps, corresponding to the ancient gneiss of Gastaldi and the central gneiss of Von Hauer. The succeeding mica-schists, often garneti- ferous and calcareous, with alternating gneiss and lime- stone bands, have also a great volume. The hornblendic schists play a less important part in the series. Though these sometimes contain a little mica, or a little chlorite, chloritic schists are rare, and the stage of the talc-schists, indicated by Lory, is not mentioned by Renevier in his description of the Simplon. § 62. The term of Primitive schists, as employed by Lory and by Renevier, is not extended to the gray lus- trous schists, already noticed as forming the upper part of the great series included by Gastaldi in the pietre verdi. These upper schists are by Lory regarded as altered trias, a view in which Renevier acquiesces. They are, for the most part, soft, glistening, and talcose in aspect, and have been variously described as argillo-micaceous and argillo- talcose schists, being sometimes, according to Lory, true sericite-schists.* They closely resemble the crystalline schists with hydrous micas which abound in the Primal and Auroral divisions of Rogers (Taconian), as seen in eastern Pennsylvania. These schists in the Alps are traversed by veins of calcite and of quartz, and include, besides great beds of quartz-rock (often a detrital sand- stone), beds of limestone, sometimes micaceous, of bana d and of white granular marbles, of dolomite and of gyp- sum. This latter, in the subterranean exposures made in * Bull. soc. gdol. de France, z., 29. / jjft; 1. '.^if*K.. 4J8 THE GEOLOGICAL HISTOUY OF SERPENTINES. [X. the Mont Cenis tunnel, is represented by anhydrous sul- phate of lime (karstenite), and is accompanied by rock- salt and sulphur. Magnesian silicates are also found in this group ; nodules of talc are imbedded in the karsten- ite, chlorite occurs in veins and layers, and beds of ser- pentine (and of euphotide, according to Gastaldi) are interstratified with these shining argillo-talcose schists. § 63. The resemblances in mineral character between these upper argillo-talcose schists with chlori*^<3 and with interstratified serpentines, and the lower or true pietre- verdi zone, are obvious. Lory has moreover remarked the likeness between these upper schists with limestones and the mica-schists with limestones in the horizon of the newer gneiss series (included by him in the Primitive schists) as leading to the confounding of the two. This resemblance, he suggests, "may have thrown some ob- scurity " upon the relations of these various rocks, and the structure of the region. It will not have escaped the notice of our readers that in the description of the section of the Simplon there is no recognition whatever of the great mass of serpentines, euphotides, and other opliiolitic rocks belonging to the pietre verdi, which elsewhere are found at the base of the newer crystalline schists, occupy- ing a horizon between the older and the younger gneisses. § 64. It will also be noted that Lory places a horizon of talc-schists, with chloritic rocks, etc., at the summit of the newer gneisses, and the view naturally suggests itself that Lory has himself confounded the lustrous schists of the upper series and their magnesian rocks, with the great lower ophiolitic zone. This latter would appear to be wanting in the section of the Simplon, where it is not noticed by Renevier. Lory thus places above the younger gneisses a talc-schist series to which he refers many of the types of rocks met with in the great ophiolitic and talc- schist zone, which elsewhere underlies these younger gneisses, and in which the protogines are probably in- cluded. In this way the apparent discrepancy between us sul- y rock- und in :ai'sten- of ser- di) are lists, between nd with 3 pietre- rlced the )nes and 1 of the Primitive ro. This some ob- ocks, and caped the ^le section er of the oplnolitio where are ts, occupy- r gneisses, a horizon summit of rests itself schists of the great lear to be it is not le younger lany of the and talc- le younger [robably in- >y between X.) EOCKS OF THE ALPS, 469 Lory and all the observers hitherto mentioned is ex- plained, as proposed by the present writer in 1881. The rehitions observed in the Biellese, as already noticed, sug- gest tliat the younger gneisses were deposited unconform- ably, alike upon the older gneisses and the great ophiolitic group, as is the case in numy otiier regions. § 65. In like manner, according to Lory, the lustrous schisis themselves, with included serpentines (which he regards as contemporaneous eruptions), rest directly up- on the ancient gneisses in the Levanna, between Susa and Lanzo. Other evidences of a want of conformity between these various groups of ancient schists in the Alps are not wanting. At the Col de Mont Gendvre, as described by Lory, there appears through the lustrous schists a great "mass of non-stratlHed rocks, comprising euphotides, ser- pentines, variolites, and various rocks of passage between these types." These ophiolitic rocks, which correspond to the lower part of the pietre-verdi zone of Gastaldi, are regarded by Lory as eruptive, and have not been recog- nized in his scheme of the divisions of the primitive schists. Their appearance among the lustrous schists is thus, according to him, an irruption in the midst of the trias, instead of being, as we should rather regard it, a protrusion of a portion of the pietre-verdi zone in the midst of the lustrous schists, which are here unconform- ably superimposed upon it, as elsewhere upon the ancient gneisses. § 66. The history of the upper or argillo-talcose schists of the section under consideration will be found discussed at some length by the present writer in a review of Favre on the geology of the Alps in 1872. It was there shown that these, though very distinct from and unlike the underlying micaceous, hornblendic schists and gneiss, are really crystalline schists, and very unlike the normal trias of the region, to the horizon of which they had been referred by most geologists. The section of them afforded by the Mont Cenis tunnel was then and there discussed, MM ',•,1^.(1 I 'I: 'B tf .r ■■^-. i .^ . , ■ m0 U I 470 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. niid many reasons were given for rejecting the notion of their triassic age, and for assigning them to the eozoio period. As was shown in u subsequent note to that re- view, Favre, after publishing his book in 1867, was led to adopt the view advanced by Gastaldi in 1871, that these schists were pre-carboniferous, though probably paleozoic, a conclusion which the latter subsequently exchanged for that of their eozoic age, as maintained by the present writer since 1872.* § 67. The section traversed by the St. Gothard tunnel furnishes important details for Alpine geology. This work, beginning at Goschenen, on the north, ends at Airolo on the south side of the mountain, the entire dis- tance being 14,920 metres. The first 2000 metres from the northern portal are in the massive rock of the Fin- steraarhorn, called by various observers granite or granitic gneiss, and by Stapff regarded as an older gneiss than that of the remaining part of the section. Between this and the mountain of St. Gothard is included the closely folded synclinal basin of Urseren, while the southern jjortal, at Airolo, is on the northern side of the similar basin of Ticino; the great intermediate mountain-mass of highly inclined and faulted strata, presenting a fan-shaped arrangement. The basin of Urseren holds, folded in gneiss and mica-schist, a group of strata consisting of ar- gillites, sometimes calcareous and often graphitic, with gray lustrous, unctuous sericite-schists, together with quartzose layers, and others which, from a development of feldspars, pass into an imperfect gneiss. With these are interstratified granular crystalline limestones, white or banded with gray, with dolomite and karstenite. Some of the limestones included in this synclinal have afforded indistinct organic forms, and the series has been referred, like the similar rocks noticed in previous sections, to the mesozoic period. A repetition of these is met with in the * Amer. Jour. Science (3), iii., pp. 1-15, also Chem. and Geol. Essays, pp. 333, 336, and 347. X.] ROCKS OF THE ALPS. 471 Ticino basin, on the south side of the mountain. Apart from these, the great mass of strata along the line of tlie tunnel consists of micaceous gneisses and mica-schists with hornblendic bands, the whole having the characters of the younger gneissic series, and very distinct from the older gneiss of the northern portion.* If this latter be the central gneiss, the pietre-verdi zone is here absent. I have not seen the gneiss of the Finsteraarhorn, but, having examined the gneisses and mica-schists of the St. Gothard and the Ticino, can affirm that they have the lithological characters of the Montalban series of North America and of the younger gneiss and mica-schists of Gastaldi and Von Hauer, in which they were included by the Austro-Hungarian geological survey. (§ 60.) The serpentines of the younger gneiss, as seen in the St. Goth- ard section, will be described in part vii. § 68. With regard to the presence of granites in these regions, Cordier, as cited by Lory, long ago asserted that true granites, occurring in veins or transversal inclusions, arc rare in the Western Alps. He, however, excepted some masses, of which the granites of Baveno may be taken as a type, and others, which are rather veins of segregation (endogenous) than of injection.f For the rest, Cordier regarded the granites of the Alps as strati- fied rocks. Gastaldi, going still farther in his protest against plutonism, admits, in the regions examined by him, none but stratified rocks of aqueous origin, and has included in his sections masses that I regard as iffneous and intrusive rocks, but which are by him confounded with true indigenous gneissic rocks under the title of " recent gneiss and granite." As regards the porphyry mentioned in the sections of Neri as above the recent gneisses, and that placed by Gastaldi above the lustrous schists, it would appear that * For full details of this section see Profll g^ologique du St. Gothard, etc., par Dr. F. M. Stapff; Berne, 1881. t See the author, in Chem. and Geol. Essays, p. 331. '1 i 111 ! 472 THE GEOLOOICAL HISTOUY OF SERPENTINES. IX. the latter employed this term in a very vague sense, since he speaks of the foldspathic and quartziferous porphyries of this region as presenting great varieties in structure and in composition, and as passing into other rocks, notably into granites, from which it is often dillicult to separate them.* Ho seems, under the general term of i)orphyry, to have included both stratified rocks at different horizons, and intruded masses of various kinds. § G9. From the various descriptions and sections of the Alpine rocks, which we have here considered, it ajv pears that they may be included in four distinct groups, which are as follows in ascending order: — I. The central or ancient granitoid gneiss, with occa- sional quartzites and crystalline limestones, bearing graph- ite and many crystalline minerals. This group we refer, with Gastaldi, to the Laurentian. II. The great group of the pietre verdi proper, in which, besides serpentines and ophiolitic rocks, are in- cluded bands of limestone, and also apparently certain gneissoid rocks, the protogine or the talcose gneiss of Lory and Taramelli. (§§ 61, 78.) It is worthy of remark that although Gastaldi, like Neri, Gerlach, and Von Hauer, placed the great group of recent gneiss and mica-schists above the true pietre-verdi zone, Avhich he declared to be confined between the older and the newer gneiss, he, in his last published sketch, indicated, besides this, another horizon of "recent gneiss and granite" (not elsewhere noticed by him) intercalated in the pietre-verdi zone, as thus limited, and probably corresponding to these talcose gneisses. This second or pietre-verdi group, we refer with Gastaldi to the Huronian. III. The younger gneiss and mica-schist series," with hornblendic varieties and intercalated crystalline lime- stones, and in some cases with serpentines and eupho- tides. This group, upon the lithological characters of ♦ Studii, etc., part H., p. 34. ^ I ' i x.i ROCKS OF THE ALPS. 478 wliich we have already insisted (§§ 58, 55, 67), wo regard as the ie[)r('soiitutivo of the Montalbaii. IV. The upper lustrous schists, with gypsums and kar- stcuite and tale, with interstratiliod srvjeutines, (^uartz- ites, often sandstones, argillites, dolomites, micaceous limestones, and bandea and statuary marbles. This group, as wo have already indicated, presents many re- scmbhmces with the great Lower Taconic or Taconian series of North America. In it are included by Gastaldi, the crystalline limestones of the Apuan Alps, which yield the statuary marbles of Carrara and of Massa. § 70. In the Western Alps there is, so far as is known, no evidence of the lower paleozoic rocks, — the sandstones with anthracite, which succeed the crystalline schists, containing in many places a carboniferous flora. The ,3ame, according to Gastaldi, is true in the Maritime Alps and the Apennines, where, in many cases, he finds the crystalline schists overlaid by the anthracitic series. Thus, in the valley of Macra, above the seipentines are found calcareous schists with crystalline limestones and quartz- ites, which are successively overlaid by the carboniferous sandstones, the limestones of the trias, with their charac- teristic fauna, the lias, the cretaceous and the nummulitic beds. At Torre Mondovi, the serpentines are overlaid by fossiliferous triassic limestones, while in the valley of Bormida they are directly succeeded by the marls, sand- stones, and conglomerates of the lower niiocene, and in the valleys of Staffora and Polcevera by the alberese and the macigno of the eocene. The supposed pre-curbonifer- ous fauna found by Michelotti in the limestones of Chaberton, has, on farther examination by Prof. Mene- ghini, been shown to be of triassic age.* § 7L Passing now from the mainland of Italy, we come fo Corsica and Elba. The serpentines of the former island have long been known to geologists, and have with- in the last few years been especially studied by liollande, • Gastaldi's letter to Zezl, In 1878, already cited. 1: I'' mm I ■;, h\ ')i .■«■ tiA.-'»:tm-x»L«fJ^. -.H >aB .avtf ' rifniiififiifr "'li'^miffliiilJ 474 THE GEOLOGICAL HISTORY OP SERPENTINES. CS. Coquaiid, Dieuiefait, and Lotti. Coquand, who described the serpentines of Corsica in 1879, and who, lilce Hollaude, regards them as eruptive, supposes them to be in part very ancient, and in pai't tertiary, since, according to liim, some of them overlie the nummulitic beds.* Fellati, whose essay we have already cited, refers however the whob of the serpentines of this island to a pre-paleozoic period, and Dieuiefait, who described these rocks in 1880,t de- clares that Coquand's reference of the serpentines found near Corte to the tertiary is based on an error of observa- tion. He moreover asserts that the serpentines of Corsica are stratified sedimentary rocks belonging to a single geological horizon, at which they may be traced continu- ously for a length of more than 200 kilometres from Corso along the northeast coast of the island. The geological succession, according to him, is as follows : — 1, stratified protogine ; 2, gneiss ; 3, lustrous schists ; 4, saccharoidal limestones ; 5, schists more or less talcose ; 6, schists en- closing sei'pentines of many varieties; 7, clay-slates; 8, black limestones with carbonaceous matter ; 9, beds often detrital ; 10, infra-liassic limestones, with Avicula contorta. § 72. Lotti, who has since studied these rocks,:^ con- firms fully the observations of Dieuiefait. He describes the crystalline limestones, white or banded, with grayish, greenish, or lead-colored talcose or silky schists ( holding a mica, sometimes apparently damourite or sericite), in which are found layers of serpentine. The serpentine itself is generally scaly in texture and glassy, but granu- lar vai'ieties are met with including veins of epidote, others with altered crystals of olivine, and also ophical- cites. The gneisses beneath the serpentine zone pass into quartzose mica-schists, often including almond-shaped masses or segregations of quartz and feldspar, sometimes * Coquand. Bull. Soc, G^ol. de France (3), vil. t Dieuiefait. Coniptes I'l-ndus do J'Acad, des Sciences, xci., p. 1000. X Lotti, Appunti Geologici sulla Corsica ; Boll, del Comitato Geolo- gico, anno 1883, Nos. 3-4. ' X.] EOCKS OF THE ALPS. 175 with large plates of mica. It would appear, from the descriptions of Lotti, that these serpentines of Corsica belong to the upper horizon defined by Gastaldi, above the recent gneisses, and in what we have designated as the fourth group of Alp'iue crystalune rocks. (§69.) The underlying proto^fine is, according to Lot+^^i, a talcose gneiss. § 73. The resemblance of these rocks to those asso- ciated with similar serpentines on the neighboring island of Elba is declared by Lotti to be very close. There also the serpentinic horizon is underlaid by gneisses and mica- schists, as in Corsica. He concludes witli Gastaldi that the great crystalline zone oi the Alps is connected through the Maritime and Ligurian Alps with the similar rocks of Corsica and Elba. Resting upon the ophioll^ic strata in Elba are found, according to Lotti, paleozoic carbonaceous slates containing Orthoceras, Cardiola, Actinoirinus^ and probably also graptolites. Lotti, however, while he asserts the great antiquity of all of the serpentines of C'jrsica, and part of those of Elba, maintains the existence in the latter island of other serpentines, which, like those of Tuscany, he refers to the eocene period. A similar question is raised with regard to the granites of the two islands. Thus Pare to, who regarded as ancient, or at any rate pre-triassic, the granites of Corsica, admit- ted for the granites of Elba, Monte Cristo, and Giglio a post-eocene age, a view which is also sustained by Lotti, while De Stefani, on the other hand, assigns the Elban granites to prr-triassic time. I can scarcely doubt that all of these granites, as well as tlie ophiolites both of the various islands and the mainland, will be found, as main- tained by Gastaldi, of pre-paleozoic age. § 74. If we turn to the island of Sardinia wt find a series of pre-Cambrian .crybtalline schists, said to consist, in their upper portions, of argillites, sometimes talcose, sandstones, crystalline limestones, and dolomites. These, which are referred by Bornemann to thf< Huronian or ii 470 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. I j i ;• n> ii, ^:!iii' pietre-verdi zone of the Alps, are overlaid, as was first shown by De la Marmora, by a series of uii crystalline limestones, shales, and sandstones, containing an abundant lower paleozoic fauna.* Of this, the upper Cambrian (Ordovician)t forms were long since described by Mene- ghini. The subsequent studies of Bornemann, in 1880, showed at the base of the series a zone marked by Para- doxides, Conocephalites, Archeoci/athus, etc., wliicli have also been examined by Meneghini, and establish the existence of a Lower Cambrian horizon. The writer had, in 1881, the pleasuire of examining at Bologna, in company with James Hall, a collection of these fossils. Above the Ordovician beds in Sardinia is found a great mass of limestone, of undetermined age, remarkable for its beds and included masses of lead, silver, and zinc-ores. § 75. We have now shown that these crystalline rocks, which, in parts of the mainland, are directly succeeded by tertiary sediments, are in different areas overlaid by various subdivisions of the mesozoic, and finally by car- boniferous, Ordovician, and Cambrian sediments, thus dis- proving the view of the older geologists, who assigned to these same crystalline recks a paleozoic or a mesozoic age. It is instructive to mark the steps by which this view has, in the process of investigation, been left behind. In Neri's section the older gneiss and the pietre verdi proper are called azoic or protozoic, but the recent gneiss is con- jectured to be paleozoic. Lory, however, included the latter in the primitive series, but claimed the lustrous schists as altered trias, while later, Gxstaldi, and with him Favre, placed even these in the paleozoic, until at last we find Gastaldi adopting the conclusion first put forward by the present writer in 1872, that the whole of these crystalline rocks are to be referred to pre-Can\brian time. * Bornemann, sur les formations stratifides anclennes tie la Sardalgne. Comptes Ilendus du Congros Gcol. Inter, de Bologne, pp. 221-232. t Scepo^t, Essay XL, § 17. X.] EOCKa OF THE ALPS. 477 § 76. The Gtory of the crystalline marbles of Carrara, now included in this series, is not less instructive. They were regarded as eruptive by Savi, who taught that dolo- mites and limestones had been poured out in a fused state, alike in secondary and in tertiary times,* and even indicated what he supposed to be centres of eruption. The marbles of Carrara, with their associated schists, have since been called cretaceous, liassic, rhaetic, infra-carboniferous, and pre-paleozoic.f They were in 1874, in the second part of Gastaldi's Studii, included in the rocks of the pietre-verdi zone, the term heiu^. then used in its larger sense, as em- bracing not only tL> urue pietre verdi, but the whole crys- talline series above the ancient gneiss. § 77. This was also clearly stated by Jervis, in his elab- orate work on the mineral resources of Italy,J a veritable treasury of information, most carefully and systematically arranged. In his first volume, in a tabular view of the geology of the Alps, he had already adopted the views of Gastaldi, and placed the whole of the crystalline stratified rocks above the ancient gneisses in a "pre-paleozoic group," which he regarded as synonymous with the pietre-verdi zone. In his second volume, in a similar tabular view of the geology of the Apennines and the adjacent islands, he further insists upon the same view, and puts above the ancient gneiss, in what he calls the pre-paleozoic period, the great series of " stratified azoic rocks," including not only the ophiolites, and the recent gneisses and other crystalline schists, but the saccharoidal and compact mar- bles of the Apuan Alps. (Loc. cit., p. 9.) It is to be remarked, as shown both bv Jervis and Gastaldi, that this great younger crystalline series is the metalliferous * Boue, Guide du g^ologue voyageur, II., 168. For the views of others as to the eruptive origin of crystalline limestones, see my Chemical and Geological Essays, p. 218, and oii^e, pp. 00, 01, and 228. t For a jiotice of some of the various views which have'lioon put for- ward with regard to the age of these marbles, see Lebour in the Geologi- cal Magazine for 1876, pp. 287 and .383. X 1 'i "esori Sotteranei del' Italia, 3 vols. Svo, Turin, 1878-1881. ill 478 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. r 1- tone of Italy, containing much cupriferous and niccolifer- 0U8 pyrites, in veins and interstratified beds, together with crystalline iron-ores, lead, zinc, and gold. § 78. With the general succession of the Alnine rocks already given, we may compare the observations of Tara- nielli in the Valtelline, where he describes the ophiolites as lying below a great gneissic and granitic series, from which they are separated by a garnetiferous hornblendic rock and saccharoidal limestones. The lowest division in the series, as observed in the Valtelline, is, according to Taramelli, a quartzose talc-schist, upon which reposes tu: serpentine in heavy continuous beds, having all the appearance of a stratified rock, followed by potstone, that is to say, steatite or chlorite. To this succeed, in ascend- ing order, hornblendic and epidotic rocks, associated with crystalline limestones, often talciferous ; then schistose amphibolite, talcce gneiss, talc-schists, and eclogite, and finally a coarsely crystalline glandular gneiss, itself over- laid by granitic and associated hornblendic rocks. This apparent reversal of the succession as defined by Gastaldi and others, suggests the probability that we may have in the Valtelline an overturn of the strata, such as is well known in many parts of the Alps and elsewhere, placing the more ancient rocks above the younger ones.* § 79. It is here the place to notice the mode of occur- rence of the serpentines which, in Saxony, are found interstratified in the granulite series of the Mittelgebirge. The granulite proper may be described as a fine-grained, gray, laminated binary gneiss, consisting essentially of orthoclase and quartz, but often containing garnet, and sometimes cyanite and andalusite. By an admixture of mica, it passes, through ordinary gneiss, into mica-scbists, which are abundant in the series. In it are also inter- stratified diohroite-gneiss, sometimes in great beds, and a greenish hornblendic gneiss, as well as the so-called gab- broe of the region (like that of Neurode). These occur * Boll. Soc. Geologica Italiana, I., p. 14. .' X.] EOCKS OF SAXONY. 479 in larger or smaller lenticular interstratified masses, to •wliiuh the distribution of the diallage in a granular labra- dorite base gives a well defined gneissoid structure. In this same series, tlie serpentine is found in interstratified beds, occasionally garnetiferous, and sometimes associated with Iherzolite. § 80. I have not seen these rocks on the ground, but have examined a large collection of them in Leipzig, with the assistance of my friend. Dr. Hermann Credner, and was struck with their close resemblances to the rocks with which I am familiar in the newer or Montalban series of gneisses and mica-schists throughout the Atlantic belt of North America. The muscovite-gneisses of the Erzgebirge, with their occr ^onal layers of limestone and of hornblende- rock, and their intercalated and overlying mica-schists, I also refer to the same general horizon. It is in these, it will be remembered, that are found the abundant con- glomerate beds described by Sauer, the pebbles in which consist chiefly of varieties of granitoid gneiss, resembling closely those cf the ancient gneiss of the Alps (Biellese) and the Laurentian gneiss of North America. These are, however, as I have seen, accompanied by pebbles of crys- talline limestone.* Mention should also be made, in this connection, of the existence of similar conglomerates in Sweden, at Soljoarne, where pebbles of ancient gneiss and granite are found, at several points, imbedded in fine-grained schistose gneiss, in calcareous mica-schist, and also in a red halleflinta, the strata of all of which are shown to rest unconformably upon the older granitoid gneiss.f § 81. It will be remembered by students in geology that in 1870 the present writer announced his conclusion that there exists in North America, besides the Lauren- tian gneisses, " a great series of crystalline schists, includ- * Zeitschrift f. d. ges. Naturwiss, Band liii. ; also, Geol. Mag., January, 1882, and Bull. Soc. G^ol. de Fr., x., 20; also, Amer. Jour. Science (3), xxvi., p. 197 ; and ante, p. 255. t Hummel. Cm Sveriges Lagrade Urberg, etc., Stockholm, 1875, p. 30. ^J 11 nw^HH || 1 m ' BbBMmBj. ' tf i II IS IP •' 1''.'.* Iheii ' ¥ W^^SSKK LI 1 h . ' 91 'Wmm F(F. i^l^M^K! pi ! , 1 1 p It \:^ n y. 1 B II -ffunimiMMniii THE GEOLOGICAL HlSTOllY OP SERPENTINES. [X. ing mica-schists, staurolite, and chiastolite-schists, with quartzose and hornblendic rocks, and some limestones, the whi le associated with great masses of fine-grained gneisses, the so-called granites of many parts of New England." * These rocks were especially indicated as oc- curring in the White Mountains of New Hampshire, but were also said to be found to the northwest of Lake Su- perior, as well as in Ontario and in Newfoundland, in which last two regions they were believed to rest uncon- formably ujjon the Laurentian gneiss. In both of these latter localities, there were provisionally associated with this group some higher limestones, with crystalline schists, and for the whole the name of Terranovan was suggested. § 82. In the following year, 1871, in an address before the American Association for the Advancement of Sci- ence, these rocks were farther noticed, under the name of the White Mountain series. The higher limestones and schists (which were not found on the geological section then described) were, however, excluded. This great series of younger gneisses and mica-schists was then as- signed to a horizon above the Huronian, and as a distinct- ive name was desirable for o series so conspicuous in American geology, that of Montalban (from the latinized name of the White Mountains) was proposed in the same year.f It was at the same time shown that the view held by most American geologists, that these rocks were altered paleozoic strata, was untenable, and that they were to be regarded as pre-paleozoic. At a later period, the higher limestones and schists, at first associated with these newer gneisses and mica-schists, were referred to the Lower Ta- conic of Emmons — the Taconian series.J When, in 1870 and 1871, 1 thus attempted to subdivide * Amer. Jour. Science (3), 1, 85. t Proc. Amer. Assoc. Adv. Science, 1871, p. 6; also Chdm. and Geo!. Essays, pp. 194, 244, 282; Das Ausland, Dec. 25, 1871, p. 1288, and Azoic Rocks, p. 181. t Azoic Rocks, pp. 201-211, 215. 0& }, with jstones, crrained °i New jtl as oc- lire, but :^ake Sii- Uand, in t iincon- of these ited with \e schists, mggested. ess before nt of Sei- le name of itones and cal section This great 'as then as- , a distinct- ipicuous in lie latinized Sn the same e view held ;ere altered were to be , the higher [these newer Lower Ta- bo subdivide jLita. and Geol. |l288, and Azoic X.] ROCKS OF THE ALPS. 481 the crystalline schists above the ancient gneiss of North America, and to define, above the liuronian, a younger series of gneibocs with niica-shists, I was not aware that Von Ilauer liad already been led by his studies to similar conclusions for the Eastern Alps, and had discovered above the great pietre-verdi zone, a series of gneisses with micaceous schists, as indicated in divisions 5 and 6 of his section (§ 60). Gastaldi, in 1871, and for years after, in- cluded these, with all the crystalline schists found above the ancient or central gneiss, in one great group of newer schists, which he assimilated to the Huronian. In review- ing this subject, in 1878,* I pointed out that the upper- most crystalline schists of the Western Alps should be separated from the Huronian, and compared them with the Taconian, while I noted the fact " that gneisses and mica-schists similar to those of the Montalban are found in many parts of the Alps." It was not, however, until after my studies among these rocks in 1881, that I referred the newer gneissic series of that region to the Montalban, for the two-fold reason that it occupies a similar strati- graphical horizon and is lithologically indistinguishable from it. § 83. Not less important in this connection is the suc- cession of crystalline rocks in Eastern Bavaria, which may be compared with those of Saxony. We have, in ascending order, according to GUmbel, first, the red or variegated gneiss, called by him Bojian, Avhich is followed immediately by the newer gray or Hercynii - gneiss, his second division, and by a third, the Hercyniun mica-schist series, occasionally hornblen'^'ic. To this succeeds, in the fourth place, the Hercynian primitive clay-slate series, which is immediately overlaid by Lower Cambrian fossil- iferous rocks. This primitive clay-slate series contains interstratified beds of limestone, sometimes doloniitic, at- taining in places a thickness of 350 feet, and associated with siderite, which gives rise by epigenesis to valuable * Azoic liocks, p. 245. I 'If ni^eaMSi S: t 482 THE GEOLOGICAL HISTOKY OF SERrENTlNES. [X. > 'Mi deposits of limonite along its outcrop. With these lime- stones are found varieties of hornblende and serpentine, accompanying which is the Eozodn Bavaricum of Giimbel. § 84. The Ilcrcynian gneiss is described by Giimbel as including nnicli gray quartzose and micaceous gneiss, with frequent beds of dichroite-gneiss, granulite, serpentine, hornblendic schists, and crystalline linipstones. With these are associated Eozoon Canadense-, from which Giim- bel supposed this upper gneissic series to represent the Laurentian, a view which was accepted by the present writer, when, in 1866, he translated and edited Giimbel's pai)er * for the Canadian Naturalist, and has since been ex- pressed by him elsewhere ; coupled with the suggestion that the Bojian might correspond to the Ottawa gneiss which underlies the Grenville series, the typical Lauren- tian (Lower Laurentian) of the Canadian survey. We are not, however, as yet prepared to recognize a subdivis- ion in the older gneisses of continental Europe, and meanwhile the analogies between the great Hercynian gneiss and mica-schist series combined, and the younger gneisses and mica-schists of Saxonj'' and of the Alps, lead us to refer what Giimbel has described as the newer gneiss series of Bodenmais and the Danube, to the same horizon as the younger gneisses of Gastaldi and Von Hauer, — the Montalban series, which in eastern Bavaria would seem, as in the Simplon, to rest directly upon the older gneiss, the Huronian being absent. The Hercynian clay-slate series, with its crystalline limestones, may correspond to the fourth group of the Alpine rocks, the argillo-talcose scliists, which we have compared with the American Taconian. V. — THE SERPENTINES OF ITALY. § 85. Returning to the Italian Alps, we have now to call attention to a very important conclusioi reached by * Giimbel, tjber der Vorkommen von Eozoon in dera Ostbayerischen Urgeblrge, Miinchen Akad. Sitzungsb., 1866 (l),pp. 25-70; also Can. Nat- uralist, iii., 1868, pp. 81-101. X.1 SERPENTINES OF ITALY. 483 \ Gastaldi, with regard to the geographical relations of the pietre-verdi zone ; using tho term in its larger sense, as embracing all the newer crystalline rocks, or those above the ancient gneiss. In 1871, in the first part of his memoir on the Western Alps, he declared it ;8 liis opinion that " all the serpentinic masses of the Tuscan and Ligu- rian Apennines, and the serpentines, ophicalcites, sacchar- oidal limestones and granites of Calabria, are but a pro- longation of this zone." In this were included, as we have already seen, tlie Apuan Alps, and, farther westward, a large part of the Maritime Alps. In support of these views he pointed out the mineralogical identity of the ophiolites and other crystalline rocks in the Alps and the Apennines. To the same horizon he also referred the so- called ophitic terrane of the Pyrennees. § 86. Gastaldi further called attention to the fact that ophiolitic rocks often appear in the form of isolated peaks or hills, for the reason that the accompanying crystalline schists and calcareous rocks, opposing less resistance, have been removed by decay and erosion, adducing many in- stances in support of this among the Alps. This being the case, he adds, we are not to be surprised when in the Apennines we find isolated masses of ophiolite rising out of the midst of surrounding Jurassic, cretaceous, or ter- tiary strata, which conceal the rocks that accompany the ophiolite. Thus it is, he adds, that "the notion has arisen in the Apennines that the serpentines, diorites, etc., are always eruptive rocks." They are, in his view, to make use of the happy expression of Roland Irving in describ- ing a similar occurrence, "protruding, but not extruded." These views were reiterated by Gastaldi in his letter to Zezi in 1876, when he asserted that the skeleton of the Apennines is a continuation of that of the Alps, and that the crystalline rocks of the Apennines are Alpine rocks. From the summit of Mont Blanc, he declared, they may be followed, more or less concealed by overlying strata of more recent date, to the Danube, to the plains of France, 484 THE GEOLOGICAL HISTORY OF SERPENTINES. Pf. 0'M f-,i' i!,'l to the Mediterranean, and along the peninsula which sepa- rates this sea from the Adriatic ; * assertions which he sup- ported in 1878 by many detailed observations, to be noticed farther on. § 87. These bold generalizations of Gastaldi have met with but partial acceptance in Italy, as may be seen by tlie discussions in 1881, and the publications of the 11. Comi- tato Geologic© and the Societti Geologica Italiana in 1881 and 1882, already referred to in § 43. Pellati, in his sum- mary, declares that the views of Gastaldi as to the anti- quity of the Alpine pietre verdi are confirmed by the work of Baretti and of Taramelli, the latter of whom clearly shows that the view, entertained by so many, that these rocks are carboniferous or triassic, is inadmissible. Hence, these ancient serpentines are by Pellati designated as pre- paleozoic (eozoic). This view he extends to all the ophiolitic masses situated in the Alps, to those of Calabria, and also to those of the Apennines west of the meridian of Genoa, those to the east of this meridian being in- cluded in the eocene. (§ 47.) § 88. Regarding the so-called eocene serpentine, and its associated rocks, Pellati observes, " as to its composi- tion, it differs but little from the older serpentine, the differences remarked being principally in a structure or- dinarily less schistose, and in a greater frequency of sub- ordinate ophiolitic rocks ; euphotides, eurites, diorites, va- riolites, ophicalcites, etc., more or less decomposed. The masses of proper serpentine are ordinarily more scattered, and of smaller dimensions, having almost always gabbros and beds of phthanite and jasper around them." Cossa, it is true, has remarked in the specimens examined by him that the mineral species bastite is more common in the eocene, or Apenniiie, than in the eozoic, or Alpine, ser- pentines, but, with this possible exception, the mineralogi- cal and lithological associations of the two are apparently identical. In fact, Pellati admits that it is in some cases * See the author, on Azoic Eocks, p. 245. I. Pf- ell sepa- he sup- ^ to be lave met )u by tbe R. Coini- a in 1881 1 liis sum- tbe auti- tbe work m clearly tliat these 5. Hence, ted as pre- to all the ,f Calabria, le meridian L being in- X.] SERPENTINES OF ITALY. 485 difficult, if not impossible, to distinguish between them. Within the great basin lying to the east of the meridian of Genoa, and embracing, as we have seen, the so-called tertiary serpentines, we are informed by him that ''the paleozoic and mesozoic rocks are generally very thin, and often are entirely absent, in which case the floor of jiietre verdi, or greenstones, is directly overlaid by the tertiary, and in fact by the very eocene which includes the younger serpentines. This is the case in the vicinity of Genoa, upon thfc right bank of the Polcevera, where the green- stones come in direct contact with the shales and the limestones of the npi)er eocene, and here it becomes doubtful whether, along this line of outcrop, por.ions of tertiary ophiolites are not mixed and confounded with others of the pre-paleozoic period." These supposed ter- tiary ophiolites " have a very great resemblance to those of the eocene of eastern Liguria, and present, moreover, a large development of the rocks which Issel has desig- nated as amphimorphic (§ 92). Thus, near Pietra Lavez- zara, for example, ophicalcites are exploited which are precisely like the green marbles of Levanto. In this same locality, moreover, argillites having the aspect of those (jf the eocene appear to dip beneath the ophiolites." In support of the belief that these seemingly tertiary ophio- lites are really eozoic, however, W3 are told that their outcrops present lines of continuity, connecting these ser- pentines with those of which the eozoic origin is un- doubted. We have seen (§ 41) that Professor Bonney, in his studies of the serpentines of Italy, fails to remark iiny distinction between the serpentines thus separated by the Italian geologists, since he describes as similar, both in mineralogical characters and in geognostical relations, the ophiolites lying to the west and those lying to the east of tlie meridian of Genoa. I shall, farther on, have occasion to refer to my own observations of some of these localities. § 89. The older school of Italian geologists, as already noticed, supposed the serpentines to have been erupted, 1. ''"PI i m Ml m 48(3 THE GEOLOGICAL HISTORY OF SEllPENTINES. [X. |l";<« r- ,.,.' :i liUe biisalts, at difToroiit ^eolo^iciil periods, and applied tliis view, not only to thoso wiiicli are evidently included among eozoic rocks, but also to thoso which rise among the teitiiiry deposits. Tiie study of tiie ophiolitic masses of eastern Liguria and of Tuscany, induced the earlier geolo- gists, like Savi, Pilla, and I'areto, to refer them to various ages between the cretaceous and the pliocene, but more recent observers have been led to include all of these ophiolites in the upjjcr eocene. This view was first advanced for those of the mainland of Tuscany by De Stefani, in 1878, and has since been maintained by Lotti, Taramelli, Issel, Muzzuoli, and (!!apacci, among others. The horizon in the ui)per eocene to which these observers refer the serpentines in question, consists of argillaceous and marly shales, alternating with beds of limestone and sandstone, and is below the argillaceous limestones with fucoids and nemertilites, but above the sandstone known as macigno, which is found at the base of the eocene in Liguria. § 90. As regards the origin of serpentines, Peliati re- marks that the recent studies of Italian geologists have led to hypotheses which ditfer widely from those formerly received, according to which serpentines were regarded as plutonic or eruptive, having come to the surface after the manner of volcanic lavas, or, at least, like certain massive trachytes, in a pasty state, or one of igneous semi- fluidity. Gastaldi, he adds, "from his studies of the an- cient serpentines of the Alps, regarded them, however, as sedimentary rocks, modified by subsequent hydrothermal actions operating at great depths in the earth." He compared their formation to that of the accompanying gneisses, mica-schists, chlorite-schists, crystalline lime- stones, diorites, and even the granites, syenites, and por- phyries of the Alps, to all of which he ascribed an aqueous origin. § 91. This hypothesis has not, however, been favorably received as an explanation of the origin of the so-called X*) SEUPENT1NE8 OF ITALY. 487 toi'tiiiry opliiolitoH of tlio Tuscan iuul Li•■•: i't'^'P.^". 1'1'f ms, however, kilicious or jas- fjaspers of Tus- L 11. Accad. del I the same year, Piorlzons in the 1 in thin layers i-t of crystalline fmangancse, and pies, leading the , deposits. rest directly upon the ophiolite, or with the intervention only of a layer of comminuted serpentine described by Capacci as an ophiolitic sand (arenaria ojiolUica). These overlying strata have a general inclination from the ser- pentine, that is to say to the eastward, of from 20° to 70°, but in some sections, as to the east and northeast of Figline, are represented in Capacci's sections as nearly horizontal, with small undulations exposing in valleys the phthanite, and even the serpentine, beneath the alberese. § 98. On the western side of the hill, where, as already said, these eocene strata appear nearly up to the summit, and plunge beneath the ophiolite, their inclination, as seen along the southwest border, is from 54° to 64° to the northeast. Here is observed a significant fact, which is- shown in the sections of Capacci, — namely, that the pre- viously noted relations of tlie serpentine, phthanite, and alberese are reversed. While on the eastern slope these three rocks appear in the ascending order just named, we find on the opposite flank of the hill, in the ascending section, alberese, phthanite, and serpentine; — the serpen- tine overlying the phthanite, and the latter the alberese. The natural and obvious interpretation of these facts Is that we have here simply an inversion of the natural order, resulting from an overturn of the strata on the western side of the hill. § 99. The ophiolitic mass itself is not simple, but, as described and figured by Capacci, is essentially composed of two layers of serpentine, with an intercalated lens of euphotide. Besides this rock, which has been tlie object of repeated studies, the last by Cossa, this lithologist has described associated masses of diabase, while Capacci has observed others of dioritic rocks, including a green variety distinguished as gabbro-verde, sometimes becoming vario- litiv., as well as the so-called gabbro-rosso, which, as there seen, is an iron-stained, somewhat calcareous dioritic rock, concretionary in structure, and apparently in a decom- posing state (§ 40). 492 THE GEOLOGICAL HISTORY OF SERPENTINES. PC. § 100. Capacci's view of the relations of these various rocks to one another, and to the accompanying eocene strata, is in accordance with the hypothesis ah-eady set forth (§ 93). He legards the ophiolite of Monteferrato as a great lentieuhir or almond-shaped mass (un' amigdala ofiolitica)^ " intercalated, in perfect concordance of stratifi- cation, among the strata of alberese and galestro of the eocene formation," which have been subsequently tilted, so as to give to the whole series an c stward inclination. In accordance with this conception, he supposes that, at a certain time during the accumulation of the eocene strata, there came, from a rupture in the earth's crust, a sudden effusion of an aqueous magnesian magna, which was spread out beneath the sea, and was subsequently overlaid by a continuation of the eocene beds, as before. The silicious sediment constituting the phthanite which, on the west side, is seen to underlie the ophiolite is, in this view, a portion of previously deposited and altered shale, while the phthanite on the east side is another portion of a similar sediment, subsequently laid down upon the ophiolite. § 101. The ophiolitic mass is thus, like all the other serpentines of Tuscany, of eocene age. The various rocks which enter into its constitution appear in the form of "lenticular masses or almond-shaped concentrations," of which the euphotide and the gabbros are examples. The gabbro-rosso is found in masses at the contact of the ophiolite with the phthanite, and results from the altera- tion of a diabasic rock by the action of thermal waters. These have also changed the galestro into the phthanite, found both above and below the ophiolites, and in perfect confo'-mity with the adjacent eocene strata, " which have all their distinctive characters, and present no traces of alteration or of metamorphism," "the action which pro- duced the phthanites being local, particular, and variable." Recomposed rocks, made up of grains and fragments of serpentine, in a cement generally calcareous, are found on SJ SERPENTINES OF ITALY. 493 the confines of the serpentine and at its contact with tlie phthanite. This is especially seen at Poggio, on the southeast side of the hill, wliere I found a veritable con- glomerate of fragments of serpentine imbedded in a paste of silicious slate. These facts, as well remarked by Capacci, show that previous to the deposition of these eocene beds, the serpentine-mass along the shores v.2 a shallow sea was subjected to a process of disintegration ; "and that, moreover, the formation of a serpentine corresponds to a kind of pause in the deposition of the eocene strata." The ophicalcites, in like manner, are found at the limits of the serpentine, and are breccias or conglomerates with a calcareous cement. § 102. I have thus given, in great measure in language translated from Capacci's memoir, the principal facts observed at Monteferrato, which I have, for the most part, verified. They, however, appear to me incon- sistent with the hypothesis propounded by the modern school of Italian geologists, and with the eocene age of the ophiolitic mass in question. The effusion of a great mass of aqueous material from the earth's interior into the eocene sea, its subsequent arrangement and crystallization into mountain-masses of euphotide, diorite, and serpentine, the elevation of these, and their subsequent disintegration to form the ophiolitic sands and conglomerates already described, riiark a geological period, and a revolution which ought to have left some traces in the surrounding eocene deposits. These, however, we are to believe, in accordance with the proposed hypothesis, continued after this event to be laid down precisely as before, — the alberese and the galestro previous and subsequent to the ophiolite making with this one conformable series. This process, moreover, we are told, was here confined to an area whose greatest extent was less than three kilometres, and was repeated at a great number of localities in the Italian tertiary basin, in all cases giving rise, not, as in ordinary eruptions, to a single kind of rock, but to a 494 THE GEOLOGICAL HISTORY OP SERPENTINES. l^. group of different rocks, indistinguishable in character from those which are known to be found in contiguous regions interstratified in crystalline schists of eozoic age. § 103. The only explanation which seems to me admis- sible, and one which is in complete harmony with the facts, is that this area of serpentine, with its associated euphotides, etc., was an eroded and uncovered mass in the midst of the eocene sea; that around its base was deposited the disintegrated material which i'^orms the ophiolitic sands and conglomerates, followed by the silicious sediments which make up the phthanite, and by the limestones and shales of the middle eocene. The s\i^- sequent movements of the eartli's crust, which caused the folding of these strata together with the intruding mass of eozoic rock upon and around which they were depos- ited, has resulted in the production of an overturned synclinal on the western side of the hill. § 104. As I have elsewhere insisted,* in cases like the present, where newer strata are found in unconformable superposition to older ones, the effect of lateral move- ments of compression involving the two series, is fre- quently to cause the newer and more yielding strata, along their border, to dip towards or beneath the older rock. These overlying strata, where they abut against their marginal limit, which was the ancient shore-line, will, in the conditions supposed, assume, according to local circumstances, either an anticlinal or a synclinal form. In the former case, the inclination of the strata towards the older mass, which forms a resisting barrier, follows necessarily, even though the elevation of the arch be slirlit. In the case of a synclinal fold or inverted arch, we have the strata dipping away from the older rock at a greater or less angle, as seen at the eastern base of Monteferrato, — the strata appearing in their natural order of superposition. When, as is frcviuently the case, this inclination passes beyond the vertical, giving I'ise to * Geological Magazine (January, 1882), ix., 39. X.] SERPENTINES OF ITALY. 495 an overturned synclinal, the same strata will appear to pass in reversed order beneath the overhanging mass of older rock, as along the svestern border of Monteferrato. It is hardly necessary to recall the fact that sharp or inverted folds, whether synclinal or anticlinal, are often attended with dislocations and vertical displacements. It may seem superlluous to insist upon these obvious prin- ciples of geological dynamics, but I have had occasion to notice that they are sometimes overlooked or misunder- stood even by teachers of the science to-day. § 105. Professor Bonne3% who, as we have seen, holds to the igneous origin of ophiolites, finds in the manner in which portions of the stratified silicious rock rest upon the serpentine near Figline what he regards as a " com- plete proof" of the eruptive nature of the serpentine, placing "the intrusive character of the latter beyond all doubt," while he is also satisfied that the great mass of euphotide (included by him under the name of gabbro) is "intrusive in the serpentine."* Whatever view may be held of the origin of these two rocks and their lela- tions to one another, the occurrence of the layers of recomposed ophiolitic rock (arenarla ojioUtlca and con- glomerato ofiol'dlco) interposed, as already described, be- tween the ophiolitic mass and the beds of phthanite, and even, as I observed in one section along the southeast base of the hill, the presence of fragments of serpentine in the latter, forces us to the conclusion that these sedi- mentary strata were deposited upon the ophiolite, so that the theory of the eruption of the latter since the deposi- tion of the eocene beds is untenable. § 106. The examinations which I have been able to mako of the ophiolitic rocks of Eastern Liguria, where I spent a little time near Sestri Levante, under the guid- ance of Prof. G. Uzielli of Turin, were such as to leave no doubt in my mind that we have here, as maintained by Gastaldi, portions of an ancient stratified series rising out * Geol. Magazine, Au'jtust, lS7i\ vol. vl., p. 302. 13 THE GEOLOGICAL HLSTORY OF SERPENTINES. [X. of the overlying eocene. In addition to the varieties of serpentine, and of euphotides, diorites, diabases (the ani- phimorphic rocks of Issel and Mazzuoli), we find eurites, jaspers, epidutic and steatitic rocks, with occasional lime- stoues, and various types of argillites, including the hypophthanites of these authors. The whole series, including its masses of pyrites, more or less cupriferous and niccoliferous, presents a close resemblance to the group of strata accompanying the serpentine of the Iluro- nian series in Eastern Canada, with which I have long been familiar. These rocks are well seen along the valley of the Acquafredda — near which I found, in an eocene limestone, grains of the underlying serpentine, as also evidences of a considerable dislocation since the deposi- tion of the eocene strata. My observations at this point served to strengthen my conviction that the ophiolite of Monteferrato is also but a small protruding mass of the same series. I was enabled, subsequently, as already noticed (§ 54), to examine with Signor Quintino Sella a portion of the ophiolitic series of admitted eozoic uge, as seen in the Biellese, in the province of Novara, and to confirm the judgments of Gastaldi, Cossa, Bonney, and others, as to the apparent identity of these ancient ophio- lites with those found in Eastern Liguria. § 107. We have already described, in §§ 22, 23, the mass of eozoic serpentine which in Staten Island, New York, rises from out of the horizontal or gently inclined cretaceous and triassic strata that have been deposited around its base. If now we conceive this region to be subjected to such movements as those which, along tlie eozoic belt a little farther south, have compressed tlio Primal and Auroral strata against the northwest base of the South Mountain, and given them a southeast dip, we should have a phenomenon not unlike that presented by Monteferrato; that is to say, a lenticular mass of ancient serpentine rising along the outcrop of southeastward dip- ping mesozoic rocks, and differing only by the accidental s. IX. •ieties of (the am- [ eurites, nal liiiie- cliug the le series, jpriferoua je to the the Iluro- have long the valley an eocene ^Q^ as also the deposi- [, this point Dphiolite of aiass of the as already tmo Sella a zoic uge, as rara, and to ionney, and cient ophio- \.i THE GENESIS OF SERPENTINES. 497 circumstame that these, on the two sides, belong to dif- ferent mfsoioic horizons. VI. — THE GENESIS OP SERPENTINES. § 108. As regards the origin of the serpentine-rocks, ■\ve liave already noticed briefly some of the hypotheses which have been proposed. Although those which sup- pose them derived by metasomatic changes from alumi- nous or calcareous rocks, either exotic or indigei ous, such as granites, diabases, granulites, or limestones., may be considered as now nearly obsolete, it may not be amiss to recall the fact that they represent two distinct and oppo- site schools, which agree only in admitting an unlimited alteration or change of substance in previously formed rocks, through aqueous agencies. The first view, which may be described as a general metasomatic hypothesis adapted to plutonisra, m that which derives not only ser- pentine but limestone from ordinary types of feldspathio rocks, such as granites, granulites, gneisses, diabases, and diorites. The integral conversion of all of thsse into serpentine by the complete elimination of the alumina, alkalies, and lime, and the replacement of these bases by magnesia, have be^ii maintained by many writers of repute belonging to the school in question.* § 109. Others still have supposed that the same rocks might be changed into limestone, by a complete removal of the silica, also, and the substitution of carbonate of lime. This extreme view has found its boldest and most consistent advocates in Messrs. King and Rowney, who * Bonney, who maintains tlie origin of serpentines by the hydration of eruptive clirysolite rocks, has, in his paper already cited, given many reasoi-a tor rejecting the notion of the formation of serpentines by meta- somatosis from the basic feldspathic roclcs so often associated therewith. The observed relations of the two are. in his opinion, wholly opposed to this view, and he insists upon the difficulty of conceiving that such a pro- cess of change . hould be limited to certain parts of a great mass, while leaving adjacent portions unaltered. From their distinctness, he is even led to the conclusion that the serpentines and their accompanying eupho- tides and diorites belong to successive periods of eruption. Pz~ .'•fWl" *■ '^'■' ' ^■■■1 II I m fm^,»mim "•,>ML.'^*iiiji'iHtix i \ i i III 1 1 1 If |l m iiili,:! '. 498 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. not only assert this origin for the limestone-masses found in the gneisses of Sweden and tlie Hebrides, but imagine that the bedded crystalline limestones, many hundred feet in thickness, which are interstratified in the Laurentian gneissio series of North America, and have been traced in continuous lines cf outcrop for hundreds of miles, have resulted from such an entire transformation of corre- sponding portions of the granitic, gneissic, and pyroxenic rocks of tlie series.* These very ingenious writers further imagine that serpentine also, — to which they assign, in ac- cordance with tlie received views of this school, an origin, by metasomatosis ("or, as they call it, methylosis), from dolerite, melaphyre, diorite, euphotide, and other supposed plutonic rocks, — is itself subject to a similar change into limestone. The existence of ophicalcites, the presence of masses of serpentine, and of such serpentinic structures as Uozoon Canadense, in limestone, are but so many evi- dences to them of a still uncompleted conversion of ser- pentine into limestone. § 110. Opposed to this view of the genesis of serpen- tines and limestones by change of substance, from plutonic rocks, is that which may be described as a general meta- somatic theory adapted to neptunism, and which, recog- nizing the aqueous and sedimentary origin of limestone, would derive from it, by alteration, not only serpentine, but the various other silicated rocks mentioned above. Illustrations of this are seen in the supposed conversion of limestone into dolomite, and of this last into serpen- tine, both of which views have found many advocates. The probable change of limestone into granite and into gneiss, was suggested by Bischof, and Pumpelly subse- quently, in 1873. proposed to explain the genesis of the bedded petrosilex-porphyries or halleflintas of Missouri by the transmutation of a stratified limestone, of which por- * See, for a discussion oi the views of this school, the author's Chem. and Geol. Essays^ pp. 324-325; also, An Old Chapter of the Geological Record, by King and Rowuey, 1881, chapters vii. and xii. 3. [X. X.] THE GENESIS OF SERPENTINES. 499 38 found imagine tired feet lurentian traced in iles, liave of corre- pyroxenic jrs further sign, in ac- , an origin, L>sis), from jr supposed change into presence of I structures ) many evi- L-sion of ser- s of serpen- L-om plutonic eneral meta- vhich, recog- ,f limestone, serpentine, fioned above, id conversion into serpen- ly advocates, jnite and into npelly subse- snesis of the jf Missouri by [of which por- ae author's Chem. [of the Geological 11. tions are found intcrlaminated with the pctrosilex.* He, at the same time, suggested a similar origin for the hema- titic iron-ore which accompanies these porphyries. § 111. With this second hyp(^tlie8is of the origin of serpentines may be mentioned another, not, however, involving mctasomatosis, which has sometimes been dis- cussed, and which was suggested by the present writer in 1857, from tlie results of certain experiments on the arti- ficial formation of silicates of lime and magnesia by the reaction between carbonates of thecie bases and free silica in presence of lieated solutions of alkaline carbonates. Such a reaction is not without its significance, and, as I have elsewhere shown, has doubtless played a part in the local development of protoxyd-silicates in sediments in the vicinity of igneous rocks and of thermal alkaline waters ; but as an explanation of the genesis of great masses of comparatively pure silicates, such as chrysolite, serpentine, and steatite, it is obviously inadequate, and was abandoned by the writer in 1860 for the view main- tained below.f Even if we could suppose the presence of sedimentary beds containing the requisite elements in proper proportions, it can be shown that the reactions required for the production of silicates were inoperative in the very regions where serpentine and steatite are found, since side by side with beds of these there are met with, in many places, bed^ of dolomite and of magnesite intimately mixed with quartz, sufficient in amount, if combined, to convert the accompanying carbonates into corresponding silicates. § 112. There remain, then, to explain the origin of ser- pentine, besides the three hypotheses just noticed, three others already mentioned, to which we must again refer. First of these, we have that which supposes the material of terpentine to have come from the earth's interior as an * Geological Survey of Missouri, Iron ores, etc., pp. 25-27; also the author, on Azoic Rocks, p. 194, and ante, p. 103. t Chemical and Geological Essays, pp. 25, 297, 300. ' f,i il! VA » •:» / .f»s*iiMaa^..umu:im^.. I "! t I !- r4 i 600 THE GEOLOGICAL HlSTOltY OF SEUPENTINES. IX. igneous fused mass consisting essentially of chrysolite, which by 8ubse(iuent liydnition has been changed into serpentine. Tiiis str'ctly plutonic hypothesis being, how- ever, by many geologists held to be incompatible with observed facts in the geognosy of serpentine, one which has been called hydroplutonic, and has already been set forth at length in these pages, has found advocates. These, conceding that the gcognostical relations of ser- pentine recjuire us to admit that it was laid down from water, have conjectured that a material so unlike that of ordinary aqueous sediments was ejected from the earth's interior, not in a state of igneous lluidity, but as an aqueous magma or mud, consisting essentially of a hydrous silicate of magnesia, which subsequently consoli- dated into serpentine, and even into chrysolite and ensta- tiiie. This view, as we have seen, is maintained by a sciiool of Italian geologists, and Daub' c, while holding to the origin of serpentine by the hydration of a plutonic chrysolite-rock, supposes this to have passed into a liydrous condition before its ejection.* § 113. There are, however, no facts in the history of vulcanism to justify this strange hypothesis of an erupted magnesian mud. The materials known to us as volcanic muds and ashes do not differ essentially, as regards their constituent chemical elements, from other detrital matters, and the origin of this conjecture may perhaps be traced to the unfounded assumption that chrysolite is peculiarly a plutonic mineral, and that rocks in which it and other magnesian silicates predominate are presumably plutonic in their origin. It is at best but a survival of the belief in a subterranean providence, which could send forth at pleasure from its reservoirs alike granite and basalt, chrysolite-rock and limestone, quartz-rock and magnetite. A rational science, however, seeks in the operation of natural causes for the origin of these various and unlike mineral masses, and endeavors to explain their production * Geologic Experimentale, p. 542. vywolite, red into njr, hoNV- b\o with 10 which been set (.Ivocates. ,8 of ser- owu i^om le that of lie eavth'a but as an [ally of a tly consoli- aud eiista- ained by a lile holding i a plutonio ;o a hydrous Z4 THE GENESIS OF SERPENTINES. 601 in accordance with known chemical and physical laws. Enlightened geologists are now agreed as to tiie aqueous origin of limestones, of dolomites, of iron-oxyds, and o( ([Uartz, by processes wliieh are intelligible to every chem- ist, and the formation in the humid way of the native silicates of magnesia is equally simple and intelligible. § 114. It was, as already set forth in these pages, after a careful study of natural mineral-waters and sediments, and of the chemistry of artificial magnesian silieates, that the present writer, in 18G0, ventured to assert the aqueous origin of the masses of native magnesian silicates, and their formation by reactions between the soluble silicates of lime and alkalies from decaying rocks and the mag- nesian salts of natural waters.* This view, although adopted by Delesse, as we have shown in § 11, and also, soon after, by Giimbel, by Cre'lMcr, and by Favre,t has not found general recognition. I have, however, to record the recent adhesion to it of Dieulefait, ths eminent chem- ist and geologist of Marseilles, wdiose arduous and original studies have already placed him in the front rank of stu- dents in terrestrial chemistry ; aiul also of Stapff, the learned and acute geologist of the St. Gothard tunnel. § 115. The conclusions of Dieulefait, as to the sedi- mentary character of the ser[)entines of Corsica, have au'eady been mentioned (§ 71). He rejects the plutonic hypothesis of the origin of serpentines, for the following reasons : The frequent alternation of very thin beds of serpentine with others of schists and of limestone equallj^ thin ; the changes in the constitution and composition of the serpentinic layers ; these, being in one place pure ser- pentine, become gradually mingled with carbonate of lime, which at length constitutes a largf proportion of the rock, and also forms lenticular masses .n the midst of the calca- reous serpentines. To all thes^, which are common to the serpentines of North America, we may add, as noted else- * Hunt, Chem. and Geol. Essays, pp. 122, 296, 317. t Ibid. pp. 304, 305, 347. |l 502 THE GEOLOGICAL HISTORY OF SERPENTINES. ff. , i m' 1 1 D! where, the frequent occurrence of grains, nodules, layers, or lenticular masses of serpentine in beds of crystalline limestone. Dieulefait notes, moreover, the absence of any signs of igneous action at the contact between the serpen- tines and the underlying schists. He next adverts to the lijdro})lutonic hypothesis, and pertinently asks on what gruuiids we are authorized to suppose the ejection of muds of magnesian silicate from the earth's interior. § 116. His own conclusion is that, while these serpen- tines are sedimentary rocks in the most complete accepta- tion of the term, the mud or sediment which gave rise to them was not ejected from below, but was formed in estu- aries of the sea, by reactions between the silicious matters derived from the decay of pre-existing rocks and the mag- nesian salts of the sea-water; in which connection he insists upon the frequent metalliferous impregnations of the serpentines, as derived in like manner from the older rocks. This view of Dieulefait's, set forth in 1880,* is, as Lotti remarks, no other than " the hypothesis enunciated by Sterry Hunt," twenty years earlier.f Lotti, for his * Comptes Rendus de I'Acad. des Sciences, xcl., 1000. [t This has since been clearly stated by Dieulefait himself, in a recent elaborate memoir on " Les Roches Ophitiques des Pyrenn^es," the result of a scientific mission confided to him in 1880 by the Minister of Public Instruction in France (Ann. des Sciences Geologiques, 1885, xvi.). There- in, using the word ophite as synonymous with the term ophiolite, em- ployed in these pages to designate not only serpentine, but its associated euphotides, diabases, diorites, etc., he writes: "The whole of the reason- ing, and the facts already resumed, lead to the conclusion that the ophitic and serpentinous rocks are of sedimentary origin; they have come into the condition in which we now see them entirely through the influ- ence of chemical reactions in the wet way, and without ever having suf- fered the action of heat from without. . . . Following this conclusion, I consider it my duty to explicitly formulate the following declaration: In France, an honored veteran in geology, Virlet d'Aoust, first enunciated for the Pyrennees the view that the ophiMc rocks are of sedimentary ori- gin. This opinion was soon after accepted by a geologist of great merit, Garrigou, to whom France has not sufficiently rendered justice. But the philosopher who first set forth the question of the sedimentary origin of the ophites in all its bearings, is the illustrious chemist and geologist of Canada, Sterry Hunt. When, in a time which I hope is not far off, the > ,1 ES. tX. 3S, layers, rystalline ice of any he serpen- erts to the 3 on what )n of muds Bse serpen- te accepta- Tave rise to Lied in estu- ous matters ad the mag- Qnection he egnations of )in the older 1880,* is, as enunciated otti, for his iself , in a recent lees," the result nister of Public 5r,,xvi.). There- Ti opliiolite, em- Lit its associated ,i(> of the reason- that the ophitic have come into irough the influ- ever having suE- this conclusion, dug declaration: ;, first enunciated 'sedimentary orl- st of great merit, justice. But the raentary origin of , and geologist of Ls not far off, the X.1 THE GENESIS OF SEEPENTINES. 50;] part, while still reserving himself on the question of the supposed tertiary serpentines of Italy, adds, after his own studies of those of Corsica : " In any case, it is impossible, as Dieulefait has said, to regard the phenomena offered by these ancient serpentines as due to eruptions, either of igneous or hydroplutonic magmas. The serpentine has either been deposited as such, as maintained by Sterry Hunt, and by Dieulefait, o' is a sedimentary rock subse- quently altered." * We shall notice later on the views of Stapff on this subject. § 117. The masses of rock known as serpentine are far from homogeneous in composition. Apart from the ad- mixtures of carbonate of lime, dolomite, and magnesian carbonate, which often enter into their composition, they occasionally include, besides the hydrated silicate, serpen- tine, the anhj'drous species, chrysolite and enstatite or bronzite, and more rarely the hydrous species, talc ; sili- cates differing widely in density, in chemical stability, and in the oxygen-ratios between the silica and the fixed bases ; that for chrysolite being 1 : 1, for enstatite 2 : 1, for talc, approximately, 3 : 1, and for serpentine 4 : 3. These differences, in the hypothesis of the aqueous origin of ser- pentine, may well depend upon variations in the composi- tion of the generating soluble silicates, and upon the balance of affinities between silicic and carbonic acids in the watery manstruum, rather than upon the subsequent transformation of one magnesian silicate to another by addition or elimination of silica or magnesia. The asso- ciation, in the same mass, of anhydrous chrysolite with ser- pentine is generally regarded as evidence of the change of chrysolite into serpentine ; but, while admitting the con- conception of the sedimentary origin of the ophites sliall have definitely talcen its place in science, the present geologists, and, above all, those of a future generation, will never forget that the promoter and one of the most active workers in this great and fruitful scientific revolution was Sterry Hunt."] * Lotti, Appunti Geologici sulla Corsica; Boll. R. Comitato Geologico, anno 1883. "■fc' f » i'!i.i^.Sji.;!*«.*rt«.*jjw, 504 THE GEOLOGICAL HISTORY OF SERPENTINES. [X. h version, under certain conditions, of both enstatite and chrysolite into hylrous silicates, the view which supposes the chrysolite or the enstatite to be simply an instance of the crystallization of an anhydrous silicate in the midst of an amorphous hydrous silicate, is more consonant with the hypothesis of the aqueous origin of serpentine-rocks. It is well known that Scheerer, from his studies of the asso- ciated chrysolite and serpentine of Snarum, was led to reject the notion of the derivation of this serpentine from a previously formed chrysolite, and to maintain a simulta- neous formation of the anhydrous and the hydrous mag- nesian silicates.* A somewhat analogous case is presented in the occur- rence of grains of anhydrous alumina or corundum found in the earthy and amorphous aluminous hydrate, bauxite, which forms beds in uncrystalline cenozoic rocks.f The notion which has been advanced that the bauxite has come from the hydration of previously formed beds of corundum, is obviously untenable, and we must regard this anhydi'ous alumina as formed by crystallization in the midst of the uncrystalline mass of hydrated alumina. De Senarmont, in the decomposition of aqueous solutions of chlorid of aluminium, at 250" C, observed a simultane- ous production of anhydrous alumina in the form of corundum, and of hydrous alumina as di">spore, both crys- tallized. J § 118. The late studies of Arno Behr throw further light on the association of hydrous and anhydrous spe- cies. He has found that solutions of dextrose, within very narrow limits of temperature and concentration, yield crystals either of hydrated or anhydrous dextrose. -U * Scheerer. Pogg. Annalen, Ixvlii., 319, and Amer. Jour. Science [2], v., 389, vl., 201, also xvl., 217. t Devllle, An. de Ch. et de Phys. f.S], Ixi., 309, and IJvnt, Origin of Some Magneslan and Aluminous Rocks. Amer. Jour. Scl., 18G1 [2], xxxli., 281; also, Chem. and Geol. Essays, p. 326. t Comptes Rendua de I'Acad. des Sciences, 1856, xxxU., 762. ^< If aci THE GENESIS OP SERPENTINES. 605 and that under certain conditions we can obtain an ad- mixture of the two, as tlie result of simultaneous crys- tallization.* An illustration of the influence of small variations in composition on the result of a chemical pro- cess, under conditions otherwise similar, is afforded by the recent experiments of Friedel and Sarrasin on the artifi- cial production of albite in the wet way. \\'^hen a solution of silicate of soda mixed with silicate of alumina, in the proportions required to form the soda-feldspar, was heated in close vessels to from 400° to 500° C, no albite was formed, but crystals of the hydrated double silicate, anal- cite ; silica, soda, and some alumina remaining in solution. When, however, an excess* of the alkaline silicate was employed, the whole of the silicate of alumina was con- verted into a crystallized anhydrous compound, which was albite.f § 119. Much obscurity still surrounds the question of the conversion of chrysolite into serpentine. In the first place, it is to be remembered that the process is one which does not, under ordinaiy circumstances, take place at or near the surface of the earth, since chrysolite-roc.:s, whether exotic masses or indigenous crystalline schists, are often met with, presenting no evidence of such change. This is Avell seen near Montreal, where the hills of chryso- litic dolerite, demonstrably of pre-Silurian age, as well as fragments of the same rock imbedded in Silurian con- glomerates, alike contain only unaltered anhydrous chryso- lite. This mineral, on exposed surfaces, is subject to a sub-aerial decay, analogous to that suffered by pyroxene and amphibole, by which the magnesia and a large propor- tion of the silica are removed, leaving a residue of ferric oxyd, as long since observed by Ebelmen. The change of chrj'solite into serpentine must then be distinct fi-om * For these facts I am indebted to a private communication from Dr. Belir. See also his paper in Jour. Amer. Chem. Soc, in 1882, vol. iv., p. 11. t Comptes Rendus de 1' Acad, des Sciences, July 30, 1883. jU '■^Hiikmm ■-'.■^l^^¥!^"/^k^.'mM ^JaSishJimv^s-s.-if- ,>w.v:- ciiasB;' 506 THE GEOLOGICAL HISTORY OP SERPENTINES. tX. that going on under the influence of atmospheric waters near the surface. § 120. One hundred parts by volume of chrysolite, with a spfcjific gravity of 3.33, if converted into a serpen- tine of specific gravity 2.50, without change in its con- tent of silica, must lose one-eighth of its weight of magnesia, and acquire the same amount of water instead, while, at the same time, its volume will be augmented by one-third, or to one hundred and tliirty-three parts. I have long since discussed this matter in connection with Schecrer's views as to the relations of these two mineral species, noticed in § 117. A simple hydration of chryso- lite would yield, not serpentine, but villarsite. Serpen- tine, when subjected to dehydration and fusion, yields, as was shown by the experiments of Daubrii2»#2^f**f'' 612 THE GEOLOGICAL HISTORY OP SERPENTINES. [X. i; ■ < ! mass, intercalated conformably in the stratification (like the layers of eulysite in the gneiss of Tunaberg, in Sweden), and now appears, as the result of numerous breaks and displacements, outcropping in a series of little lenses, the line joining which intersects at a sharp angle the schistose lamination of the beds. Near to the fissures which, with displacements, cut the mass, the rock adjoin- ing the serpentine is stretched out and pushed back (etiree et re fo idee'), both at the surface and in the interior of the tunnel." § 129. This displacement in one case, on the surface, was found equal to 450 metres, and the adjacent strata were bent in the form of an inverted C. Tiie m.ixinmm thickness of the serpentine at the outcrop Avas 100 metres, and the thickness of 440 metres, which it attains in ^he line of the tunnel, is believed by the author to be due to the accumulation, by the movements described, of succes- sive portions of one and the same lenticular mass : a con- clusion which is illustrated by a great number of minute observations. He adds, " The fissures along which this heaping together must have taken place, present striations produced by the sliding of the rock ; they are coated with a steatitic matter, and sometimes filled with a friction- breccia. Farther proofs of this crushing are found in the abrupt discontinuity of the schistose and compact portions of the serpentine, and in the indented outline presented by the upper surface of the serpentine-mass ; a detail not represented in the profile." The author farther says: " Although we would not consider the serpentine to be an intrusive rock, we must remark that it could not have had precisely the same [mechanical] sedimentary origin as that which we have sui)posed for the micaceous gneiss which encloses it. We may regard it as originally a deposit of hydrated silicate of magnesia, formed by springs, and en- closed between the sediments which gave rise to the mica- schists." The hydrated magnesian silicate is supposed by our author to have been subsequently converted into an- E3. IX. tion (like uibevg, ill numerous es of little harp angle the fissures :)ck adjoiu- ished back the interior the surface, acent strata le maxiumm 1 100 metres, ,taiiis in +he to be due to ;d, of succes- mass : a con- )er of minute ig which this ient striations e coated with 1 a friction- found in the ipact portions ine i)resented a detail not farther says: ■ntine to be an not have had 7 origin as that i gneiss which y a deposit of )rings, and en- ,se to the mica- is supposed by erted into an- S.) STUATIGRAl'HY OF SERPENTINES. 613 hydrous chrysolite, etc., which by a later hydration has generated serpentine, portions of chrysolite still remain- ing in the mass. It may be questioned whether the phe- nomena require this hyi)othesis of a double change for their explanation. The serpentine contains imbedded, in some portions, not only chrysolite, but hornblende, talc, and garnet. Intercalated with the serpentine, which is often distinctly stratified, are layers of schistose talc, of compact chlorite, of actinolite-rock, of ferriferous dolo- mite, and of mica-schist. The serpentine itself is chromif- erous, and also contains magnetite. § 130. Stapff farther adds : " The curious modifica- tions of form which the mass of serpentine has suffered from the effect of faults, etc., correspond to those of the adjacent micaceous gneiss, but in the case of the former they have been better studied, for the reason that it u more easy to define the limits of these forms. If we sup- pose in the section, in place of the serpentine, a mass of ordinary micaceous gneiss subjei 3d to all the movements of displacement and elevation which we have here dis- played, we should perceive nothing more upon the profile than a uniform surface of micaceous gneiss, with some interlacings of beds. It cannot, however, be denied that movements arrested by the hard and tough mass of the serpentine have produced in the neighboring rocks per- turbations much more intense than would have resulted from similar movements acting upon a more tender rock." (Loc. cit., pp. 43-44.) It would be difficult to illustrate more clearly than Dr. Stapff has done, the manner in which movements in the earth's crust may affect inter- stratified masses of unequal hardness and tenacity, giving rise to accidents which simulate to a certain extent those produced by the intrusion of foreign masses, and may thus lead different observers, as we have seen, to opposite conclusions with regard to the geognostical relations of rocks like serpentine and euphotide. m ^m iK^ 1 J ,; ■ 1^'^ i I! W I' >. >l 614 THE GEOLOGICAL HISTORY Or SERPENTINES. PC CONCLUSIONS. The following are the chief points regarding serpentine and ophiolitic rocks which we have sought to set forth in the preceding pages : — 1. To siiow historically the diversity of opinions as to the geognostical relations of Kerpentine and related rocks, which have heen regarded by some writers as eruptive and of igneous origin, and by others as aqueous and sedi- mentary. 2. To show how, from the hypothesis of their eruptive origin, came tlie application of that of metasomatosis, and also to set forth the hypothesis of the aqueous origin of serpentine, explaining how silicates of magnesia may, on chemical grounds, be looked for at any geological horizon. 3. To indicate the various horizons at which serpen- tines are found in North America ; and first, those of the Laurentian, of the Huronian, and of the younger or Montalban gneisses ; in which connection we have noticed the serpentines of Chester County, Pennsylvania, and those of New Rochelle, Hoboken, and Ma'diattan and Staten Islands, all of which are regarded as indige- nous stratified rocks ; the apparently intrusive character of the serpentine of the latter locality being explained. 4. We have further described the occurrence of serpen- tine among the Taconian rocks in Pennsylvania, and also among the gypsiferous rocks of the Silurian series at Syracuse, New York. 5. Having noticed some points regarding the nomen- clature of serpentin., and related rocks, and Bonney's account of the serpentines of Cornwall, and of parts of Italy, we have considered the serpentine-bearing rocks of the Alps, in which we show four great groups, in ascend- ing order, which are the older gneiss, the pietre-verdi or greenstone series, the newer gneisses and mica-schistp, and the still younger lustrous schists, corresponding re- IE8. serpentine jet forth in iuions as to jlated rocks, as eruptive )U8 and sedi- heir eruptive letasomatosis, qneous origin iiag^iesia may, ,ny geological which serpen- jt, those of the he younger or -tion we have rennsylvania, and Manhattan irded as indige- ^•usive character ng explained, rence of serpen- llvania, and also durian series at X.) CCMCLUSIONS. 615 sitectivoly to the Lauroiitian, Iluronian, Montalban, luul Tucoiiiiiii of North America; the second aiul thud of these being the Puhidian and the Grampian of iJreat Britain.* Serpentiney, It was shown, occur in the Alps intcrstratified in the second, third, and fourth of tliese groups, the youngest of which includes the marbles of C.'arrara. 6. The view that this youngest group is mesozoic, is discussed, and the relations of all these groups of crystalline schists to the fossiliferous rocks of the main- land, and of those of Elba and Sardinia, are sot forth, showing their pre-Cambriau age ; while it is maintained that the ophiolKcs and other crystalline rocks which have there been referred to tl;e tertiary are but exposed por- tions of these pre-Cambriau rocks. 7. The crystalline rocks of the Simplon and the St. Gothard, and those of Saxony and Bavaria, are considered, and are compared with the younger gneisses of North America. 8. The relations of the so-called tertiary terpentines to the surrounding strata are elucidated by a detailed discus- sion of the mass of Monteferrato, in Tuscany, which 's regarded as of pie-Cambrian or eozoic age. * It re.- Jus to be seen whetlier the Arvonian series, which is essen- tially composed of stratified Uii'leflinta or petrosilex-rocks, passing into quartziferous porphy.ies, and is largely developed at the bas» he Huronian in parts of North America, and of Great Britain, is not repre- sented In the Alps. Since we have seen the serpentines, Iherzolites, euphotides, diabases, and even the marbles of the Alps and other regions, removed from the category of eruptive mesozoic and cenozoic masses, and shown to be regularly interbedded members of pre-Cambrian stratified series, it is, I think, a leglti-. ate subject for inquiiy whether the quartzi- ferous porphyries which are so largely developeil at Uotzen, and elsewhere in the Alps, and have been regarded as eruptive rocks of Permian age, may not prove to belong to a stratified series, the equivalent of the Arvonian, with which, to judge from descriptions, analyses, and speci- mens, they bear a close resemblance. For an account of these rocks of Botzen by one who regards them as plutonic, see Judd in the Geological Magazine for 1876, vol. xiii., pp. 200-214, and for details with regard to the history of the Arvonian series, see the author ia 1880, American Jour. Science [3] (xix., pp. 274, 278, et seq.); also ante, page 409. l<4' '% . # 616 THE GEOLOGICAL HISTORY OF SEKPENTINE3. [X. 9. TL'3 various theories proposed to explain tlie genesis of serpentines are considered, and that of their aqueous origin is adopted. 10. The geognostieal history of chrysolite is discussed, and the essentially neptunian origin of many chrysolite- rocks is mdntained. 11. The contradictory views as to the geognostieal rela- tions of serpentine are considered, and an attempt is made to show that the appearances of intrusion, upon which some have insisted, are explained by subsequent move- ments of the strata in which the serpentines are included. >t til XI. THE TACONIC QUESTION IN GEOLOGY. In the investigation of the age and relations of the crystalline stratified rocks, it became necessary to consider a great series of strata which by Maclure had been culled Transition, but by Eaton were, for the greater part, included in his Primitive divisions, tlioiigh placed by him stratigraphically between the gneisses and related loclis on the one hiind, and the paleozoic rocks on the otlier. That at a later period this intermediate series, through the extension of the doctrine of regional metamor- phism, came to be regarded as a local alteration of the lower part of the paleozoic, and that it bus been a source of much controversy, are facts well known to all geolo- gists. It had been the task of the writer to attempt with some success the unravel- ling of the famous Cambrian and Silurian controversy into which the errors of Murchison had introduced confusion, and still it now remained to essay the heavier task of solving the greater problem presented by a vast series of rocks, not less widely spread, which had perplexed the geologists of a generation ; a problem closely con- nected, moreover, with the Cambrian and Silurian controversy, and involving still wider discordances of opinion, greater contradictions, and more important results for the science of geology. Partial and limited observations, partisan spirit, a neglect of geological literature, and false notions of metamorphism, had each con- tributed to obscure the question. A personal examination of localities throughout tlie country, a critical study of the literature of the Taconic controversy, and a candid discussion of all the facts there made known without regard to the precon- ceived opinions of myself, or of others, were evidently necessary for the solution of the problem. The reader of the following pages must judge to what extent these ends have been attained. This essay was presented to the Uoyal Society of Canada, the first portion, to the end of § 135, on the 23d of May, 1883, and the remainder on the 21st of May, 1884. These two portions have been published respectively in the first and second volumes of the Transactions of the Society. Additional para- graphs regarding the Green Pond Alountain range of New Jersey, the Taconian and Keweenian of Lake Superior, and the Keweenian and Cambrian of Texas and the great American b.isin, have been added, giving the results of later studies. i I II i ' ltl 111 -I I. — INTEODUCTION. § 1. The history of those stratified rocks which in eastern North America have been called the Taconic series, is one of many contradictory opinions, and of much obscurity. Taken in the larger sense in which the name was at one time applied, this history moreover in- cludes, besides those rocks to which the appellation of Taconic or Taconian was subsequently restricted, another important series, sometimes called the Upper Taconic, which, under the names of the Hudson-River group and 517 H m wt 'm w \ w m ' 'In J m i m ^1 m 'i\ !'^ M ■ 'mjmmki&iiisi;^! 518 THE TACONIC QUESTION IN GEOLOGY. [XL the Quebec group, has been the subject of prolonged controversy. It may here be noted that one of the latest writers on the subject, whose views will be discussed in the present essay, still maintains for this latter series the name of Taconic* The questions involved in this his- tory are of fundamental importance, and have hitherto been involved in so much misconception that it seems desirable at the present time to give a concise view both of the fabts and of the various theories which have been held with regard to the whole of the rocks in question. For this purpose we must go back to Amos Eaton, to whom rightly belongs the honor of having laid tlie foun- dations of the American school of geology, so worthily continued by his pupils, James Hall, George H. Cook, and the late Ebenezer Emmons. It is now half a century since, in 1832, appeared the second and revised edition of Eaton's " Geological Textbook," from which we may gather his matured views as to the geological succession in northeastern America. From this, and from his pre- vious " Geological and Agricultural Survey of the Erie Canal," published in 1824, I have elsewhere endeavored to frame a connected statement of these views,t which is here brieflv resumed. § 2. Dividing the stratified rocks of northeastern America into five great groups, — namely: I. Primitive; II. Transition ; III. Lower Secondary ; IV. Upper Second- ary; V. Tertiary, — Eaton supposed that each of these groups "com mei»ced with carboniferous slate, and termi- nated with calcareous rocks, having a middle formation, the centre of which is quartzose." This three-fold divis- ion and alternation in each great series, which Eaton regarded as universal, was, so far as I know, the first re- cognition of the principle, now so generally understood, of cycles in sedimentation. § 3. These three divisions evidently correspond to ar- * Marcou, Bull. Soc. G^ol. de France, 1880 ; (3), ix., p. 18. t Azoic Kocks, etc., pp. 24r-29. 0U prolonged the latest Lscussed in series the LU this his- ve hitherto it it seems j view both L have been qviestion. 3s Eaton, to id the foun- so worthily ye H. Cook, alf a century vised edition hich we may •al succession from his pre- Y of the Erie •e endeavored ws,t which is northeastern I. Primitive; ;pper Second- each of these lite, and termi- Idle formation, [iree-fold divis- which Eaton |w, the first re- lly understood, [-respond to ar- }), ix., p. 18. XL] THE TACONIC QUESTION IN GEOLOGY. 519 gillites, sandstones, and limestones, but in the application of his scheme its author allowed himself considerable liberty of interpretation, and referred to his first or argil- laceous division, not only clay-slates, but the great body of crystalline schists. In this way, the first division of the Primitive series was made to embrace both the gneisses of the Adirondacks and the Highlands of the Hudson, and the unlike crystalline rocks of New England ; includ- ing, besides gneisses, various hornblendic, chloritic, and micaceous schists, in some of which strata the occurrence of graphite was held to justify the title of " carbonifer- ous," applied to these rocks as a whole. Following this first division of the Primitive series (I. 1), Eaton recog- nized in western New England the second or silicious division (I. 2), and the third or calcareous division (1. 3), represented respectively by the Granular Quartz-rock and the Granular Lime-rock or marble of the Taconic range. § 4. Succeeding these, came the rocks of his second or Transition series. Of this, the first or carboniferous divis- ion was the Transition Argillite (II. 1) which in many localities directly overlies the Primitive Lime-rock, and consists in part of roofing-slates, with coarser and more silicious layers, and in part of soft unctuous micaceous schists. To this Transition Argillite succeeds, according to Eaton, the First Graywacke or Transition Graywacke, representing the second or silicious division of the Tran- sition series (II. 2), and consisting of the so-called gray- wacke-slate, with sandstones and conglomerates. The base of the First Graywacke was declared to rest uncon- formably upon the Transition Argillite. § 5. The geographical distribution of the First Gray- wacke — a very important point in our present inquiry — was carefully indicated by Eaton. " It is seen resting on the Argillite, near Col. Worthington's on the Little Hoosic, near the eastern limit of Rensselaer County. On ascending the western hill or ridge, the graywacke-slate, rubble, and millstone-grit [elsewhere indicated by Eaton ^ym ii ill ''1 mm mm li'- I i&M»^s*.fei^aft^■it'*«j;■A«»8;.i.'s>;■;!.a«>»*•c 620 THE TACONIC QUESTION IN GEOLOGY. txi. as making up the First Graywacke] are found in succes- sion. This ridge extends from Canada, through the State of Vermont and Washington, Rensselaer and Columbia Counties in New York." Elsewhere, we are told by Eaton that the rubble or conglomerate of this First Gray- wacke " forms the highest ridges between the Massachu- setts line and the Hudson." He also supposed that the Shawangunk Mountain of Ulster and Orange Counties, on the west side of the Hudson, now referred by New York geologists to the h orizon of the Second rrray wacke, "is a continuation of the grit and rubble of the First Graywacke of Rensselaer County." * § 6. To the third or calcareous division of the Transi- tion series (H. 3) was referred by Eaton, what he called the Sparry Lime-rock, found at the summit 'of the First Graywacke, to the east of the Hudson. In the same divis- ion also was included a group of strata lying to the west of Lake Champlain, which he designated the Calciferous Sand-rock and the Metalliferous Lime-rock. In this latter region, however, these Transition limestones were found to rest directly upon the lower division of the Primiui -e series ; the whole of the intermediate divisions beJig absent. § 7. The Transition limestones in this western area were, according to Eaton, directl} followed by the third or Lower Secondary series, the first division of which ( III. 1 ) was described as an argillite or gray wacke-slate, and the second (HI. 2) as a sandstone or millstone-grit ; the two together making what he called the Second Gray- wacke ; t^'eclared by him to be indistinguishable from the First or Transition Graywacke, except by the fact that it overlies the Transition limestones. This Secondary Gray- wacke is thus clearly indentified with the strata subse- quently called the Utica slate, the Pulaski or Loraine shales, and the Gray or Oneida sandstone. Succeeding it, were the Lower Secondary limestones, including the Niag- • Geological Textbook, 2d ed., pp. 74, 93, 123. [XL. Q. bucces- the State Columbia told by irst Gray- Massachu- L that the Counties, \ by Hew rraywacke, the First ;he Transi- b he called f the First same divis- to the west Calciferous ;n this latter I were found le Primi ti '6 Lsions be).l.').\ il' '. . W i ■'*t !,f,; i! 522 THE TACONIC QUESTION IN GEOLOGY. [XI. and Black-River limestones; 5. Utica slate; 6. Loraine shale ; 7. Gray sandstone ; 8. Medina sandstone. The numbers 7 and 8 were, as is well known, subsequently separated from the Champlain division, and joined to what was called the Ontario division of the New York system. § 9. This order was evident to the west of Lake Champlain, where the strata are nearly horizontal, and rest in undisturbed succession on the crystalline rocks of the Primary series. To the east of the lake, however, and thence southward along the valley of the Hudson, is found the belt of disturbed strata, dipping generally to the eastward, which had been called by Eaton, the First or Transition Graywacke, the distribution of which has been given in § 5. This belt, described by Emmons as consisting of a great thickness of green, red, and gray sandstones and conglomerates, with green, purple, and black slates, and some associated limestones, was by him now referred to the horizon <.f the Loraine shale and the succeeding sandstones (Nos. 6, 7, and 8), or, in other words, to the Second Graywacke, lying above the Trenton limestone; which latter, according to Emmons, appears, in some localities, to dip beneath this graywacke. By Eaton, however, the strata of the same belt had been assigned, under the name of the First Graywacke, to a position below the same Trenton limestone. § 10. These views were published by Emmons in 1842, at which time, as we see, he dissented from the opinion of Eaton as to the stratigraphical horizon of the First Gray- wacke of the latter, and adopted that which had been put forth by Mather, to be mentioned below. As regards the quartzite and limestone of the Primitive series, and the. Transition Argillite, which, according to Eaton, intervene stratigraphically, as well as geographically, between the crystalline schists of the Primitive and the First Graywacke, Emmons supposed that these three divisions constitute a distinct group or series, which, from [XI. , Loraine le. The sequently ioined to ^ew York ; of Lake ontal, and e rocks of iwever, and Hudson, is enerally to 1, the First which has Emmons as i, and gray purple, and was by him lale and the |or, in other the Trenton )ns, appears, pvacke. By 5lt had been ywacke, to a mons in 1842, ,he opinion of } First Gray- ich had been , As regards ire series, and ig to Eaton, eographically, litive and the ; these three |s, which, from XI.] GEOLOGICAL SURVEY OF NEW YORK. 523 its development in the Taconic hills of western "Massa- chusetts, lie named the Taconic system. This he regarded as distinct from, and older than, the New York system lying to the westward of it. § 11. The survey of the Southern district of New York was a'^signed to Mather, who, in his final report on the region, in 1843, described the southward extension of the various groups of rocks just mentioned, and main- tained, in opposition to both Eaton and Emmons, that the Taconic system of the latter was a modification of the Champlain division ; the quartzite being supposed to correspond to the Potsdam, the marble to the Calciferous, Chazy, and Trenton, and the argillite to the Utica and Loraine ; for which latter subdivision he adopted, as a synonym, the name of the Hudson slates. § 12. As regards the gray wacke-belt east of the Hudson River, this consisted in part, ac Hng to Mather, oi che same slates in a disturbed and altered condition, and in part of higher strata, belonging to the horizon of the Oneida and Medina subdivisions of the New York system. He supposed, with Eaton, that the belt of these rocks, continued from Canada, through Vermont, and along the east side of the Hudson, was prolonged south- ward, on the west side, in the Shawangunk range ; and that the Green-Pond Mountain range, in New Jersey, was also a portion of the same belt, which it lithologically resemblco. Thus, in the view of Mather, the whole series of Eaton, from the granular quartzite of the Primitive up to the top of the Second Graywacke, was made up of the rocks of the Champlain division, with some still higher strata. He confounded the First with the Second Gray- wacke, and supposed both the clay-slates of the former, and the underlying Transition Argillite of Eaton to be nothing more than local modifications of the Utica and Loraine shales. Lideed, as is well known, Mather went 80 far as to regard the Primitive crystalline schists them- selves as a farther modification of the same Champlain j| fegSBj^*', '■ ■K' ^^^^Bs'^ -i Hi! ^n V HIHif''il'il HkHHm P'r - ' '': ^^HmH S'''i ' 11 ^^^^HKi ^' ^ l^raili ^^Hlf 1 H|B||\i ^^^Hii'tl'' t ' ■'!CSS^i»Ta"aE:?r~rmru: u *&i. ' 524 THE TACONIC QUESTION IN GEOLOGY. PSti series. Emmons, as we have shown, while adhering to the views of Eaton in other respects, adopted, at this time, Mather's conjecture as to the horizon of the eastern graywacke-belt. § 13. Tlie name of Hudson slates had already, in his fourth annual report on the Southern district of New York, been given by Mather to the strata which he re- garded as equivalent to the Loraine shale ; described by Emmons as occurring in Jefferson and Lewis Counties, in the Northern district. These strata were farther studied by Vanuxem in the Central or intermediate district, which included the counties of Oswego, Oneida, Herkimer, and Montgomery, extending southeastward along the valley of the Mohawk. The rocks found in this district were first described by Conrad as dark shales (the Utica slates) suc- ceeded by fossiliferous lead-colored shales alternating witl gray sandstones, well displayed at and near Pulaski, in Oswego County. At the summit of these was a sand- stone quarried for grindstones, and in Oneida County the series was overlaid by a quartzose conglomerate. These were at first called by Vanuxem (who succeeded Conrad in the charge of the survey of this district) the Pulaski shales and sandstones, and they clearly correspond to the Loraine shale and the Gray sandstone of Emmons. As these shales were also regarded, both by Emmons and by Vanuxem, as identical with the Hudson slates of Mather, Vanuxem included them in what he called the Hudson- River group ; a name which, in subsequent geological and paleontological publications, has generally replaced that of Loraine shale, as being synonymous with it. § 14. The Hudson-River group, however, according to Vanuxem, embraced two distinct divisions, the upper, a highly fossiliferous member (being the Pulaski shales and sandstone), found west of the Adirondacks, in Jefferson, Lewis, and Pulaski Counties, and disappearing to the southeastward, in Oneida County. The lower member of the Hudson-River group, as defined by Vanuxem, was aC an. Ihering to d, at this he eastern 3ady, in Ws ict of New hich he re- escribed by Counties, in ther studied strict, which erkinier, and the valley of ■ict were first ja slates) suc- ernatingwitl ir Pulaski, in ; was a sand- da County the lerate. These leeded Conrad :) the Pulaski •respond to the I Emmons. As Lnmons and by Les of Mather, jd the Hudson- geological and replaced that of It. |er, accordmg to Is, the upper, a llaski shales and Iks, in Jefferson, Ippearing to the [lower member ot Vanuxem, was XI.] G'^OLOGIOAL SURVEY OF NEW YORK. 525 named tin I'Vankfort division, from Frankfort, in Herki- mer Cuuiu^ and was described as consisting of greenish argilliies and sandstones ; which underlie the Pulaski shales to the northwest, as far as Jefferson County, constitute, in Herkimer and Montgomery Counties, the only representa- tive of the Hudson-River gronp, and extend eastward, througli Schenectady, Albany, and Saratoga Counties, to the Hudson River. This lower division of tlie group was said to contain none of the organic remains of the Pulaski or up])er division, but to include some graptolitic shales. To this lower division, Vanuxem supposed, might belong the thick masses of contorted argillaceous strata, of "con- troverted age," along the Hudson valley. He farther remarked that the two divisions of the Hudson-River group " are not co-extensive with each other. The lower one enters from the Southern district, along the Mohawk, and extends north by Rome, thi'jugh Lewis into Jefferson County. The upper division first appears in Oneida County, and from thence, west and iKjrth, is a co-associate of the Frankfort slate or lower division." These two divisions Vanuxem insisted on treatij.^ separately, "inclining to the opinion that they ought not to be put together in local geology." * He, moreover, declared that the two divisions, although in juxtaposition in parts of New York, occur separately in Pennsylvania. The Pulaski shales, having in all respects the same characters as in New York, it was said, are found in the Nippenose valley, west of the Susquehanna ; while the Frankfort slates and sandstones are seen to the east of the North Mountain, in the Kittatinny or Appalachian valley, and include the roofing-slates of the Delaware. The Oneida conglomerate, which in Oneida, New York, according to Vanuxem, rests upon the Pulaski shales, is seen in Herkimer County, overlying directly the Frank- fort slates and sandstones. The same conditions, accord- ing to Horton (Mather's assistant in the Southern district * Geology of the Third district of New York, pp. 60-67. i\ ■^1^^-:^ i.i ', '. I < 626 THE TACOXIC QUESTION IN GEOLOGY. oci. of New York), occur in Orange County, wliere the sand- stone of Sliiiwangunk Mountain is said to rest unconform- ably upon the edges of the Argilli^e and Graywacke series. § 15. The table on page 529 will show the relations of the various groups of stj-ata already noticed, by a com- parison of the divisions established by Eaton with those adopted by the New York geologists, and by others. It should here be repeated that Eaton insisted upon the fact that the Argillite is unconformably overlaid by the First Graywacke. He wrote, " while European geologists have described a change of direction at the meeting of the Lower and Upper Secondary, in which the latter rests un- conformably upon the inclined edges of tho iormer, in North America this change takes place at the meeting of the Argillite and the First Graywacke." He was careful to distinguish between the bedding and the slaty cleavage of the Argillite, the plates of which, he tells us, " form a large angle with the general direction of the rock." His diagrams, moreover, show both the non-conformity of stratification between the two, and the independent slaty cleavage of the lower series.* Eaton did not distinguish the Potsdam sandstone on the west shore of Lake Champlain from what he called the Calciferous Sand-rock, there underlynig the Metallif- erous Lime-rock, — a term (borrowed from Bakewell) by which he designated the Trenton limestone, with its sub- divisions, including what he called the Birdseye or Encri- nal marble, and the underlying Chazy. The Calciferous Sand-rock he described and figured as in part marked by geodes (a very distinctive character), and represented it as the equivalent of the somewhat dissimilar Sparr}^ Lime- rock, found, to the eastward, at the summit of the First Graywacke. Of this Sparry Lime-rock, he both desig- nated and figured two varieties, which he called " veiny " and " tessellated." The correctness of these and of other • ♦ Geological Textbook, pp. 63, 72, 74. the sand- inconform- Grraywacke iie relations I, by a com- \vith those - others. It pou ibe fact by the First ologists have etiug of the tter rests un- lu; iormer, in lie meeting of le was careful slaty cleavage Is us, " f o"" ■'^ ^e rock." His ^conformity of ependent slaty XI.] OEOLOaiCAL SURVEY OF NEW YORK. 627 descriptions by Eaton, will be acknowledged by those who examine carefully the rocks which he described. § 10. In the Lower Secondary of Eaton, what he named the Corniferous or Cherty Limo-rock, with its beds of chert (called by him "stratified horn-rock"), is the Upper Heldcrberg of later geologists, and his Geodiferous Lime-rock is as clearly the Niagara; the Lower Helder- berg limestone, and the succeeding Oriskany sandstone, now regarded as the basal member of the Devonian, not being then recognized. Besides the regular division of each group into triads of argillaceous, silicious, and calca- reous rocks, which lie regarded as normal, Eaton admitted the existence of what he called subordinate or interposed strata. To this class of abnormal rocks, he referred, in the Lower Secondary, the Onondaga group, with its marls, salt, and gypsum, and also the hydraulic limestone or Water-lime above it; all of which may be regarded as interpolated between the Niagara and the Helderberg hraestones. In the same subordinate class, also, were in- cluded by him the red beds of the Medina and the iron- ores of the Clinton. § 17. It will be remembered that the Potsdam of Emmons, which (like the Calciferous Sand-rock) is often wanting at the base of the Champlain division, was un- known to Eaton, and hence does not appear in our table, from which what he regarded as subordinate strata are also omitted. The Calciferous Sand-rock of Eaton, and the underlying Potsdam sandstone were, by Emmons, de- clared to be represented, to the eastward, by the great development of strata included in the Sparry Lime-rock and the First Graywacke, to which, as a whole, he gave the name of Taconic slates, and later that of Upper Taconic. He farther declared, in 1860, that the Primor- dial zone in Bohemia, which includes Barrande's first fauna, " is in co-ordination with the upper series of the Taconic rocks." * ♦ Emmons, Manual of Geology, p. 89. i 'J I i t .1 it m 'A ill p n 14. Il i' Il ■i i ii' il il I mi ii 1- '"' "p p! I t : iLin.ii 623 THE TACONIO QUESTION IN OEOLOOT. [XI. Tlio name of Ordovician (sometimes contracted to Or- doviun) which we have introduced in this table, was pro- posed by Lapworth, in 1879,* to designate the group of paleozoic rucks found in Wales between the base of the Lower Lhind(jvery and the base of the Lower Arenig. Tliese, conesponding essentially to the Upper Cambrii'ii or Bala groi'p of Sedgwick, — the second fauna of E..r- runde, — were, as is well known, by a mistake in strati- grapliy, joined by Murchison to his Silurian system, under the name of Lower Silucian; and have also since been called Siluro-Carabrian and Cambro-Silurian. By making of this debated ground a separate region between the true Silurian above and the great Cambrian series below (the Middle and Lower Cambrian of Sedgwick), Lapworth has squght to get rid of the confusion in nomenclature, and to restrain the attempts of some to extend the name of Silurian downwards even to the base of the Cambrian itself. This new division is convenient in American geology from the fact that it includes the group of strata between the base of the Silurian (Oneida) sandstone and the base of the Chazj limestone ; the latter, together with the Trenton, Utica, and Loraine divisions, being equiva- lent to the Ordovician. The name was given in allusion to the Ordovices, an ancient British tribe inhabiting North Wales. [Hicks, to whom we owe so much of our knowledge of the paleozoic rocks of Great Britain, has recently proposed to extend the term Cambrian above the limits assigned by Sedgwick, and to regard it as including three divisions, Silurian, Ordovician, and Georgian ; the latter name, for V e middle and lower divisions of the original Cambrian, being derived from the St. George's Channel, along which its principal groups in Wales are displayed.!] The question of the relations of the great Keweenian * Geological Magazine, vi., p. 13. + Geol. Magazih- 3885, iil., 359. For a detailed account of the f!am- brian and Silurian question, see Cbem. and Geo^ Essays, pp. 349-386. '■ t .' ^ XI.] GEOLOGICAL SURVEY OF NEW YOIIK. 629 eat Keweeman ;count of the 0am- ,ys, pp. a49-386. o mi 'A H W 5 Q M IS >> h-t o CO O o o CO '«k: if ! ■ \W '\\ "'-'^-'^ ?;l n^-'ia nilki i'-fc''avSi3ii^iA4..;ii:i:ftfei>w.wi^^ 630 THE TACONIC QUESTION IN GEOLOGY. [XI. series (unknown to Eaton), already noticed on page 415, which lies at the base of the Cambrian, and above the Ta- conian, will be discussed at length farther on. § 18. We have placed at the base of the column, as representing the gneisses and other crystalline rocks (Eaton's Lower division of the Primitive series), the names of five groups: Laurentian, Norian, Arvonian, Huronian, and Montalban, the distinctness of which, in our opinion, is now established, alike on stratigraphical and lithological grounds, both in North America and in Europe. The following words, published in 1874, before the recognition of the Arvonian, are still applicable : "The dis- tribution of the crystalline rocks of the Norian, Huronian, and Montalban series would seem to show that these are remaining portions of great distinct and unconformable series, once widely spread out over a more ancient floor of granitic gneiss of Laurentian age ; but that the four series thus indicated include tlie whole of the crystalline stratified rocks of New England is by no means affirmed. How many more such formations may have been laid down over this region, and subsequently swept away, leaving no traces, or only isolated fragments, we may never know; but it is probable that a careful study of the geology of New England and the adjacent British prov- inces may establish the existence of many more than the four series above enumerated." * III. — GEOLOGICAL STUDIES IN PENNSYLVANIA. § 19. The reader's attention is now called to the two districts in Pennsylvania mentioned in § 14 ; where the present writer has been enabled to confirm the observa- tions of Vanuxem. To the west of the Susquehanna, in Mifflin County, is the Kishacoquillas valley, an eroded anticlinal valley, having a rim of Oneida sandstone (the Levant, or No. IV., of Rogers), which is the summit of the Second Graywacke of Eaton, and is conformably * Huut, Chemical and Geological Essays, p. 281. [XI. 11 page 415, )Ove tlie Ta- e column, as . alliiie rocks series), the 11, Arvonian, , of whicli, in tratigrapliical n erica and in 87 4, before the ble: "Thed^s- ian, Huronian, that these are unconformable ancient floor of -, the four series talline stratified iffirmed. How jeen laid down t away, leaving we may never d study of the ant British prov- j more than the 1 ,Ued to the two ^ 14 ; where the tirm the observa- Susquehanna, m lalley, an eroded la sandstone Cti^e is the summit 01 d is conformably ays, p 281. XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 631 overlaid, on both of the monoclinal slopes, by the Medina and Clinton beds. Passing downwards from the massive sandstones of the rim to the centre of the valley, we find alternations of sandstone layers with sandy shales, suc- ceeded, in descending order, by the Utica slate and the Trenton limestone ; all of which are well characterized, both lithologically and paleontologically. The whole series, from the summit of the sandstone to the base of the limestone, here presents apparently one unbroken stratigraphical succession, cor"esponding to that already described as occurring in the central district of New York (§ 13), and to what is seen along the north shore of Lake Ontario, in Canada. A similar condition of things occurs in the Nippenose, the Nittany, and the other so- called coves or limestone valleys, which are found in central Pennsylvr.xiia, and, like that of Kishacoquillas, are eroded anticlinals. Accounts of these will be found in Rogers' "Geology of Pennsylvania," Vol. I., pp. 460-511; and also in Report T., on Blair County, by Franklin Piatt, of the Second geological survey of the State. The latter tells us that " there is no appearance of non-conformability here between III. and IV "; that is to say, between the Loraine shale and the succeeding Oneida- Medina sandstones. Within these valleys, there appears, beneath the fossiliferous limestone, a great mass of mag- nesian limestones, several thousand feet in thickness, abounding in ores of iron and of zinc, ard identical with the limestones of the Appalachian valley. § 20. When we pass from the central region of Penn- sylvania to the east of the North or Kittatinny Mountain, we find, along the western border of the Appalachian valley, the sandstone, No. IV., which constitutes this, monoclinal ridge, resting upon a great series of schistose rocks, declared by Vanuxem to belong to the Frankfort or lower division of his Hudson-River group, in which he included the roofing-slates of the region. The contact between the overlying sandstone and these rocks is not, m ^^^H Ii- ^1 ^I^^^^Hiil& ''' ^H^BI ^^^^H^^kI BFj ^^HEi' i\^. ^HhRji^ JS i' ^^^^^^IHi'l' ui 1 ■Hill j««Mfl^^E|raB iji; ,ST ■ ■ '■ "iiiim!^ ■' *i •■ i' SffiSMSBl' ' !■ '1 \ \ \: ^^^^^^Eki \ i ^^tK^^m' W 9- ^^HJ^B''' }- HHiiy IVImkiR t '^- ' I'^i ' ' ^ '1 ftl^ ijHfetflaM ii >?4€si4iW4«:4*»#aS*Wi*(«w**>/ 532 THE TACONIC QUESTION IN GEOLOGY. [XI. I! however, as in the central valleys, one of conformable passage, through intercalations, into an underlying series of fossiliforous shales, but, as may be observed at the Lehigh Water-Gap, one of non-conformity. H. D. Rogers noted the fact that the conglomerates of tlie sandstone. No. IV., here include "many rounded iiebbles and frag- ments of the three underlying formations which intervene between it and the Primary rocks at the bottom of the series." * He recognized among the pebbles portions of the Primal sandstone, of chert derived from the Auroral limestone, and of the Matinal shites. The presence of all these, which I have verilied, is sufficient to show the complete stratigraphical break which here separates this Silurian sandstone from the subjacent argillites. These latter are seen along the banks of the Lehigh, resting in apparent conformity upon tlie Auroral limestone, — which, with its overlying and interstratified scliists, and its subjacent qui., tzite, makes up the Lower Taconic of Emmons. § 21. As the result of my observations in these two regions of Pennsylviuiia, I stated, in 1878, that the passage in the central valleys "from the Upper Cambrian sVales into the Silurian sandstones is gradual, and that there is no stratigraphical break ; although, as shown by Rogers, such an interruption occurs between these same sandstones and the underlying slates along the northwest border of the great Appalachian valley." f This non-conform- ity has been questioned by Professor Lesley, but my own observations at the Lehigh Water-Gap are confirmed by those published by I. C. White, in 1882, in his Report G. 6, of the Second geological survey of Pennsylvania (pp. 150, 151), and I repeat his statement that "the proof seems conclusive " that the Silurian sandstone, IV., here rests unconformably upon the underlying slates. Of these, we have already spoken as entirely distinct from * Second Annual Report on the Geology of Pennsylvania, 1838, p. 30. t Chein. and Geol. Essays, 2d ed., preface, p. xxi. XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 533 niformable ying series ved at the . D. Rogers sainlstone, ;s and frag- jh intervene ittom of the , portions of the Auroral presence of to siiow the separates this Uites. These 3high, resting limestone, — sd schists, and 'er Taconic of t in these two hat the passage anibrian sVales I tliat there is wn by l^^gers, iame sandstones rthwest border is non-conform- jy, but my own pe confirmed by n his Report G. jnnsvlvania (pp- Bhat'^'^the proof Istonc, IV., here ing slates. ies, in regions where the effects of sub-aerial decay are seen to 'oi^siderable depths, are converted into the so- called brown hematite ores — limonite and turgite — which are found imbedded in soft clayey and generally iiighly inclined strata, the results of the decomposition and partial solution of the limestones and their associated schists. The limonitic ores of this horizon are extensively mined along the outcrop of these Taconic rocks, from Vermont +^0 Alabama; and, as lias been shown by the concordant observations of many investigators, have been derived by eT)igenesis, in some cases from the sulphid, and in other cases from the carbonate of iron ; both of [i 1.. '' ' :r ■ I m' ' 4 ?■ 4 Hi -n', 636 THE TACONIC QUESTION IN GEOLOC /. [». Ul which, in the deeper workings, are found unaltered. Crystals of magnetite are sometiiues disseminated through these schists, as well as thin layers of compact hema- tite, both of which are occasionally fcnnd in tlie clayey beds with the limonites. The massive granular magnetic ores of this horizon in Pennsylvania are generally asso- ciated witli small quantities of pyrite and chalcopyrite, and frequently yield by analysis a little cobalt. They are distinguished from the magnetites of the older rocks by their generally finely granular texture and feebler cohesion, as well as by the characteristic imbedded minerals. The hematite is often a very soft unctuous micaceous variety, and both magnetite and hematite are occasionally found in grains disseminated in the soft granular sandstone layers. The limonites are often man- ganiferous, and are sometimes accompanied by manganese- oxyds, which are doubtless derived from corr*^spo]tding manganesian carbonates.* Associated with the limestones of this series are sometimes considerable interbedded de- posits of zinc-blende, and oxydized ores of zinc are found at the outcrops of these. f § 27. As regards the thickness of the strata which in the central region of Pennsylvania underlie the sand- stone, No. IV., we find in the Kishacoquillas and Nittany * The few carbonated ores from this horizon in Pennsylvania, which have been analyzed, are more or less manganesian ; one of them yielding to McCreath 5.0 per cent of manganesian carbonate. A massive fawn-colored carbonate, with a specific gravity of 3.25, found in layers in the so-called Primordial slates of Placentia Bay in Newfoundland, gave me by analy- sis 81.6 p6r cent of manganesian carbonate, and 15.4 per cent of silica for the most part soluble in a dilute alkaline solution, besides traces of ferrous, calcareous, and magnesian carbonates. It was partially incrusted with black crystalline manganese-oxyd, evidently of epigenic origin. Amer. Jour. Science, 1859, vol. xxviii., p. 374. t The chief facts in the mineralogical history of these rocks will be found in my volume already cited ; Azoic Rocks, etc., pp. 201, 206. See also a description of the Cornwall Iron-.Mine, etc., Proc. Araer. Institute Mining Engineers, vol. IV., pp. '.jI[)-o'2'>, and two notes on the Tacouic System, and on the Genesis of Iron Ores, published in the Canadian Naturalist for December, 1880 ; besides a farther discussion of this sub- ject, pp. 261-268 of the present volume. . "i XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 637 valleys, respectively — for the Loraine shales, a thickness of 1200 and 700 feet, for the Utica slate, 400 and 300 feet, and for the fossiliferous Trenton beds, 400 and 300 feet. Beneath the latter, in these central valleys, lies a great mass of niagnesian limestones interstratitied with schistose beds ; the whole called by Rogers, Auroral, and supposed by him to be, like the similar rocks in the east- ern part of the State, the representatives of the Chazy and Calciferous divisions of the New York system. To this succession of limestones, as observed at Bellefonte, Rogers assigned ii thickness of over 5400 feet, of which the upper 600 are highly fossiliferous ; while the great underlying portion is destitute of fossils, or contJT.>: jit few and un- determined organic forms. The most complete section of the strata below the sandstone. No. IV., in the central region, is that lately measured by Mr. Saunders, in Blair County ; where, beneath 900 feet representing the Loraine and Utica shales, are found not less than 6600 feet of strata, inchiding, at the top, the fossiliferous Trenton beds, whose thickness is not separately given, and, near the base, intercahited sandstones and shales.* A sum- mary of this section gives, in descending order : — Feet Sandstone, No. IV — Upper shales (Utica and Loraine) 900 Limestones and dolomites, including tlie fossiliferous Trenton 5400 White sandstone 40 Limestone with sandstone and shales 1100 7500 § 28. In none of the sections of these rocks exposed in the eroded anticlinal valleys of the west, has anything been found corresponding to the older crystalline groups which, along the border of the southeastern region, under- lie the base of this series. For the rest, these lower non- fossiliferous strata present similar mineralogical characters * Second geological survey of Pennsylvania, Report T, by Franklin Pratt, pp. 18, 48-69. ;:,■! a-.^jj.fiw. J_v Mfer tm-^Jm *•«- 538 THE TACONIC QUESTION IN GEOLOGY. [XL to those of the great valley to the southeast, and include extensive deposits of limonite (imbedded in clays, which are decayed schists in aitu)^ as well as ores of zinc ; both of which are largely mined in Blair County. § 29. If we turn from these central valleys of Penn- sylvania to what Rogers culled the southeastern area, that is to say, the regions lying to the southeast of the Kittatinny Mountain, we fiuf^ a very different condition of things. In place of the 9'. j feet of fossiliferous shales measured in Blair County between the limestone below and the overlying sandstone, we find not less than 6000 feet of unfossiliferous strata. As long since measured by Rogers, on the west side of tlie Delaware River, at the Water-Gap, there are 6102 feet between the base of the sandstone, No. IV., and the underlying Auroral limestone.* Mr. Chance, in a later section in this vicinity, makes them above 3900 feet, and Lesley concludes from observations on the Sus- quehanna that they have an aggregate thickness of not less than 6000 feet, which agrees with the early measure- ments of Rogers. The characters of this great group of strata in tlie Kittatinny valley, included Loth by Rogers and by the second geological survey in the Matinal divis- ion, are exceedingly variable, and they present important local differences. The roofing-slates already mentioned (§ 20) are confined to a small area in the northwest part of this valley, occupying a narrow zone lying from one to three miles south from the base of the Kittatinny Moun- tain, and extending from a point in New Jersey a few miles east of the Delaware Water-Gap, across the Dela- ware and Lehigh, and a few miles west of the latter river. These roofing-slates were assigned by Rogers to the lower part of the group in question. According to Chance also, who has lately examined them, they are very low in the series, and of no great thickness ; but are affected by such sharp flexures that the dips on both sides of the anticlinals and synclinals are nearly parallel, so that the apparent * Second Annual Repci\ 1838, p. 35. ^d include lys, which zinc; both 3 of Penn- tstern area, .east of the ^t condition jrons shales le below and 6000 feet of idbyl^Jge^S' 5 Water-Gap, Indstone, No. Mr. Chance, a above 8900 5 on the Sus- ckness of not jarly measure- ^veat group of loth by Rogers :Matinal divis- sent important ady mentioned northwest part ng from one to ttatinny Moun- V Jersey a few .cross the Dela- the latter river, ers to the lower r to Chance also, very low in the affected by such oftheanticlinals Lat the apparent 35. XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 639 thickness of the roofing-slates is much augmented.* In the region to the west of the Lehigh, in thu counties of Berks and Lebanon, these Matinal sUites include a great amount of coarse arenaceous rock, and rise into bold hills. Some parts consist of heavy gray sandstones with conglomerates, and bluish or grayish shales with thin- bedded limestones. Large portions are characterized by a predominant reddish or reddish-brown color, with inter- stratified beds of yellow or fawn-colored shales, and are said by Roger;; to resemble the strata of the Medina and Clinton, above No. IV. Mention should also here be made of the existence of considerable masses of conglomerate made up of more or less completely worn pebbles of the Auroral limestone in a calcareous cement, which are found at several points in the great valley, and have been described by Rogers as resting upon the Auroral limestone.f § 30. From my observations in this region, in 1875, when I had an opportunity of seeing the rocks of this group at several points in the Appalachian valley between the Lehigh and Schuylkill Rivers, I was struck with their great resemblance to the First Graywacke of Eaton (the Upper Taconic of Emmons, or Quebec group), as seen from the banks of the St. Lawrence at Quebec, to the valley of the Hudson ; which, it will be remembered, was, by Mather, confounded with the Second Graywacke (§ 12). It is apparent from a section to be seen a little west of the Lehigh, below Slatington, that the coarse red and gray sandstones, with red shales and conglomerates, overlie the roofing-slates of the valley; and their geo- graphical relations are such as to suggest an unconform- able superposition. § 31. Regarding the rocks of this valley, I expressed, ' in 1878, my belief " that besides the Auroral limestones, * Rogers, Geology of Pennsylvania, I., 247 ; also, Second Geol. Survey Penn., Report G 6 ; pp. 340, 363. t Geology of Pennsylvania, I., 252. m I ) h I ' T< L V 640 THE TACONIC QUESTION IN GEOLOGY. Pit. with their succeeding argillites, and the unconformably superimposed (Oneida) Silurian uonglon'.erates of tlie North Mountain, there are, to the west of the Lehigh Uiver, portions of two intermediate -.rniations. One ol these, marked by red-colored sandstones, conglomerates, and shites, appears to be the same with the Upper Taconic or Cambrian belt; which has been traced by II. D. Rogers, Mather, Emmons, Logan and the writer, with some inter- rui)tion8, from New Jersey to Canada, aUnig the great Appalachian valley. The other is an impure black earthy limestone, becoming in parts a soft, thinly bedded flag- stone, which was seen lying, at moderate angles, above the blue limestone of the valley, not far from Copley, and was the supposed to belong to a different series. It is apparently the same with the Trenton beds recognized by Professor Prime in that vicinity," * as mentioned below (§34). § 32. We have noted the evidence of a stratigraphi- cal break between the slates of the great valley and the overlying Levant (Oneida) sandstone in the Kittatinny Mountain, and have shown that the conglomerates of the latter include numerous pebbles derived alike from the underlying Primal, Auroral, and Matinal rocks. If now we turn to the central valleys we find, as already stated, no evidence of any stratigraphical break ; but, on the con- trary, a passage downwards from the Oneida sandstone to the underlying Loraine and Utica slates. We still, how- ever, find in these sandstones similar conglomerates to those of the Kittatinny range. This is well seen in Jack's Mountain, on the eastern border of the Kishacoquillas valley, where the Levant division is described by Rogers as consisting in its lower part of four hundred feet of sand- stone ; of which he says, it contains " numerous pebbles of white quartz, of Matinal slate, and of the harder Primal strata, and is really a conglomerate." The upper member of the Levant, which is still thicker, is also a conglomer- * Hunt, Azoic Hocks, p. 215. iforinably 58 of the te Lehigli , One ol lomerates, 31- Tacouic D. Rogers, ome iuter- tlie gveat iack earthy ecUled Hag- gles, above Oopley, and lies. It IS cognized by ioned below XI.] GEOLOGICAL STUDIE8 IN PENNSYLVANIA. r)41 ate, holding in parts quartz pebbles, in addition to which, "flat liuni)8 and pebbles of red shale occur througliout the whole mass."* Tiie jiebbles of these conglomerates, which I have examined in situ, have evidently nothing in common with the fossiliferous strata below tiiem, but are derived from older rocks, like those of the Kittatinny valley, and include large (juaiitities of the characteristic red shales which we have already noticed. § 33. The thickness of the Auroral limestones in the great valley is less than farther west, being, according to Chance, about 3000 feet ow the Sus(iuehanna ; while at Bethlehem and AUentown, in Lehigh County, they measure about '2000 feet, according to Prime, who thinks their maximum thickness there may be 2500 feet. These Auroral limest;nies, with their immediately associated schists and limonites, have been carefull}' studied by Prime in the count}^ just named, and are described by him in Reports D and D 2, of the second geological survey of Pennsylvania. Schistose layers, with limonite, are there occasionally intercalated in the limestone, but the principal bodies of clay, or decayed schist, holding this ore, are, according to this observer, found at two horizons, the one near the summit and the other at the base of the limestone, between it and the underlying quartzite ; which, also, includes in this region, schistose bands with hydrous micas, limonite, and occasional layers of red hematite. § 34. These Auroral limestones and shales were, as we have seen, supposed by Rogers to be the equivalents of the New York series from the base of the Calciferous to the summit of the Birdseye and Black-R'ver divisions; the Trenton limestone proper being, according to him, represented in the eastern area only by some beds of argillaceous limestone, uhich were by Rogers included in the Matinal division of his classification. According to Prime, "the Trenton or fossiliferous limestone seems to • Rogers, Geol. of Peiin., Vol. I., p. 473. ■( I ^Hc .s^, ^. ^, ^. IMAGE EVALUATION TEST TARGET (MT-3) A % 1.0 I.I 11.25 Ik ■ 50 m m IS 1.4 m us ■"IS 2.0 1.6 ^ ^ /: f ^^J ^ .^ ^> /A % %* '-'w 7 Photographic Sciences Corporation 23 WBST MAIN STREET WEBSTER, N.Y. 14580 (716) •72-4503 \ •^ >^ t/j 542 THE TACONIC QUESTION IN GEOLOGY. [XL occur only at a few points in the valley," having been recognized by its fossils at one locality only. It is here dark-colored, earthy, and uncrystalline, and associated with argillaceous beds which yield a hydraulic cement. These, which are supposed to belong to the same horizon, are found at several other places in the region, overlying the magnesian limestone. Prime also mentions one local- ity where forms referred to Euomphalus and Maclurea are met with, indicating the horizon of the Chazy ; while in another an undesoribed Lingula occurs. Peculiar funnel- shaped markings, not very unlike the Scolithus of the underlying quartzites, have also been found in the magne- sian limestoue in one place, and have been referred to the genus Monocraterion, which occurs in the Eophyton sand- stone of Sweden. For farther notice of these organic form3 see the author's volume on " Azoic Rocks," p. 206, and also Professor Prime's Report D 2. § 35. The Primal division of the series under con- sideration is, in the northeast part of the great valley, in Pennsylvania (where it rests unconformably upon the Laurentian gneiss), a thin and irregular deposit, and, ac- cording to Rogers, is sometimes wanting ; in which case the Auroral limestone reposes directly upon the gneiss, as may be seen in Lehigh and Northampton Counties. In the North- Valley Hill, in Chester County, and farther to the northeast, in Lehigh County, the Primal quartzite, often with Scolithus, is seen to rest, with a thickness of from twenty to fifty feet, directly upon the Laurentian gneiss. These basal beds in Chester County include some micaceous and schistose layers, and are followed by the Upper Primal slates and the Auroral limestones. The rock is sometimes granular, and often detrital, while at other times it is a hard granular or even flinty quartzite. Farther to the soutliwest, in Berks County, the Primal quartzite becomes more continuous and thicker, rising in- to high ridges. § 36. The conditions above noticed show the deposition XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 643 the deposition of these rocks over an uneven subsiding gneissic area, and a conformable overlapping of the Primal beds by the suc- ceeding Auroral limestone. As described by the writer in 1876, " they were evidently deposited over a subsiding continent, with bold shores ; so that while the Primal has in places a great thickness, it is elsewhere very thin, or entirely wanting beneath the Auroral, which rests directly upon the older crystalline rocks." * The characters of the Primal are best seen farther to the west, where, in the broader part of the basin, it is brought up by undulations from beneath the Auroral, and appears as a complex group of considerable thickness, v."ith alternations of quartzites, argillites, and crystalline schists, beds of iron- ores, and intercalated limestone-layers ; the latter consti- tuting, as well described by Rogers, beds of passage into the overlying Auroral limestone. Rogers defined the group as a Primal sandstone, with slates above and below ; but it is occasionally less simple, since what he called the Upper Primal slates may include interstrati- fied sandstone-beds, sometimes of considerable thickness. Thus, in a section near Parkesburg, on the North- Valley Hill, described by him, a mass of 200 feet of yellow sand- stone is found, with 300 feet of slates above, and 350 feet more below, lying between this upper sandstone and the white Scolithus-sandstone beneath, which here measures fifty feet ; the section being as follows, numbered in de- scending order : — Feet. 0. Auroral limestones — 1. Upper Primal slates with sandstone layers .... 300 2. Yellow sandstone 200 3. Laminated slaty beds 360 4. Middle Pr-mal sandstone, with Scolithus .... 50 5. Lower Primal slates 300-400 A section at Chickis, on the Susquehanna, also described by Rogers, gives a still greater thickness of strata referred * Harpers' Annual Record for 1876, p. xcvl. .,M» to ..eP.n..;t.eW,^o.t.» se.es not ..„. exposed. Wehave, as before— ^ ^^ 1800 1. Upper Primal slates. •••••.•.''*...' ^7 2. White sandstone dw : :i'S«.n;^*Womh^: : : we Shan notice fan^-J^e-ha^^^^^^^ Primal slates as seen el'«™l'»™ " f„ Pennsylvania and .n other Sta^s ^.^ ^ j R S7. Since the time 7^«" ''Marches in various parts JestigationsinPen„sj« e-ai^^^ ^^^^^ of the Atlantic bf '' ^"^j'^^h were known to h.m, between the ancient gneisses w ^^^^ .^^ and the base of the I'^'^l^'l^ talline stratified localities, one or more f J^J Huronian, and of the rocks. 0£ these, Pf'^''^"? i*t^hich have been called younger gneisses and ■»''=.»-^*'^*JJogical resemblances to kntelban, present certain mineralog ^ ^^^^^ ^^ ^^^^^ the schists o£ the Lower Pnmal^^^^ ^^^ ^^^^^^^^ interstiatified with '^"'i ov'^f jf,, aUtinctly crystalline was stated by Rogers, more «^ „„,,, Rogers con- Sd thrc^s^^rf.- -rth^r s:- Sa^lotLrrgralnameof-semi-met. morphic schists." separates the Scolithus- I 38. No stratigraph.calbreak sop ^^^^ ^^^^ ^^^ Jdstono from the ^'J^^.'^Xmorphic schists below ,vhole of these r"" lltone great group. Th s was this horizon were included '» Jfation of the Paleozoic described as a d'"'r'''li'oe of organic remains, was series; but, from f*hs and from the more ancient distinguUhed alike ft^J^'^;"™ jj^e name of the A.ou> - or so^alled Hypozoie g»;~^Kogers, among the rocks series. There ^f.-^'Ja^ physLl break, or horizon of the Atlantic belt, but one p j XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 545 i not being Feet. 1800 27 300 )f the Lower ribution, both his geological I various parts e shown that tnown to him, are, in many aiine stratified an, and oi the ave been called lesemblances to las well as those Auroral, are, as nctly crystalline 3S, Rogers con- with portions ot ,e older gneiss: of "semi-meta- tes the Scolithus- ,es, and thus the )hic schists below group. This was of the Paleozoic anic remains, was the more ancient ,ame of the Azoic s, among the rocks break, or horizon of unconformity, throughout the immense succession of altered crystalline sedimentary strata," namely, that at the summit of the ancient or Hypozoic gneiss ; and " one paleontological horizon — that, namely, of the already discovered dawn of life among the American strata. This latter plane or limit, marking the transition from the non-fossiliferous or Azoic deposits to those containing organic remains, lies within the middle of the Primal series of the Pennsylvania survey ; that is to say, in the Primal white sandstone, which, even where very vitreous, and abounding in crystalline mineral aggi-egations, con- tains its distinctive fossil, the Scolithus linearis." § 39. In the opinion of Rogers, the whole series of strata below the Levant sandstone had, in the southeast- ern area, been the subject of alterations, which had given to them the characters of crystalline rocks. I have else- where set forth at some length the views of Rogers on this point, and have shown that his conclusions with re- gard to the so-called Azoic rocks were not clearly defined, and that, in his opinion, it was often difficult, if not im- possible, to distinguish between the upper portions of the Hypozoic and certain parts of the Azoic series. It would appear from his descriptions, and from my own examina- tions in the region, that portions of Huronian, and of Montalban, were by him included in the Hypozoic ; and other portions of the same or of older rocks, in the Azoic^ or even in the Upper Primal slates. Both these, and the Primal quartzite itself, were by Rogers supposed to have been changed into feldspathic rocks ; and he has described as alt-^red Upper Primal, a great group of such rocks seen in the South Mountain to the south of tlie Susque- hanna, which we shall proceed to notice. § 40. Leaving the Mesozoic red sandstones at Gettys- burg, and passing westward over the South Mountain, by Caledonia Spring to Chambersbuig, we meet first with a belt, more than two miles wide, of crystalline rocks, regarded by Rogers as in part Upper and in part Lower (|P3^* m I fipiiil PI 'I iHil I i 111 « i 546 THE TACOKIC QUESTION IN GEOLOGY. [XI. Primal slates ; the latter represented by talcose, chloritic, and epidotic schists, with diorites, and the former by what were called by Rogers, "jaspery rocks," and "reddish jaspery slates." These, which I first saw with Dr. Per- sifor Frazer, in 1875, were found to consist of petrosilex or compact orthofelsite, often becoming porphyritic from the presence of crystals of feldspar or of quartz. I then compared tliem with the similar rocks found along the coasts of Massaclmsetts and New Brunswick, and on Lake Superior, all of which I at iuat time included in the lower part of the Huronian, but have since been led to regard as an independent series, identical with the Arvo- nian of Kicks ; which, in Wales, appears to be interposed unconformably between the Laurentian (Dimetian) below, and the Huronian (Pebidian) above. § 41. To this series also belongs a great thickness of petrosilex-rocks, often porphyritic, and associated with small portions of soft, unctuous micaceous schists, occur- ring in central Wisconsin, where they overlie conformably a great mass of vitreous quartzites, which, from the inter- calation of similar micaceous layers, apparently belong to the same series with the petrosilex. These rocks, origi- nally described by Percival as altered Potsdam sandstone, were by James Hall, in 1862, referred to the Huronian, with which they are also classed by Irving, who has since described them.* I have recently examined these rocks, in situ, as seen on the Baraboo River in Wisconsin, and have found them indistinguishable from the petrosilex- beds of Pennsylvania and of our Atlantic coast, and from the typical Arvonian of Wales. § 42. These petrosilicious strata, presenting many va- rieties in color and in texture, have a great thickness in the South Mountain, west of Gettysburg, where they generally dip southeastward at high angles. With them are seen in some parts, apparently interstratified, thin * See Geology of Wisconsin, 1877, vol. ii., pp. 501-521; also Hunt, Azoic Rocks, p. 232. XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 547 5, chloritic, ler by what \ "reddish bh Dr. Per- f petrosilex liyritic from ,rtz. I then d along the ick, and on jluded in the 3 been led to ith the Arvo- be interposed letian) below, ; thickness of isociated with schists, occur- te conformably from the inter- jntly belong to •se rocks, origi- dam sandstone, the Huronian, , who has since I'ed these rocks, Wisconsin, and the petrosilex- coast, and from anting many va- at thickness in ■irg, where they es. With them ,er'stratified, thin 601-621; also Hunt, bands of argillite, with chloritic and epidotic rocks, such as I have found with the similar petrosilicious rocks on Passamaquoddy Bay, on the Atlantic coast. This crys- talline series is, to the westward, un conformably overlaid by a belt, about a mile and a half wide, of sandstone, with conglomerates, generally with a northwestern dip, consti- tuting what is known as Green Ridge. This is followed by a repetition of the petrosilicious rocks, again with high southeast dips, and by a great mass of chloritic and epi- dotic strata, overlaid to the westward, as before, by a considerable thickness of Primal sandstone, which dips in that direction beneath the Primal slates and Auroral lime- stones of the Appalachian valley. § 43, In this remarkable section, it is evident that the crystalline rocks, upon which the Primal quartzite rests unconformably, belong to one or more older series, dis- tinct from the Laurentian, and representing both the Huronian and the petrosilex or Arvonian series. I was thereby confirmed in my opinion, expressed in 1871, that the crystalline schists regarded by Rogers in this region as altered Lower Primal and Upper Primal, are both of them older than the Primal quartzites, and belong to one or more distinct series. These conclusions were an- nounced in the Proceedings of the American Association for the Advancement of Science for 1876 (pp. 211, 212), and also in Azoic Rocks (pp. 18 and 193). Frazer, who has since devoted much time to the study of the region, agrees with me in placing the crystalline rocks of the above section in the Huronian, including under that name the accompanying petrosilex group; and regards the quartzites as there forming the basal member of the Primal series.* § 44. From the observations given in § 36, it is appar- ent that the Primal series of Rogers, where most largely developed in Pennsylvania, includes several repetitions of quartz-rocks, sometimes vitreous, sometimes granular, and * Thfese pr^sent^e k la faculty des sciencea de Lille, etc., 1882^ 648 THE TACONIC QUESTION IN GEOLOGY. 00. occasionally detrital and conglomerate in character, alter- nating with softer schistose strata. This will be farther illustrated in a succeeding chapter, by observations in Virginia and elsewhpre; when it will also appear that repetitions of these quartzites are met with below the horizon of the Scolithus-sandstone. In many cases, a quartzite, often a conglomerate, is found to constitute the basal member of the series, which rests unconformably upon different groups of the older crystalline rocks,— Laurentian, Arvonian, Huronian, or Montalban. Inas- much as portions of the latter two groups were by Rogers confounded with the Lower Primal slates, it will require careful examination, in each case, to determine whether we have really to do with the older rocks, or with strata belonging to his Primal series. Notwithstanding the division of the latter into Azoic and Paleozoic, based by Rogers upon the appearance, in the midst of the Primal, of the Scolithus-sandstone, it is to be remarked that the Primal slates, both above and below this horizon, really constitute, with the rocks of the Auroral, and a portion of the Matinal in the south- eastern area of Pennsylvania, one great continuous series. Similar schistose and micaceous layers are found inter- calated alike among the Primal quartzites and the Auro- ral limestones; while the accompanying masses of slate often include minor beds of quartzite, and others of gran- ular limestone. The intimate relations of these various rocks were noticed by Rogers, who mentions what he calls " the alternations of Primal slate and Auroral lime- stone," and " the limestone at the passage of the Primal into the Auroral." The Lower Primal slates were else- where described by him as alternations of " talcoid sili- cious slate, talco-micaceous slate, and quartzose micaceous rocks," usually schistose, besides otl !r strata which are nearly pure clay-slate. Portions of the Matinal, in like manner, were said by him to be " a semi-crystalline cluy- slate* partially talcose or micaceous." Later studies XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 549 actev, alter- 1 be farther jvvations in appear that 1 below the any cases, a onstitute the iconformably ine rocks, — alban. Inas- ive by Rogers Lt will require mine whether or witli strata iter into Azoic appearance, in sandstone, it is oth above and I the rocks of ,1 in the south- ntinuous series. re found inter- 5 and the Auro- masses of slate others of gran- of these varioiis utions what he id Auroral lime- re of the Primal slates were else- of "talcoid sili- rtzose micaceous strata which are Matinal, in like L-crystalline clay- Later studies have shown these strata to abound in hydi'ous micas, and more rarely to contain talc, chlorite, and related species. Some beds in the Primal slates are apparently feldspathic in composition, since they are changed by sub-aerial decay info clays resembli»ig kaolin. § 45. The continuous belt of Primal and Auroral rocks stretching along the southeast base of the North or Kittatinny Mountain, is bounded on the south, in its extension between the Delaware and Schuylkill Rivers, by the so-called South Mountain. Beyond the Schuylkill, at Reading, this Laurentian range is, so far as known, represented only by one small mass, a little west of the town. Its disappearance at the Schuylkill, to rise again south of the mesozoic belt, in the northern part of Chester County, permits a great extension of breadth of the Pri- mal and Auroral rocks to the southward, in the counties of Chester, Lancaster, and York, where they appear, both to the north and the south, from beneath the broad and somewhat irregular belt of mesozoic sandstone which, from the Delaware to the Susquehanna, crosses the State in an east and west direction. From the Susquehanna to the line of Maryland, however, the trend of this belt is to the southwest. The Primal and Auroral strata, along the south and east of the mesozoic, occupy the limestone- valleys of Lancaster and York Counties, with which the narrow limestone valley of Chester County, lying to the eastward, is, as Frazer has shown, continuous. § 46. The South Mountain, which, as we have seen, is effaced between the Schuylkill and the Susquehanna, re- appears to the southwest of this rive' , in the broad ridge of crystalline rocks, already described in § 40 as found in Adams County, between the continuous limestone-valley on the northwest, and the mesozoic on the southeast. In tliis ridge of Huronian and Arvonian rocks, the Lau- rentian has not yet been recognized. It, however, as already remarked, appears in Chester County, betAveen the mesozoic and the Chester limestone-valley. In addi- i; tj' M I ooO THE TACONIO QUESTION IN GEOLOGY. [XI. tiun to tliis, I pointed out, in 1876, the existence of a subordinate Laurentian axis, south of the limtstone- valley just named, crossing the Schuylkill in Buck Ridge, near Consholiocken.* This ridge bears upon both flanks the Montalban gneisses and mica-schists; while between these and the Laurentian, on the south side of the axis, there is seen on the river an intermediate mass of hornblendic and chloritic schists, with serpentine, ensta- tite, and steatite, which may be an intervening outcrop of Iluronian. § 47. Returning now to tlie Primal and Auroral rocks, the distribution of which has been defined, we remark that it is chiefly along the border of the mesozoic belt that the Primal schists, with their accompanying crystal- line iron-ores, already noticed (§ 26), are best exposed. Examples of these ores are seen at Boyerstown, and near Reading, at Wheatland, Cornwall, and Dillsburg, on the north side, and at the Warwick and Jones mines, on the south side, of the mesozoic sandstone. Rogers, in his third annual report on the geology of Pennsylvania, in 1839, referred these iron-ores to the mesozoic or "middle secondary red sandstone " series, giving, as examples, besides the mines just mentioned on the south side of this belt, the Cornwall mine on the north side. In his final report, in 1858, however, he referred these crystalline iron-ores and their enclosing schists to the Upper Primal slates. He regarded the iron as an original constituent of the sediments, but supposed it to have been re-arranged "by some agency connected with the metamorphism of the strata." Lesley, in 1859, in his " Iron Manufacturers' Guide," described these same ores under the head of " Primary," with those of the gneisses and pre-paleozoic crystalline rocks; at the same time referring with ap- proval to those who regard these ores "as of middle secondary, and not of primary age." Subsequently, in the same volume, he noticed the later view of Rogers * Azoic Bocks, page 200. QY. >^ xistence of a [\e limtstone- t Buck Uitlge, n both flanks vhile between e of the axis, diate mass uf peutine, ensta- ling outcrop of Auroral rocks, led, we remark 3 inesozoic belt panying crystal- a best exposed, stown, and near )illsburg, on the [ones mines, on Rogers, in bis Pennsylvania, in ozoic or "middle tg, as examples, south side of this ;ide. In his final these crystalline the Upper Primal ".ginal constituent 3 been re-arranged metamorphism of on Manufacturers' nder the head of t and pre-paleozoic referring with ap- :es "as of middle Subsequently, m er view of Rogers XI.] GEOLOGICAL STUDIES IN PENNSYLVANIA. 651 already stated, and apparently accepted it, at least for the ores of Warwick, of Cornwall, and of Chestnut Hill, where nuigiietite is closely associated, in adjacent strata, with limunite. Frazer, however, in 187G, still niaiiitaiiiod the early view of Rogers for the ores of Dillshurg, iu Adams County, which he describes as included in the mesozoiu series,* and they are so classed in MoCreath's Report M 3 (preface, page x) of the Second geological survey, in 1881. § 48. From my own somewhat extended studies of all the localities known along the two borders of the meso- zoic belt in Pennsylvania, I am constrained to maintain the opinion expressed by me in 1875, mat the ore-beds near Dillsburg form no exception, but that these, with the deposits of ore at Cornwall, at Wheatland, in the vicinity of Reading, and at Boyerstown, on the north, as well as those of the Warwick and Jones mines, on the south, all belong to the same ancient horizon. That they are met with chiefly along the borders of the mesozoic- saudstone belt, as I then said, " is due to the fact that these ancient ore-bearing rocks, from their decayed con- dition and their inferior hardness, have been removed by denudation, except where protected by the proximity of the newer sandstones, or by eruptive rocks, as is the case at the Cornwall mine." f There, as I have i)ointed out, the dikes from the neighboring raesozoic area have served as barriers, and have preserved from erosion a great mass of magnetic iron-ore. § 19. The stratigraphical relations of these ore-bearing rocks serve to show that they must be referred to the Primal schists which underlie the mesozoic sandstones.. These latter, which are generally regaixled as of triassic age, form a continuous belt from the banks of the Hudson southwestward across New Jersey and Pennsylvania into • Second Geological Survey of Penn., Report C, page 71; and Trans. Amer. Inst. Mining Engineers, v., 133. t Ibid., iv,, 320. 662 THE TACONIO QUESTION IN OEOLOOT. OD. Virginia. Throughout this region, as is well known, theso newer rocks have everywhere a moderate and very uniform dip to the northwestward, of from ten to thirty degiee.s, and were deposited upon the worn surfaces of the previously folded Primal and Auroral rocks, which have contributed largely to the materials of the mesozoic. These older strata,, unlike the latter, present everywhere considerable undulations, with dips, sometimes at high angles, alike to the northwest and the southeast. The unconformably overlying mesozoic rocks, now themselves affected by a gentle and pretty uniform inclination to the northwest, agree nearly with the older rocks in strike; and the coincidence which thus appears between the mesozoic and the northward-dipping outcrops oi the older rocks readily explains how the two have been con- founded. § 50. In the vicinity of Dillsburg, where numerous openings for iron-ore have been made, the dip of the enclosing strata, so far as observed, is to the northwest. The same condition is seen at Wheatfield, to the east of Cornwall, where se/eral lenticular masses of magnetite have been mined ; but at Fritztown, less than half a mile to the southward, the similar ore-bearing strata dip to the southeast. Again, at the Roudenbusch mine,, near Read- ing, is a bed of magnetite which had, in 1875, been mined for a distance of 480 feet down the slope of the bed ; the dip being thirty degrees in a direction S. 30° E. At the Island mine, also near Reading, is a similar opening for ore, which had been followed 240 feet on the incline, with a dip of forty-five degrees to the southeast ; while immediately to the north of this opening is a slope with a still steeper dip to the northwest, on what appears to be the same ore-bed ; indicating the presence of an anticlinal in the ore-bearing strata. At Boyerstown, still farther east, where the mesozoic lies along the southeast flank of the South Mountain, there is opened, at its margin, a mine in which the ore-stratum had, in 1875, been followed ,Y. ^^' veil known, ate auvliich the whole series dips, and beneath which it seems to pass. The strike of the Pvimal rocks is N. 60° E., while that of the eozoi. is N. 30° E. The lower portions of the Taconic series are here some- times concealed by faults. § 67. It is hardly necessary to repeat that the great Lower Taconic belt, as above defined, includes the Primal and the Auroral, together with a portion of the Matinal of Rogers, in Pennsylvania, where some localities were examined and described by Emmons. In our account of these rocks, in that State, in the last chapter, we called attention to the thinning-out of the slates and quartzites in some loc : 'ities along tlie borders of the deposit, and XI.] LOWER TACONIC ROCKS. [XI. lore than •ough the lid is well ough Vir- ibama, A the valley y, will be Geology," : Berkshire Ilia given in ay compare Bine Ridge, cation from tz-rock with m the gveat I slates with L700 ieet of ,eO. strata of ing a hydrous In the lower beds holding of feldspa^^iic bed as a con- ozoic locks of eries dips, and of the Primal fi. is N. 30° E. are here some- that the great ides the Pri-nal of the Matmal localities were our account of apter, we called 1 and quartzites the deposit, and 667 even their concealment beneath the conformably overlap- ping limestone. We also noticed the appearance of these rocks of the Primal division, elsewhere, from beneath the limestones, with a volume not less than that measured by Emmons and Fontaine. The great thickness assigned to the limestones of the series in Pennsylvania is to be noted; and also the consideration that some of this ap- parent thickness of several thousand feet may possibly be due to repetitions. We have also remarked the fact that these Lower Taconic or Auroral limestones are brought up by undulations from Loneath the overlying rocks in the central valleys or coves of Pennsylvania. The same condition of things is met with in Alabama, where the Auroral limestones or marbles, with their underlying slates and quartzites, abounding in limonite, as shown by Eugene A. Smith, are exposed on the great axis which divides the coal-basins of the Black Warrior and the Cahaba; and are also brought to view by a dislocation and uplift along the southeastern edge of the latter basin.* § 68. Lying to the eastward of the Lower Taconic belt of the Appalachian valley, and generally divided from it by the range of aiicient crystalline rocks to which belong the South Mountain and the Blue Ridge, there are other areas of Lower Taconic strata found, at intervals, from Georgia to New Brunswick, often appearing as parallel interrupted belts. These are the remains of a mantle of these rocks once widely spread over the older crystalline strata of the Atlantic slope, from which, after folding and faulting, they have been in great part removed by erosion. § 59. One of these Taconic areas was, as long ago, as 1817, defined and mapped by Maclure, who described it as " a Transition belt " extending from the Delaware to the Yadkin in North Carolina, having a breadth of from * See Hunt on Coal and Iron in Alabama; Trans. Amer. Inst. Mining Engineers, February, 1883. 558 THE TACONIC QUESTION IN GEOLOGY. [XI. St ■; -lu. two to fifteen miles, and a general dip to the southeast. He pointed out its course from the Delaware, passing by Norristown, Lancaster, York, and Hanover, in Pennsylva- nia, and Frederickstown, in Maryland, through Virginia ; noted its passage beneath the mesozoic red sandstone, and its termination in Pilot Mountain, in Surry County, North Carolina. The rocks composing "this belt were described by Maclure as consisting of granular quartzite, granular limestone or marble, and various slates.* Through the Lancaster valley, as already noticed (§ 45), the Taconic rocks of this eastern belt are connected with those of the Appalachian valley. Maclure also described another area of the same rocks found on the north branch of the Catawba, at the base of the Linville Mountains, in Mc- Dowell County, North Carolina. § 60. Emmons, who had examined this belt near the Schuylkill River, in Pennsylvania, was also acquainted with its extension into North Carolina, and in his report on the geology of that State, in 1856, mentions it as one of the five areas of Taconic rocks known within its bor- ders, which are described in that report. Professor Kerr, who has since studied still farther the distribution of these rocks in that State, has delineated them on the geological map accompanying his report of 1876 These rocks pre- sent, according to him, " five principal outcrops, with two or three subordinate' ones," which may be regarded as portions of these. Referring to his report for details, it may be said that the first or easternmost belt of these rocks in North Carolina, is in part concealed under the tertiary strata east of Raleigh, but is again seen west of the Raleigh granite-range. The second, a band with a breadth of from twenty to forty miles, extends from north to south across the State, along the western border of the mesozoic area. . . • Maclure, Observations on the Geology of the United States of Am- erica, with a Geological Map, etc., reprinted from the 1st vol. of the Trans. of the Amer. Philos. ooc, new series. Philadelphia : 1817; pp. 42, 43. southeast. passing by Pennsylva- h Virginia ; dstone, and unty, North re described Lte, granular Chrough the the Taconic those of the another area •anch of the tains, in Mc- belt near the 30 acquainted in his report ions it as one vithin its bor- rofessor Kerr, bution of these the geological lese rocks pre- n-ops, with two 36 regarded as t for details, it t belt of these ^aled under the lin seen west of a band with a ends from north rn border of the mited States of Am- 1st vol. of the Trans. ,: 1817; pp. 42, 43. XI.] LOWER TACONIC ROCKS. 659 § 61. The third, designated as the King's-Mountain belt, and including, besides the mountain of that name, the elevations known as Crowder's, Spencer's, and Ander- son's Mountains, is in the southern part of the State, west of the Catawba River ; stretching through Catawba, Lin- coln, and Gaston Counties, and passing thence into South Carolina. This third belt is in the strike of that traced by Maclure from the Delaware into Stokes and Surry Counties, in the northern part of the State, and is re- garded by Kerr as a continuation of it, though interrupted for some distance between the Yadkin and the Catawba. § 62. The fourth is a great belt which, like the second, is continuous across the State, along the Blue Ridge ; the rocks in question passing from the east to the west side of that chain in the southwest part of their extension. This belt, at the Swannanoa Gap, is from six to seven miles broad, but has its greatest development in the Linville Mountains, where it includes the area of these rocks noticed by Maclure on the north branch of the Catawba, in McDowell County, and also an important section de- scribed by Emmons, on the Frenoh-Broad River, in Bun- combe County, to be noticed farther on. § 63. The fifth or western area of the Taconic rocks is confined, according to Kerr, to the southwestern part of the State, into which it extends from Tennessee, including the mass of the Smoky Mountain of the Unaka range, and stretches from Madison County, widening southward, until it includes almost the whole breadth of Cherokee County, in the southwest corner of the State. To the Taconic of this region belongs the well known section near Murphy in that county. The rocks of this belt, as seen at Paint Rock on the French-Broad River, in Madi- son County, beginning at the Tennessee line, are 1 y Kerr, and by Safford, identified as a continuous part of the Ocoee, Chilhowee, and Knox groups of the latter.* It is * Kerr. Report Geological Survey of North C"\rolina, 1875, vol. I., p. 131, and pp. 138, 139. !llr |:i fl fc;.i iw-'.llUfi'"' 5G0 THE TACONIC QUESTION IN GEOLOGY. [XI. under these names that Safford has described tlie Lower Taconic series of the Appalachian valley, as found in eastern Tennessee, of which the belt in the southwest counties of North Carolina forms a part. The Ocoee slates, and the Chilhowee or Scolithus-sandstone of east- ern Tennessee, both recognized by Emmons as Lower Taconic, represent the Lower Primal slates and quartz- ites, which, in this region, have a greatly augmented volume. In Alabama, according to Prof. Eugene A. Smith, the thickness of this sandstone is not less than 2000 feet, and that of the underlying slates, 10,000 feet. § 64. Professor Kerr, while recognizing in these rocks the strata described by Emmons under the name of Ta- conic, gave them, as he tells us in his report of 1875, provisionally, the name of Huronian, both designations appearing in thu legend of the accompanying map. In explanation of this, it is to be remarked that he then included all of the more ancient crystalline rocks of North Carolina under the head of Laurentian, which he divided into Lower Laurentian (also called granite in his engraved sections) and Upper Laurentian. The latter name (at one time used by the geological survey of Can- ada to designate an entirely different group of rocks, the Norian) was by Professor Kerr applied to the series of younger gneisses and micaceous and hornblendie schists (with included beds of chrysolite or olivine-rock), which is the Montalban series of the author. These rocks, in 1877, 1 found to rest in Mitchell County, North Carolina, directly upon the ancient granitoid gneisses of the Laurentian, the Huronian being absent. The true pkxe of this, as appears from multiplied observations, is below, not above, the Montalban, and it moreover differs entirely in its lithological characters from the Lower Taconic rocks, which are found above the Montalban hori- zon. It remains, however, to be determined whether true Huronian and Arvonian rocks may not occur in parts of North Carolina, and may not be represented by some of ZI.] LOWER TACONIC ROCKS. 561 the greenstones and the feldspar-porphyries noticed by- Professor Kerr as found in parts of the Montalban (Upper Laurentian) area of the State. § 65. The Taconic strata of North Carolina are de- scribed by Kerr as resting in some places upon the granitic rocks, and in others upon the upper or Montal- ban series, and in part made up of its ruins. Pebbles of the older crystalline rocks, in which I have recognized both gneiss and mica-schist, are often met with in the con- glomerates of the series. In the quartzites of the second belt at Troy, in Montgomery County, occur the silicious concretions regarded by Emmons as organic, and described by him under the name of Paleotrochis. Other beds of granular quartzite are flexible, constituting the variety known as itacolumite. With these, besides the usual argillites and unctuous schists, are found beds of pure massive pyrophyllite, which was by Emmons described as agalmatolite, and has been mistaken for steatite or com- pact talc, beds of which are also met With in this series. The schists are sometimes graphitic, and even include beds of graphite, ^s in the King's-Mountain belt. The quartzites of the series frequently contain cyan^te and rutile, and also include, as in Pennsylvania, both mag- netic and specular iron-ores, as will be noticed farther on, in the account of these rocks in South Carolina. The characteristic limestones — often becoming marbles — and the limonites of epigenic origin here, as in other regions, mark the series. § 66. I have elsewhere described these rocks as seen by me in the fourth belt in North Carolina, on the north branch of the Catawba, near Marion, in McDowell County, where they were first seen by Maclure ; * and have noticed the granular quartz-rock, often becoming thinly bedded and flexible, the unctuous micaceous schists, the limon- ites, and the limestones, as having all the characters of * Proc. Boston Soc. Nat. Hist., 1878, xix., p. 277, and Azoic Rocks, pp. 207, 208. ^„^ TACONIC QUESTION «■ "■^'-O''^' 1 • [XI. / 1 -a A section in this these rocks as seen in P«™f J^rm Springs, in Bun- same belt favtl.ev southwest, a^ Wa ._! ^^^^ be combe County, des«nbed .^y 1^^^ ^^,^^ ,„„ks m the compared with sm.lar B«<;t'^ ° „ Pennsylvania, and in Appalachian valley m V>rg'W»' ^ ;^^ ,,hich latter he BeLhire County, Massachusetts ^^^^, „i,h a wes^ especially compared •»• ™ ^„,„be County uncontorma- Jtd inclination, rest '"^unc". ^^^^^ „f the bly upon the eastward-dipping c y ^ ,„„giomerate bL Uidge. They P-»^^^\*; succession o£ slates ^ith a talcose P''^''^' *°"°';!^„ Jar and vitreous quartz- with interposed masses ot granui tliickness o£ to and Wo^c-^'^t^^^'^eTd W) feet of limestone, about 2400 ieet. To «rese succeed ^^^ ^,^,^,^ foUovvod by more than 150 tee^ ^^^^^^ ^^p^r- hesides a farther mass of comer ^^^^.^^^ ^^^ fectly exposed. Emmons notes 'n^^^^^^^,, ,^hich is development of l^f^'*? f"^;, Z L strata belovv the Sitr.;." ?— -5- •■«"'-"■"■ "•* sachusetts.* , , unconformable superpo- 5 67. In connection ™'* "^^ " yer crystalline rocks, silon of the Taeo"io^^,nrV»th C^^^^^^^^ and in New Emmons has noted, both mjoit the Tacom York, the appearance m ^om* P' ,^ ^nd concludes that strata, of granitic ''f .^l^'^^^^y^derlying floor, ex- thev are portions of the J"«B"' , ^be Taconic. Of ;':'ed byL folding -ddc—n of th^^^^^^^^^^ ^^^, certain interposed bands 1 « «»y« ^^^\ careful exam- . . ", gard them as interlammated r^^^^^ ^^ ^^^^ „^ber will . Jtion of the '^''l.-f "\"' '^Tp imary rocks are under- „snlt in the conviction that the p J ^^^^ ^th the lying and older '""^t'chthy geographically separate."t /edimentary rocks which *ey g^^^ ^^^^^ ^^ ^^ ^ # American Geology, II-, P-**' y. XI.] LOWER TACONIC ROCKS. )03 ition in this ^gs, in Bun- 855, may he rocks in the rania, and in Ach latter he ,, with a west- l unconforma- rocks of the conglomerate jsion of slates, itreous quartz- ite thickness ot .t of limestone, Igrained slates, se rocks, imper- .ction the larger lerates, which IS strata belovr the bat the rocks ot logically indistm- lliamstown, Mas- 'ormahle superpo- cvystalline rocks, >lina and in Nexv ,,ong the Taconic ,nd concludes that derlying floor, ex- ^ the Taconic. ^^ ,e geologist might ,^t a careful exam- s to each other will ,y rocks are undei- .Jnnection with the phically separate. T an Geology, W-.P- 26. In a later book, his " Manual of Geology," published in 1860, Emmons reproduces the figures of the sections noticed above, but gives with them only very brief de- scriptions. He there states that the maximum thickness of the Lower Taconic rocks may be about 5000 feet. Above the basal conglomerate, which is sometimes absent, there are generally, according i^ his later statement, three masses of quartzite, divided by slates, the upper of these being often vitreous, and the lower granular in tex- ture. The roofing-slates are said to occur in the upper part of the mass of slates which overlies the limestone. § 68. Passing southward from North Carolina, the Lower Taconic rocks were by Tuomey traced across South Carolina, and into Georgia and Alabama. He de- scribed them as a series of quartzites, with talcose slates and marbles, well displayed in the Spartanburg district, in the northern part of South Carolina. They are also met with in Pickens, the most western district of the State, in what is probably a continuation of the fourth belt of North Carolina, and extend across Pickens into the contiguous portions of Georgia. The belt just men- tioned has there a considerable development in Habersham County, where it has been seen by the writer, and also in the adjacent counties of Hall and Union, a region in which a considerable number of diamonds, supposed to occur in this series, have been found. There appears also to be another and more eastern belt, which, according to C. U. Shepard, passes from South Carolina into the counties of Lincoln and Columbia in Georgia. In the former of these occurs Graves Mountain, known to mineralogists as a locality of pyrophyllite and cyanite, as well as of re- markable crystals " of rutile and of lazulite, all of which are found with the granular quartzites of the series.* § 69. These rocks were the subject of extended and careful studies by the late Oscar Lieber, whose examina- tions were chiefly confined to the area in the northern * Amer. Jour. Science, 1859, xxvil., p. 36. 564 THE TACONIC QUESTION IN GEOLOGY. Bn» part of South Carolina, where, according to him, they are best seen at and near King's Mountain, in York District, and occupy a region about twenty-one miles long and from four to seven miles wide in York, Spartanburg, and Union districts. This region is the southward prolonga- tion and terminus of the third, or King's-Mountain, belt of North Carolina, which we have described as passing southward from Gaston County, and as being, in the opinion of Kerr, the continuation of the belt which from the Yadkin River is traced northeastward to the Chester and Lancaster valleys of Pennsylvania, and thence into the great Appalachian valley. § 70. The rock most characteristic of this series is, according to Lieber, the granular, more or less schistose quartzite, which, with its associated iron-ores and slates, he compared to the similar rock described by Eschwege, from the province of Minas Geraes, in Brazil, as a chloritic or schistose quartz-rock. This, from its occurrence at Mount Itacolumi, near Villa Rica, was by Eschwege called " itacolumite," ^ name which was also adopted by Humboldt and Claussen for these and related rocks as a whole, but is now commonly given only to the flexible and elastic variety of the quartzite, the "elastic sand- stone " of Martius. This variety, however, is exceptional alike in Brazil and in our Lower Taconic series ; and the designation of itacolumite was by Lieber applied not only to the whole of the quartzite, but to its interstratified schists and limestones, which he described as the Itacolu- mitic group or series. § 71. These rocks, on lithological grounds, were con- jectured by Lieber to be the stratigraphical equivalents of the Itacolumite or diamond-bearing series of Brazil, and of the similar rocks described by Jacquemont, and later by Claussen, as occurring in the diamond region of India, being the Lower Vindhyan series of the present geologi- cal survey of that country. He also noticed its probable relation to the rocks found by Helmersen and Hofmann XI.J LOWER TACONIC ROCKS. 5G5 m, they are rk District, 8 long and ^nburg, and a prolonga- luntain, belt L as passing eing, in the b which from , the Chester L thence into ihis series is, less schistose es and slates, by Eschwege, l1, as a chloritic occurrence at by Eschwege so adopted by ited rocks as a to the flexible "elastic sand- :, is exceptional series; and the vpplied not only ;8 interstratified I as the Itacolu- )unds, were con- sal equivalents ot es of Brazil, and emont, and later region of India, e present geologi- ticed its probable sen and Hofmann in Russia, in the southern Urals, which they had described as identical with the Itacolumite series of Brazil, and wliich have since been found to be dianiantiferous. I have elsewhere discussed at some lenrrth tlio liistory of these rocks,* which are again noticed farther on, in § 208. § 72. Lieber's studies are to he found in liis four annual reports of the geology of South Carolina, puhlished in 1856-60. With the third report there appears, as a sup- plement to the first three, an essay on the Itacoluniitic series, resuming his conclusions and ohservations up to the year 1859.t This same essay was also published in German in 1860.:^ The studies by Lieber are the more interesting and in- structive as they are the work of a student trained in a foreign school, and were made without any reference to the preceding investigations of Maclure, Eaton, Emmoi i, or Rogers, and apparently without the knowledge that these rocks extended to the north of the Carolinas. As his reports are very rare, and hut little known, I have thought it desirable to give in the following pages an abstract of his observations. In the four annual reports already noticed, together with the included supplement, Lieber proposed to describe the ancient stratified rocks of the State, and successively corrected and enlarged his descriptions, by collating which we are enabled to frame a connected statement of his views. Lieber divides the * Report of the Smithsonian Institution for 1882; Review of the Progress of Geology. t The Itacolumite and its associates, comprising observations on their geological importance and their connection with the occurrence of gold; a Contribution to the (Jeologic Chronology of the Southern Alleghanies, supplementary to Reports I., II., and III., by O. M. Lieber, state geolo- gist, Columbia, S. C, 18r)0. This, though having a separate title, is paged consecutively (pp. 77-149) with Report III., published in 1858, witli which it forms one volume. The relations of gold to the Itacolu- mite, and to other rocks, are considered in a subsequent part of the same volume, pp. 153-220. t The German edition of Lieber's essay appears in the Gangstudien of Von Cotta and Herrm. Miiller; dritter Band, drittes und viertes Heft, pp. 309-507. Freiburg, 1860 ,„B TACOKIC QUESTION IN OEOLOQV. t^ 6G0 , ,, r, . ' ^^ tl.rpe nfli'ts, namely, tue e,ystaUine rocks ot tbo State «Uo «27» „,:,,„„„,„„,itio Itaoolu.uitic gi'"«P |»'^1 ^'' "'For the fl.-»t-na.uea o,- „-,aaie'one of these » ""Xs ft t„taius rocks closely .,t a natural group. '"" ""^^^^^^^^^ „»«,ciated. „Uicaa,>a everywhere "'""^ «'\„ n,„ Klug's-Mouutau. « 73. His descnpuo .9 am y I ,^^^^^ ,i„|,„„,i ,c|iou, as B-''/"«T^''^:rgt^l map. with an i.leal i„ 5 «9. and of wuch a 8°°' 8«' , ^ ^n.). A de»er>pt.on Bcotiou. is given ,u l''-'!" " "^ <-[ .^ displayed, was atter- i„ Ueport l.,ut tl";;'^-""^,f "afmttllygiven,withson,o wards corrected iu Uc,o t II. an^ i^J^ l^^,,,,a«S *" bered in descending order : - 1 Banded blue crystalline limestone. I Abed of 5;-^-S£ar layers of catawbarlte (an aggregate 3. Talcose slate, ^'"" '|^ ' > t»Sn=='.orW.,.o...=»* variety. ^^^. „j micaceous l^«™'^^l*\''^\.r ' ,, Spceular ^f J^^^' ^^^^^^^^^^^^^^^ . 5 74. Beneath these, Lieberpiac ^^.^^.^i.^e, hovn- Jic division, clay-Blate '^l^l\j, of which wevo blende-slate, S^''''\^''tZTeZ roo\.s- He describes a coniectnred by him to be igneous r ^^^.^^^ ^^^^^^ >QY. po. XI.) LOWER TACONIO IlOCKa. 507 8, namely, the b-Itacolumitio ivat-niuneil i>r tUo distinuti^'ii rock* closely I. iugs-Moimtiuu e luive deline.l p, vvitli an iileal A descviptiou ayed, was aftei- given,witli8i""o substituting tlio id adding »«»"« ,on, whicli repre- by Lieber, num- wbarlte (an aggregate anded, sometimes with rs of the flexible, elastic hematite and quartz, e replaces hematite. iually into the quartzUe iiv bis 8ub-ltacolu- be, mica-slate, bovu- ,,,e of wbicb wevc ,k8. He describes jv e gray variety wttU Jt from tbe co.u.e of tbe State, and is cbists bolding browu iron-ores derived from [jyrites. Those various rocks luivc, iu his o[)inion, no necessary rehition to those above them, hut are simply tlie strata which, in different parts, unchu-- lie the Itacolumitic series. Little is said about the undorlyinf]f elay-slate and talcose slate, botii of which may perhaps belong to the Itacolumitic series. To this all the others were referred, with the possible exception of the upper or blue limestone, which was provisionally designated as super-Itacolumitic, because it is unlike the marbles below, and also is apparently above the horizon of the gold-veins, which are common to the inferior rocks of the Itacolumitic series. § 75. No measurements of the several members of the series are given by Lieber, but as seen at King's Moun- tain, he says, " its thickness will probably equal nearly a mile." As represented in the engraved section in Report I. (plate v.), it is highly contorted, and in some places shows inverted dips, tlie strike being between north and northeast. No direct evidences of organic life are seen in the series, if we except the forms observed by C. U. Shepard in the upper blue limestone at the Broad-River quarry in York District, and by him supposed to be im- pressions of stems of Equiseta, with swelling nodes.* This description recalls tlie distinctly nodose character exhibited by the so-called Scolithus of the Primal quartz- ite of Pennsylvania, and the cylindrical forms in the Auroral limestone and its accompanying strata elsewhere, which I have compared with them (§ 34).t § 76. The quartzites of the series present the charac- teristics which we have recognized in those of the Primal series elsewhere, being sometimes conglomerate, and at other times massive, compact, concretionary, or granular ; often with an admixture of a foliated mineral, which gives them a laminated character, and assimilates them to the * Shepanl, Report to the Swedish Iron Manufacturing Company, Charleston, 1854; cited by Lieber, Rep. H., p. 8?. t Azoic Rocks, pp. 137, 138, 206. I I. ^ . "I «! .1'! 1 'I i t i,l- ? 5G8 THE TACONIO QUESTION IN GEOLOGY. [XI. older crystalline schists. This interposed mineral is, ac- cording to Lieber, sometimes a mica, and at other times chlorite or talc. We have seen that the schistose strata of this series in Pennsylvania, and in North Carolina, are sometimes chloritic, or contain the species venerite, a copper-chlorite (ante, page 357) ; while at other times they consist wholly or in part of a hydrous mica (damour- ite or sericite), pyrophyllite, or true talc. All of these species are probably confounded by the common epithet, " talcose," applied to these rocks, though true talc is com- paratively rare. We have also noticed the occurrence in the schists of this series, in Pennsylvania, of serpentine, of amphibole, and of garnet. Cyanite and rutile, the latter in large and fine crystals, are not unfrequently found in the granular quartzites of the series, and stauro- lite is also met with. The lower limestone of Lieber's section sometimes contains tremolite. It is marked by dark bands, and frequently by talcose seams, which render it unfit for use as a marble. In King's Mountain, this limestone is traversed by auriferous veins, and the quartz- ites and schists of the series are also auriferous, and constitute the chief gold-bearing rocks of the southern States. § 77. The iron-ores of the series in South Carolina, other than the iimonites, are by Lieber included under three varieties. First, an aggregate of magnetite with talc, called by him catawbarite, the talc in some cases dis- appearing ; second, a schistose silicious hematite, described as a specular schist, in which foliated hematite takes the place of mica. This, by the substitution of magnetite for hematite, passes into a rock which, from a locality in Braidl, has been named itabirite. These ores occur in beds or lenticular masses ; the latter two varieties in the quartzite, and the catawbarite in the Oalcose schists of the series. § 78. The hydrous iron-ore or limonite, so abundant in this series elsewhere, received but little notice from XI.] LOWER TACONIC ROCKS. 569 Licber. He mentions, ho trover, its occurrence in the King's-Moinitain region intercalated in decaying talcose shvtes, with red clays and an underlying stratum of kaolin. The limonite is here, as in parts of Pennsylvania (§ 26), associated with anhydrous red oxyd, and Lieber conceives this, and some other similar deposits in the region, to have originated from the hydration and alteration of specular 'iron-ore, or of magnetite. This view, which has been frequently advanced by others, is, however, inconsistent with the known permanency and unalterable character of the anhydrous oxyds of iron, and, moreover, with the well known origin of the hydrous ore by epigenesis from pyrites or from siderite. Lieber himself mentions else- where the occurrence of beds of limonite, intercalated in the talcose slates of the series, and due to the alteration of masses of pyrites, which is found unchanged in depth. § 79. We owe to Prof. Henry Wurtz a valuable paper, published in 1859, on the mineralogy of the northward extension of the King's-Mountain belt, as seen in Gaston and Lincoln Counties in North Carolina. He there no- ticed the itacolumite-rock, and its supposed relations to the diamond, described the anhydrous iron-ores under the names of magnetite-schist and hematite-schist, and more- over what he called a pyrites-schist. He farther observed great interstratified beds of limonite, which he regarded as derived from the alteration of a pyrites that is found unchanged in the deep workings of these ores. With them, and elsewhere in the talcose schists of the region, he observed the frequent occurrence of black earthy man- ganese-oxyd, containing much cobalt and some nickel.* It is worthy of notice in this connection that both the magnetites and the limonites of this horizon in Pennsyl- vania generally contain more or less cobalt, as shown in numerous analyses by Genth and McCreath. The pyrites found at the Cornwall iron-mine in Pennsylvania is also cobaltiferous. r 1 '1- V,.;^1 * Amer. Jour. Science, xxvii., pp. 24-31. iU Ei^&imm'^ li> f II; ii 670 THE TACONiC QUESTION IK GEOLOGY. [XI. The magnetic and specular ores found so abundantly in the Primal series of Pennsylvania, and already described at length, are evidently the equivalents of those described by Lieber and by Wurtz, and constitute an important and widely extended ore-bearing horizon. The silicate mingled with the magnetites in many of the Pennsylvanii; deposits, is probably more nearly related to pyroxene than to talc in composition. The mineralogy of all of these deposits demands careful study, inasmuch as they belong to a distinct and well marked horizon of crystalline rocks, the importance and geological significance of wliicli has hitherto been to a great extent overlooked by American geologists. The Itacnlumitic series of Lieber, with its estimated approximative thickness of 5000 feet, being evidently the Lower Taconic of Emmons, it remains to be seen whether the upper blue limestone, provisionally regarded by Lieber as distinct, really belongs to a higher horizon, or is a member of the series. In the latter case, the upper schists and the roofing-slates of the Lower Taconic are unrepresented in this area, and have probably been removed by erosion. The best locality for the study of the whole series in South Carolina is, according to Lieber, at Limestone Springs, in the Spartanburg district. § 80. In this connection mention should be made of the occurrence of several narrow belts of Lower Taconic rocks folded in the gneiss of the Highlands east of tlie Appalachian valley, in northern New Jersey, where they have been carefully studied and described by Cook, and are well seen in the Pohatcong and Muscanetcong valleys. They also extend into southeastern New York, where little is known of their distribution, and where they have been confounded with the older Laurentian rocks, into which they were supposed, by Nuttail, Mather, and H. D. Rogers, to graduate.* In New Jersey, where Cook has shown the fallacy of this view, the Auroral limestones, * Hunt, Azoic Rocks, pages 40, 42. JY. ^^• bundantly in ,dy describee! ,ose described an important The silicate Pennsylvania pyroxene than )f all of these xs they belong ystalline rocks, 5 of which has id by American li its esfi mated Dcing evidently lins to be seen Lonally regarded higher horizon, latter case, the B Lower Taconic e probably been for the study of ording to Lieber, g district, ould be made of J Lower Taconic lands east of the ersey, where they 3ed by Cook, and icanetcong valleys. New York, where d where they have •entian rocks, into Mather, and H. D- y, where Cook has Auroral limestones, kO, 42. XI.] LOWER TACONIC EOCKS. 671 associated with limonites, and often overlaid with slates, are found resting directly on the gneiss, or with a thin intervening layer of the Primal sandstone. These strata are much folded and faulted, and sometimes present over- turned flexures, giving the whole succession an eastward inclination.* All of these rocks above the gneiss are, in accordance with the classification of Rogers in Pennsyl- vania, referred by Cook to the infra-Trenton portions of the Champlain division. The relations of the Green-Pond Mountain conglomerate, found in this region, will be no- ticed farther on. § 81. The parallel belts of Lower Taconic rocks found east of the Blue Ridge, in the southern States, and the final disappearance of these rocks beneath the tertiary to the east of Raleigh, show that they were once widely spread over the floor of the more ancient crystalline rocks which now form the Atlantic belt. To the north of New York, where this belt, greatly contracted between the James and the Hudson Rivers, again broadens, we might look for farther areas of Lower Taconic rocks iu New England, and in the provinces lying farther to the north and east. We find, in fact, to the east of the Green Mountains, in Vermont, a series of limestones with soft micaceous slates, which have been compared with the Lower Taconic, and may perhaps represent it. To this horizon may also not improbably belong the considerable areas of argillites, often roofing-slates, found in the prov- ince of Quebec, to the north of Lake Memphremagog, extending to Melbourne, and occupying what I have called the Windsor basin. These argillites overlie the Huronian schists, and are themselves unconformably overlaid by Silurian limestones, which repose alike upon the argillites and upon the Huronian series. § 82. Farther east, in Maine, are areas of argillites, and others of quartzose conglomerates, limestones, and soft talcose schists, which were declared by Emmons to * Cook, Geology of New Jersey, 1868, pp. 70, 144. if n ii (J 672 THE TACONIO QUESTION IN GEOLOGY. [XI. resemble the Lower Taconic rocks of western Massachu- setts, and to rest unconformably upon the ancient niica- schists and gneisses of the region. This series, whicli inchides the limestones of Rockland and of Camden, has, according to Emmons, a thickness, in the latter locality, of 2000 feet, and is uy him regarded as belonging to the Lower Taconic ; to which, moreover, he refers, with much probability, many of the silicious and argillaceous schists of this part of Maine. The limestones and associated rocks of C;imberland, Rhode Island, are also supposed by Emmons to belong to the same horizon.* These, the present writer has not yet personally examined. § 83. In southei'n New Brunswick, as I have pointed out, there are found numerous exposures of rocks closely resembling those of Camden. They have been much eroded, but are seen at several points along the coast, as at Fryc's Island, the peninsula of L'Etang, Pisarinco, and the mouth of the River St. John. At this last locality, a section along the Green-Ilead road, on the right bank of the river, is described in detail by Matthew and Bailey in the report of the geological survey of Canada for 1870. The strata, with a general southeast dip of about fifty degrees, have a breadth, across the strike, of 4100 feet, of which 1500 are limestones, and the remainder chiefly quartzites, often schistose, with argillaceous and some- what micaceous schists, and occasional hornblendic layers. Considerable masses of conglomerate, with silicious and calcareous pebbles, are also included in the series, tiie members of which are not improbably repeated by dislo- cations. The limestones, of which there appear to be several masses two or three hundred feet in breadth, are in part distinctly crystalline and white, or banded with blue and gray colors, and in part finely granular, gray- ish, schistose, and sometimes concretionary. They are frequently magnesian, and occasionally contain small * Emmons, Agriculture of New York, I., 97-101, and Amer. Geology, II., 20-22 ; also Hunt, Azoic Rocks, 179. iGY. wfc*. rn Massaclni- ancient nuc;i- series, whieU Camden, lias, latter locality, onging to the jrs, with much laceous schists and associated 50 supposed by * These, the ined. I have pointed 3f rocks closely ve been much ng the coast, as r, Pisarinco, and s last locality, a le right bank of w and Bailey in ;anada for 1870. p of about fifty , of 4100 feet, of emamder chiefly ceous and some- )rnblendic layers. nth silicious and II the series, the ■epeated by dislo- ere appear to be et in breadth, are 5, or banded with ly granular, gray- onary. They are [ly contain small Jl, and Amer. Geology, XI.] LOWEU TACONIC ROCKS. 573 masses of yellow serpentine, and a silvery-white mica. Portions of the limestone are apparently colored by a carbonaceous matter, and a bed of impure schistose graphite, which has not the crystalline aspect of the Laurentian graphites, is mined in these rocks near the citj' of St. John. These limestones have yielded to Sir. J. W. Dawson the remains of Eozoiin Canadense. The argil- laceous beds, sometimes schistose, and occasionally graph- itic, wliich lie between tlie quartzite and the limestones, closely resemble those found in similar associations in the Taconian areas already noticed. § 83 A. [The rocks on tlie southern slope of the Cobe- quid Hills, at the head of the Bay of Fundy, belong to the same series as those of the St. John. Nearly verti- cal in attitude, and unconformably overlaid by carbonifer- ous strata, they were long since described by Dawson as " a metamorphic series," supposed to be of paleozoic age.* They consist of a great underlying mass of quartzite, often granular, with massive beds of granular white limestone, black or olive-colored argillites, often lustrous, and a soft greenish or grayish, apparently argil- laceous rock. These strata are intersected by great veins, often brecciated in structure, and filled with sparry ferrous carbonates of varying composition, sometimes nearly pure siderite, and including portions of crystalline hematite and magnetite. The carbonates of these veins near the sur- face are changed into limonite, and have been extensively mined at Londonderry, Nova Scotia. I have there found evidences of contemporaneous deposition of iron-ores in tliin layers of crystalline hematite interbedded with the granular limestones.] § 84. This succession of crystalline limestones, quartz- ites and slates, in New Brunswick and Nova Scotia is clearly older than the un crystalline sandstones and shales of Lower Cambrian (Menevian) age, which, with their characteristic fauna, are at St. John found to be in prox- * Dawson, Acadian Geology, 2d ed., p. 682. [hi ■: 1 I0i '■ I'' :■■■ f' I i ^ I'li'^'i!':'. 1J: 1') 674 THE TACONIO QUESTION IN GEOLOGY. [XI. imity. The latter strata appear to be in part made up of the ruins of the older schists, and in one section, beds of quartzite and conglomerate, believed to belong to the limestone series, occur between the Menevian and the underlying Huronian strata. A mile or two away, how- ever, the limestone series is seen to rest upon red grani- toid gneiss, regarded as Laurentian ; and was itself de- scribed, in the report just mentioned, as an upper member of the Laurentian series. The evidence above adduced shows that we have here a great system resting uncon- formably alike on Laurentian and Huronian, and at the same time wholly distinct from the Lower Cambrian. From these facts, ami from its close resemblance to the Lower Taconic of Maine, and of western New England, it was in 1875, by the preset' t writer, referred to the Taconic series.* A great mass of similar limestones and marbles, with soft micaceous schists, described by Murray as occurring in Newfoundland between the gneisses and the fossiliferous Cambrian, may not improbably represent the Lower Taconic.f § 85. As we go northward in the Champlain valley, the Lower Taconic, which is seen in southern and central Vermont, at the western base of the Green Mountains, passes beneath newer strata. From thence northeast- ward, we have no certain evidence of the existence of this series between the latter and the belt of crystalline strata of Huronian age, which may be traced along the south- east side of the St. Lawrence valley, to a point a little farther east than the meridian of Quebec, where the crystalline rocks disappear beneath the surrounding pale- ozoic strata. If, however, we pass westward, we find in Hastings County, north of the eastern extremity of Lake Ontario, a considerable area occupied by quartzites, con- glomerates, limestones, micaceous slates, and argillites, resembling closely those of the various Taconian areas. * Proc. Bos. Soc. Nat. Hist., xvi., TjOO ; and Azoic Rocks, pp. 170-180. t Hunt, Amer. Jour. Sci., 1870, vol. I., p. 86. €. XI.] LOWER TACONIC ROCKS. ; made up of tion, beds of ■long to th.e vian and the away, how- 011 red grani- was itself de- ipper member bove adduced resting uncon- an, and at the yer Cambrian, (iblance to the New England, eferred to the limestones and ibed by Murray tie gneisses and jbably represent lamplain valley, lern and central reen Mountains, iience northeast- existence of this crystalline strata along the south- o a point a little lebec, where the surrounding pale- ■ward, we find in ixtremity of Lake jy quartzites, con- 3S, and argillites, s Taconian areas. ,oic Rocks, pp. n«-l80. These strata, whicli rest unconformably alike upon the Laurentian and Huronian rocks of the district, are them- selves arranged in several synclinals, with moderate dips, and are unconformably overlaid by the fossiliferous lime- stones of the Trenton ; the lower members of the Cham- plain division being absent throughout this region. The conglomerates include pebbles from both of the under- lying groups. Crystalline dolomites, constituting mar- bles, are found in the series, and above them, a mass of about 1000 feet of fine-grained, grayish and bluish, earthy and somewhat schistose limestones ; the whole series being estimated at 3800 feet. These rocks, which were first particularly described by Thomas Macfarlane (then of the geological survey of Canada), in 1864, were subse- quently known, in the reports of the survey, as the Hastings series ; and were by Logan, in 1866, compared with the micaceous limestone-series of eastern Vermont (§ 81). In 1875, the writer, after an examination of the three regions, compared the rocks of the Hastings series with the similar rocks of southern New Brunswick, and of Berkshire County, Massachusetts, and described them as Lower Taconic* Xt may here be mentioned that areas of Montalban gneisses, and mica-schists occur in the vicinity of the Taconian rocks of Hastings County, in Ontario. § 86. These rocks are not destitute of direct evidences of organic life, having furnished remains of Eozoon Cana- dense^ which have been described and figured by Dawson. Numerous specimens of this have been found in Tudor, "imbedded in an impure, earthy, dark gray limestone, with which, and with carbonaceous matter, the cavities of the white calcareous skeleton are filled " ; unlike those of the Eozoon from the Grenville series on the Ottawa, which are generally filled with serpentine or pyroxene. Dawson farther noticed, in some of the impure dark-colored limestones of the Hastings series from Maduc, ■ * Azoic Rocks, pp. 170-172, and p. 177. 1 it .1 , ■» yi . i' ■m ji ^ 676 THK TACONIC QUESTION IN GEOLOGY. m. II i. "fibres and gnumles of carbonaceous matter which do not conform to the crystalline structure, and present forms quite similar to those which, in more modern lime- stones, result from the decomposition of algse. Though retaining mere traces of organic structure, no doubt would be entertained as to their vegetable origin if they were found in fossiliferous limestones." He noticed also a similar limestone from the same vicinity, which is apparently "a finely lanunated sediment, and shows per- forations of various sizes, somewhat scalloped on the edges, and filled with grains of rounded silicious sand." Other specimens from the same region are said to present, on their weathered surfaces, indications of similar circular perforations, having the aspect of Scolithus or worm- burrows. Some of these markings from jNIadoc were sub- sequently figured by Dawson, and designated "annelid- burrows," with the remark that " there can be no doubt as to their nature." * These are as yet known only by a few transverse sections, and cannot, therefore, be com- pared with the cylindrical markings referred to Scolithus and to Monocraterion, in the Taconic quartzites and lime- stones of the Appalachian valley (§§ 34, 62). § 87. Brooks described in 1872 an area of rocks in St. Lawrence County, New York, lying along the northern base of the Adirondacks. They include the Caledonia and Keene iron-mines of that region, and appear as a series of folded strata, with a northeast strike, resting in apparent unconformity upon reddish Laurentian gneiss. The rocks in question consist of granular quartzite, crystalline lime- stone, and a greenish schistose rock described as magne- sian. A bed of quartzite is interstratified with the lime- stones, which include treraolite and are overlaid by the soft, greenish, gray-weathering schists, to which succeed the micaceous and earthy red hematites in lenticular masses, intercalated with similar schists and masses of * Dawson ; The Dawn of Life, pp. 110, 139 ; aad Hunt, Azoic Rocks, pp. 171-177. 'iw GY. . *^ ter wliicli do and present modern lime- JgjB. Though are, no doubt origin if they le noticed also nity, which is and shows per- alloped on the silicious sand." s said to present, ■ similar circular lithus or worm- jSIadoc were suh- :Tnated ^'annelid- can be no doubt known only by a .erefore, be com- vred to Scolithus lartzites and Ume- t,52). ^ . area of rocks m along the northern ide the Caledonia i appear as a series resting in apparent gneiss. The rocks te, crystalline lime- escribed as magne- Lfied with the lime- are overlaid by the 5 to which succeed Itites in lenticular lists and masses ot laad Hunt, Azoic lUKks. XI.] LOWER TACONIC ROCKS. 577 quartzite ; a friable sandstone, sometimes conglomerate, overlying the whole. White quartzo-feldspathic veins occur in the lower portion of the limestone. Emmons, who described this locality in 1842, and did not observe the lower quartzite, referred the overlying conglomerate to the Potsdam, and supposed the hematite, the limestone, and the greenish rock (by him called serpentine) to be all alike erupted plutonic masses. The observed thick- ness of the series, as there exposed, is, according to Brooks, not less than 700 feet, and its entire volume prob- ably much greater. Although they were by Brooks com- pared with the Lower Taconic of Emmons, I was disposed, in writing of these rocks in 1877, to regard them as a part of the Laurentian. Tliey were, at that time, compared with the crystalline limestones, with interstratified quartz- ites and conglomerates, found in Bastard in Ontario.* Farther consideration leads me to suspect that these rocks of St. Lawrence County are really an outlier of Taconian. In comparing these rocks in Maine, New Brunswick, Nova Scotia, Ontario, and northern New York with the similar rocks in the Appalachian valley, and elsewhere southwards, it should be remembered that in these latter regions the strata, in many cases, present, at their outcrops, soft materials, the results of sub-aerial decay ; whereas only their harder underlying portions are seen in the eroded regions farther northward. These varying conditions of outcrops of similar crystalline rocks in different geo- graphical areas have already been discussed at length in Essay VII. § 88. That the fauna of the Cambrian, as seen in the Menevian beds of our eastern coast, or in the so-called Potsdam which forms the base of the Cambrian in Minne- sota and in Wisconsin, marks the dawn of organic life, will now scarcely be maintained, even uy those who ques- * Brooks ; Anier. Jour. Science (3), iv., pp. 22-26 ; Hunt, Azoic "Rbiks, pp. 148 and 218; and Emmons, Geology of the Northern District of New York, pp. 92, 98. ^ 1514! f: ■:f5:kvtt;i;':|ilM 'i; 'i 14 'i * iif m THE TACONIO QUESTION IN GEOLOGY. txi. 678 show tlie existence, beneath tlu3i« ' • ti,e place *:ewherc, of '"-"j-^r '« ^Thie i. found which we l'»™ »«'«"''? '°„£ Lake Superior, a scries of aong the ""■■''"tJ;ls analtes%,ieh tl>e writer, nuartzites, impure l'""^.^*"™;' " These had been, by ?„ 1873, called the Am2k'g;7^_^^^^^ „£ «„ g^cat Logan, regarded »^ *« '°';, Copper-bearing series o Keweenian or 8o-caled .'JPP"^'i878, also attempted the region. Tbc -" "^^^ ,'J,Lnltones, and limestones to show that the o™?!""" „° A 'deriving uneonformably of the so-called Nlp-gonfoup, overly^ ^g^^ _^^^^^ ^.^^,^^._ the Animikie, are rot the •« . j,,^ g<,,Uou exam- which, so tar as "- ev^^c So'ded y^^^ ined by Logan and l"^!,''' * ^ maintained by goes, -gl>';^->r olrs. torn ifthological cons.dera- some, or, as neiu uj . -i # .ions, of a more "^"'^J'''^^ ,^, distinctness of which 5 89. The Ammikie group. ^^[ntained, has smcc from the overlying ^f^'^'^XZ^elX, and shown to been traced '^f ^'^f. \f i"ri,ontal sandstones of the underlie """""'"'"''''^I'i as belon-ring to the Potsdam St. Louis River, regarded »« be'°^= ^ a,^ Animilae Cambrian of the ^^fl^re t uu onformably upon the quartzites and ^^^'^'^^^^Z borizon of the Lower Huronian series, w°uld occupy .^ ^^^^ ^^.g,^^,, ( Taeonic. The writer »»"'^;'y„„e ota (in the vicinity this series, near Thomson, in Ml ^^^^.^^ ^^,„„eous „£ which they afford ;»°fl"S f Sded to J. W. Dawson concretions, one of f ^*J^'J'%he granular quartz.tes ; the remains of a ^eratose sp nge^ J^J^^^,^^ ,,,„ or sandstones of t^'' sene' ^^ described b5 ' • Azoic Rocks, PP- 238-241. observations ,ho westward, ig the phi^^ lere is found, or, a scries of icli the writer, , had been, by of the great Ming series of also attempted , and limestones ' unconformably Lt newer rocks; .leseHion exam- r Tluuider Bay, 8 maintained by ogical considera- actness of which ntained, has since 3II, and shown to jandstones of the r to the Potsdam ^,„the Animikie ormably upon the izon of the Lower n the argillites ot ta (in the vicinity amerous calcareous to J. W. Dawson granular quartzites 'xningledwithma^ 3S, as described bj it should be meli- us of the Menominee n 1846, referred b} 241. XI.] LOWER TACONIO R0CK8. 679 Houghton and Emraona to the Taconic system.* Tliere is reason to believe that tliese rocks of the Menominee region which, as described by Brooks and Piunpelly, in- clude great deposits of iron-ores and murbles, and appar- ently differ much from the Huronian in character, are, as 8up[)osi'd by Irving, identical with the Animikie rocks. § 90. [The publication, by the United States geological survey, in 1883, after the above was first printed, of Irving's report on the Copper-bearing rocks of Lake Superior, together with information since gathered from other sources, throws much additional light on the ques- tion of Huronian and Taconian in this region. Irving then announced his conclusion from the essential identity between the Animikie rocks (which have, according to him, a thickness of 10,000 feet) and those of the Penokie range in Wisconsin, and their close resemblance to the Iron-bearing strata of the Marquette and Menominee dis- tricts, that the whole of these constitute one great geologi- cal series. [These conclusions I have been enabled to verify, hav- ing, by the courtesy of Prof. N. H. Winchell, examined his collections of rocks from Minnesota, and been allowed the same privilege for the collections from Minnesota and the northern peninsula of Michigan, got by Dr. Rominger, who has, moreover, permitted me to consult his unpublished report of geological work in these regions in 1881-84. The Granitic and Dioritic groups of his published report of 1878-80 are by him regarded as plutonic rocks, in the former of which he embraces alike gneisses and strati- form granites, and the transversal granitic masses found in the dioritic group. This latter includes both massive and schistose, more or less chloritic varieties, and is intimately associated with the serpentines of the region, which apparently form a part of his dioritic group. [Resting in some cases upon this group, and in others upon the granitoid rocks, is a great system divided, in as- * Emmons, Agriculture of New York, i., p. 101. ff»; ■ \ . I \ 680 THE TACONIC QUESTION IN GEOLOOY. [XI. m^4 ivil!- cending order, by Roiningcu- in 1880, into a Quartzite group (which includes a Marble series), an Iron- ■ ',1 J, • m Ilk?) , j> m I'M 582 THE TACONIC QUESTION IN GEOLOGY. [ZI. pietre verdi — alike from the older granitoid and gneissic group, from the mica-schist or Montalban group, and from the great overlying Animikie or Taconian system, includ- ing the quartzites, marbles, iron-ores and argillites, is however manifest. The succession is thus brought into complete accordance with that which is found in many parts of the Appalachians, as well as in southern Europe, as pointed out in part iv. of Essay X.] § 92. Considering the pre-Cambrian age of the Lower Taconic of Emmons to be established, as well as its dis- tinctness alike from the older crystalline rocks below and from the Cambrian series above, to which Emmons had given the name of Upper Taconic — it was proposed by the writer, in 1878, to restrict the term of Taconic, — for which the alternative name of Taconian was then sug- gested, — to the Lower Taconic of Emmons.* The question as to whether the Cambrian is to be regarded as the base of the paleozoic series, or, in other words, whether the Taconian should be considered as belonging to eozoic or paleozoic time, was discussed by the author, in 1876, when he wrote as follows : " It will be found as difficult to draw the line between the eozoic and paleozoic as it is to define that between the mesozoic and paleozoic on the one hand, or between the mesozoic and cenozoic on the other. There are no hard-and-fast lines in nature ; breaks are local, and there is nowhere an apparent hiatus in the geological succes- sion which is not somewhere filled." Referring to the Lingula found by Prime in the Auroral limestone of Penn- sylvania, it was said: "This seemingly imperishable type of brachiopods may serve, like the rhizopods, repre- sented by Eozoon, as a connecting link between eozoic and paleozoic time." f Subsequently, in a paper read * On the Geology of the Eozoic Rocks of North America ; Proc. Bost. Soc. Nat. Hist., xix., 278 ; and Azoic Rocks, p, 207. t Proc. Amer. Assoc. Adv. Science, 1876, pp. 207-208 ; also. Azoic Rocks, p. 211. XI.] UPPER TACONIO P.OCKS. 583 before the National Academy of Sciences in April, 1880,* it was said of the Taconian series: "These older rocks are not without traces of organic life, having yielded in the Appalachian valley the original Scolithus, and related markings, besides obscure brachiopods; and in Ontario, besides similar Scolithus-like markings, a form apparently identical with the Eozoon of the more ancient gneisses. We may hope to find in the Taconian series a fauna which shall help to fill the wide interval that now divides that of the eozoic rocks from the Cambrian. We should seek, in the study of stratigraphical geology, not the breaks dividing groups from each other, so much as the beds of passage which serve to unite all these groups in one great system." i Pil 1 America ; Proc. Bost. n. , . ). 207-208; also, Azoic V. — THE UPPER TACONIC OR FIRST GRATWACKE. § 93. We now return to the history of the First Gray- wacke, which, as has been shown, was by Mather, in 1842, assigned, contrary to Eaton's conclusion, to a horizon above the Trenton limestone, and to the position of the Second Graywacke of the latter. Mather regarded the Taconic quartzite, limestone, and slates of Emmons as forming one continuous series with the succeeding First Graywacke of Eaton, and referred the whole succession to the various subdivisions of the New York system from the Potsdam to the Medina, both included. [Emmons, in his report on the Northern district of New York, in 1842, approached the discussion of this question with the remark that although the Taconic rocks do not appear within that district, a knowledge of them is requi- site to a correct understanding of the relations of the Champlain division. While maintaining that the Quartz- ite, Limestone, and Argillite (which Eaton had placed beneath the First Graywacke) were inferior to the Tren- ton limestone, and, indeed, to the whole New York system, Emmons therein showed a divided opinion as * Canadian Naturalist for 1880, vol. iz., p. 430. / ■ I. I' ii ill i ,m i«yi aa a 8»i |Miih < r^:frt»>»-gy of Cau- 3 contorted strata those of tiie belt horizontal (Tren- rred favorably to Emmons in 1842) torted strata, con- lerates, and Ume- tvaced in Canada ,ight bank of the a to the valley of lear Quebec south- crystalline schists 5^ere recognized by of the Graywacko- the view of Mather, eferred to the upper eclared to embrace id the immediately n-esentative of the llevy were supposed et found in the belt, ,u a conglomerate at xi.i UPrEIl TACONIC IN CANADA. rm Pointo Lnvia, which wore orrnneously supposed to bo derived f'/om tlio Trenton; and ccrtahi forms oceurring ill a liiiiiistono at Phillipsburgh, near Lake Chaiii[)huii, also regarded as of Trenton ago. In ISof), were first described, by James Hall, the gra[)t()lites of Pointj Levis, then 8i)oken of by liim as coming from "near tlio suiiiniit of the Hndson-lliver group " ; to which horizon, consid- ered as that of the Lorainc shales, tliey were, on strati- grapiiical grounds, assigned by Logiin. § 104. As early as 1840, however, as we liave seen, Emmons had, on stratigraphical grounds, assigned this Graywacke-belt to the horizon of tlie Calciferons Sand- rock, and had dechired it to contain certain peculiar rnnus of graptolites and of trilobit^c-s. This view, which was ussen- tially a return to that of Eatin, was, however, combated by all the other Americans geologists who had studied these rocks in Canada and in Vermont; C. li. Adams, W. B. Rogers, and W. E. Logan uniting, on alleged structural grounds, to place these rocks at the summit of the Champlain division, or in the Second, instead of the First Graywacke. § 105. Tt was not until 1856 that the present writer discovered in association with the graptolitic beds (-f Pointe Levis, limestones containing a hitherto unobserved trilobitic fauna, the examination of which by Billings led him to the conclusion that the strata in question were older and not younger than the Trenton limestone ; or, in other words, that they belonged to the First and not to the Second Graywacke. It was in 1861 that Logan, in a letter to Barrande, published this conclusion, then reached, and at the same time admitted the correctness of the later view of Emmons, for which this geologist had contended alone during fifteen years, namely, — that the belt of disturbed rocks which in Canada and in Vermont had been called the Hudson-River group, was in reality the stratigraphical equivalent of the lower members of the Champlain division, and older than the Trenton lime- i m 596 THE TACONIC QUESTION IN GEOLOGY. [xr. stones. These strata were the Upper Taconic of Emmons, which he had already in 1860 declared to be the equiva- lent of the rocks holding the first or Primordial fauna of Barrande. (§ 17, and ante, p. 586.) § 106. The contact of these rocks near Quebec with the underlying gneiss is concealed by the horizontal Trenton limestone of the region. The green sandstone of Sillery here lies upon the other members of the Graywacke series, and since this had been regarded as the Oneida sandstone, overlying Loraine shales, the whole series was supposed to be in its natural order of succession. Hence it was that, while admitting the change of horizon f these rocks from above to below the Trenton limestone, the Sillery sandstone, as it was henceforth called, was placed at the summit, and the limestones and graptolitic si ites of Pointe Levis, to which the name of the Levis division was given, at the base of the series ; an interme- diate portion receiving the name of the Lauzon division. The real order, however, as described both by Emmons in Vermont, and by Murray in Newfoundland was the reverse of this, and the Sillery sandstone was, in truth, the oldest member of the series here displayed. Logan, as we have seen, maintained that the typical section of southeast- ward-dipping strata at Quebec, estimated by him to measure 7000 feet, was the southeast side of an eroded anticlinal, and represented the rocks of his Quebec group in their natural order ; the Levis division at the base and the Sillery at the summit. I have long since endeavored to show, alike on structural and on paleontological grounds, that this view is erroneous, and that we have here an inverted succession. The true position of tlie Sillery is at the base of the series, and we here find exposed the eroded surface of the northwest side of an overturned anticlinal, by which the Sillery sandstone is made to overlie the younger members of the Graywacke series.* The succession is thus brought into harmony » Harper's Annual Record for 1876, p. xcviii., and for 1877, p. 167. XI.] UPPER TACONIC IN CANADA. 697 3GY. ^"^• ic of Emmons, be the equiva- ordial fauna of Quebec with the Lzontal Trenton stone of SiUery the Graywacke I as the Oneida whole series was jcession. Hence re of horizon i 'enton limestone, ^orth called, was es and graptolitic ame of the Levis n-ies; an interme- . Lauzon division, pth by Tilmmons m ,nd was the reverse ,u truth, the oldest Logan, as we have ,tion of southeast- Tiated by him to side of an eroded f his Quebec gvouv ,ion at the base and ig since endeavorec on paleontologicul and that we have :rue position of the , and we here find northwest side of an SiUery sandstone is rs of the Graywacke •ought into harmony ,11., and for 1877, P- 16^' V. ith tliat determined by Eaton, and by Emmons, in many sections farther south. This series, which had been previously called the Hudson-River group, was now by- Logan, and the Canada geological survey, named the Quebec group, and was described as a great development of strata between the Trenton limestone and the Potsdam sandstone ; which latter, Logan conceived to be repre- sented by certain black shales, that in several localities appear to pass beneath the Levis division. The rocks of this series were now, by Logan and his assistants, traced down the St. Lawrence to Newfoundland, on the one hand, and to the valley of Lake Champlain, on the other; where, however, the Red Sand-rock was supposed to represent the Potsdam. The history of these investiga- tions I have elsewhere set forth, in "Azoic Rocks," pp. 81-125. It should here be said that this view, which made the Sillery the youngest member of the series, was, in 1862, questioned by Billings, who inclined, with the writer, to place it at the base of the series ; while its evident basal pos^'^tion in Newfoundland led Logan also to express doubts, and to look upon the order assumed by him as simply provisional.* § 107. It remained, however, to determine how far this identification with the First Graywacke applied to the rocks farther south, in the valley of the Hudson, to which the name of the Hudson-River group had first been given ; and which had been declared, alike by Eaton, Emmons, and Mather, to be geographically and stratigraph- ically identical with the similar rocks in Vermont and in Canada. These rocks in the Hudson valley had been by Mather assigned to the horizon of the Second Gray- wacke, and from the occurrence in portions of them of the fauna of the Loraine or Pulaski shales, he, with Vanuxem, had, as we have seen, been led to emi)loy the names of Hudson slates and Hudson-River group as synonymous * Billings, Paleozoic Fossils, 1865. p. 69, and Logan, Geology of Canada, 180:1. p. 880. 1'# \:\ 698 THE TACONIC QUESTION IN GEOLOGY. [XI. J I with that of Loraine shales. The opposition between the view of Mather, on the one hand, and that of Emmons, now adopted by Logan, on the other, as to the horizon of the so-called Hudson-River group, was thus radical and complete. § 108. A question here arises whether it might not be possible to reconcile these two seemingly contradictory views by showing the belt of disturbed strata in question to include both the First and the Second Graywacke. These, as we have seen, were declared to resemble each other so closely as to be scarcely distinguishable save by the fact that the latter overlies the Trenton limestone (§ 7). If now, from any cause, this limestone should be absent, or should not appear in its usual character, it might vc r> well happen that the Second should appear to succeed directly the First Graywacke. That such a condition of things occurs in the disturbed region east of the Hudson, had already been affirmed by Emmons in 1846. As I have elsewhere pointed out (" Azoic Rocks," page 49), he then asserted the existence in ihis region of three distinct series of rocks: I. The Lower Taconic or Taconian limestone and slates. XL The First Graywacke, or Upper Taconic, resting in apparent unconformity upon the former, and itself partially eroded before the deposi- tion of III., which latter consists of shales and sandstones belonging to the upper portions of the Chumplain division, or the Second Graywacke, and rests unconformably, in many localities east of the Hudson, both on I. and II. ; having itself been subseqv.ently disturbed and eroded. § 109. These observations accord with many others to show the existence of ut least two important stratigraphieal breaks, with unconformity, in this eastern region : the first between the Taconic and the First Graywacke, already pointed out by Eaton, and the second at the summit of the same Graywacke series; thereby dividing, in this eastein region, the Champlain division into two distinct periods, the second one of which began with the depo- XL] UPPER TACONIC IN CANADA. 599 sition of tlie Trenton, or, rather, with tliat of the immedi- jitely subjacent Chazy limestone. The Laurentian regions of the Adirondacks and the Laurentides were not, at this early time, as has been so often said, the nucleus of a growing continent, but higher portions of a subsiding one. Upon its ancient gneiss we find reposing directly, in different localities, the Potsdam, the Calciferous, the Chazy, and the Tientou subdivisions. The dei)osition of the Trenton marks a, time of subsi- dence, daring which, along the Laurentides, the sea extended far and wide to the northward, and the marine limestones of the Trenton, overlapping the lower members of the Cham plain division, were laid down (in the regions to the north of Lake Ontario and of the lower St. Lawrence, as far eastward as the basin of Lake St. John on the Saguenay), directly on the submerged primary or eozoic rocks. After this period, and before the deposition of the succeeding mechanical sediments, extensive move- ments took place in the region of the Ottawa and Cham- plain valleys, and still farther south, which serve to throw light upon the problem before us. § 110. A striking illustration of this disturbance is shown on Logan's larger geolcg'cal map of Canada, in 1866, where, immediately south and east of the city of Ottawa, appears an outlier of Utica slate, overlaid by gray calcareous sandstone, holding the fossil remains of the Loraine, and associated with red slates ; the two possi- bly representing the Oneida and the Medina. This out- lier, with a length of about twenty miles from east to west, reposes transgressively alike upon the Trenton, Chazy, and Calciferous subdivisions. All three of these, witii a slight eastward dip towards the centre of the Ottawa basin, appear successively, in passing from west to east along the southern border of this unconformably overlying area of newer strata, which are here preserved by having been let down along the north side of an east and west dislocation ; thus testifying to a former exten- V :i,wrrTiiiii 600 THE TACONIG (iUESTION IN GEOLOGY. [XI. ,r''^ii! iKS sion of the Second Graywacke over this area, where it lies un conformably, not only on the Calciferous Sand-rock (the representative of the First Graywacke), but on the Trenton limestone itself. For farther references to the '.le tails of tliis region, which was carefully mapped by Logan, see "Azoic Rocks," page 50. We have here, in tlie valley of the Ottawa, evidences of the same conditions as v/ere described by Emmons in that of the Hudson; namely, the unconformable superposition of the upper members of the Champlain division upon the lower ones ; a break occurring at the summit of the Trenton. It is impossible not to connect "these conditions in tlje Ottawa and Hudson valleys with those already noticed in eastern Pennsylvania, and in Orange County, New York, where the Oneida sandstone, which, as we have seen, is continu- ous with the upper part of tlie Loraine shales, is found to rest unconformably upon thu strata of the First Gray- wacke. § 111. Considerable movements are thus seen to have marked both the beginning and the close of the Chazy- Trenton period, and it is evident that the absence, in any district, of the characteristic limestone of this time, be- tween the First and Second Graywackes, might result either from non-deposition or from erosion. Evidence of the latter is afforded in the area just described in the Ottawa basin ; while, at the same time, there is not wanting evidence that this limestone-mass, so well marked by its thickness, and the persistence of its lithological character over great areas in eastern North America, else wi ere thins out, and either disappears entirely, or loses its ordinary lithological characters. Thus while in Canada, at points as widely separated as Beauport, Mont- real, Ottawa, Lake Simcoe, and the shores of Lake Huron, it appears with a thickness of from 600 to 750 feet (being everywhere followed by the Utica and Loraine sluiles), it is in Lewis County, New York, diminished to 300 feet, at Trenton Falls to 100 feet, and, it is said, to thirty feet in GY. t**' , wliere it lies us Sand-rock ), but on tlie srences to tlie y mapped by have here, in Line conditions the Hudson; of the upper he lower ones ; rrenton. It is in the Ottawa aced in eastern w York, where een, is continu- ,les, is found to ;he First Gray- ius seen to have e of the Chazy- absence, in any )f this time, be- js, migbt result n. Evidence of described in the le, there is not ;, so well marked ■ its lithological North America, lears entirely, or Thus while in Boauport, Mont- s of Lake Huron, o 760 feet (being voraine shales), it ed to 300 feet, at ., to thirty feet in XI.] UPPER TACONIC IN CANADA. 601 the Mohawk valley ; tliinning out and disappearing to the southeast, according to Conrad; but, as will subsequently ap£/ear, probably represented along this eastern border by argillaceous beds, which, but for their organic remains, would not be recognized as of Trenton age. § 112. The bearing of the paleontological investigations made by the geological survey of Canada on the question of the age of the eastern Graywacke, or so-called Hudson- River group of rocks, v as discussed by James Hall, in a note to his "Geology of Wisconsin," in 1862 (page 443). He there alluded to the evidence furnished by organic remains found in the Hudson-River slates in Vermont and Canada, "which prove conclusively that these slates are to a great extent of older date than the Trenton lime- stone," though probably newer than the Potsdam. He moreover remarked that " the occurrence of well known forms of the second fauna ... in intimate relation with, and in beds apparently constituting a part of the serie?, along the Hudson River, requires some explanation. Looking critically at the localities in the Hudson valley wliicli yield the fossils, we find them of limited and of almost insignificant extent. Some of tliem are on the summits of elevations, which are synclinal axes . . . where the remains of newer formations would naturally occur. Others are apparently unconformable to the rocks below, or are entangled in the folds of the strata . . . while the enormous thickness of beds exposed is almost desti- tute of fossils." In view of all these facts, Hall, while still retaining the name of Hudson-River group as the desig- nation of the fossiliferous strata which elsewhere are found to occupy a horizon between the Utica slate and the Oneida sandstone (otherwise called Pulaski and Loraine shales), concludes that the name of Hudson-River group cannot properly be extended to the great mass of strata which have hitherto borne that name, and which, according to him, "are separated from the Hudson-River group proper by a fault not yet fully ascertained." .«i <*« Vli.';!:!i; ml n ';" g.!^iiii» 602 THE TACONIC QUESTION IN GEOLOGY. [XI. § 113. It should, however, be remembered that al- though the Hudson-River group was, through the paleon- tological publications of the New York survey, identified with the Loraine shales only, the name, as at first given by Vanuxem, was made to include two divisions, the lower of whicli, as he showed, was distinct from the upper, as appeared by its different geographicid distribution. That these two divisions of the Hudson-River group were supposed by him to be associated with a still older series, lithologically resemblijig them, would appear from Van- uxem's language, when lie wrote of "the difficulty of separating or distinguishing the slaty and schistose mem- bers of the Hudson-River group from those of greater age, with which, along their eastern border, the two (sic) are more or less really or apparently blended." § 114. Hall, while thus admitting the existence of an apparent unconformability between the older and the newer fossiliferous rocks in this disturbed region, fell back on Logan's explanation, and imagined the juxtapo- sition of the two series to be effected by a break of the strata, with an uplift on the eastern side, by which the rocks of pre-Trenton age were brought up, and were sometimes found in contact with the Trenton or Utica divisions, at others with the Loraine, and, perhaps, even with the still higher beds of the Oneida. I have else- where discussed ac length this hypothesis of a single great fault, with an upthrow of 7000 feet, imagined by Logan to extend from Alabama to the northeast extremity of the continent, in Gaspd ; and having shown its great im- probability both geographically and stratigraphically, have maintained, for ten years past, the simpler explanation of an unconformity between the First Graywacke and the succeeding members of the paleozoic series. ("Azoic Rocks," pp. 121-125.) Evidences are there given that movements of the earth's crust in these regions immedi- ately precede! the Trenton age, and that u[)on the folded, eroded, and submerged strata of the First Graywacke, as XI.] UPPER TACONIC I" CANADA. 603 upon the Taconiau and still elder series, there wero sub- sequently deposited the Trenton limestones. Where these limestones were afterwards removed by denudation, or where, to the eastward, they thin out and disappear, we may expect to find the Loraine, or the succeeding Oneida strata, in direct superposition upon these older rocks. § 115. In 1863, Logan, having followed southward into Vermont the Graywacke-belt, to which he had then given the name of the Quebec group, proceeded, in company with James Hall, to examine the same rocks in eastern New York ; where they were now described by him, as they had been by Emmons, as sandstones and conglomerates, generally with argillites, sometimes red and green, and with limestones, often schistose or concretionary, includ- ing the Sparry Lime-rock of Eaton. All of these were now declared by Logan to belong to the Quebec group, which was said by him to occupy nearl_y the whole of Columbia, Rensselaer, and Dutchess Counties ; the Sillery division being largely displayed in the first-named of these, but scarcely appearing south of it. To the westward, in approaching the River Hudson, these rocks were declared to be replaced by lithologically distinct and more recent strata, referred to the Loraine shales ; a narrow belt of which was traced along the east side of the river to a point a little above Hyde Park, where the boundary of the two divisions crosses to the west bank. The strata on both shores from thence down to the gneiss of the High- lands were referred by Logan to the Quebec group. § 116. Logan, however, as we have seen, assumed the Sillery sandstone to be the sunnnit instead of the base of the First Graywacke ; and when he became, at this time, acquainted with the underlying Taconiau marbles in Vermont, and, farther southward, imagined them to be his Levis and Lauzon divisions in 'an altered condition, and thus described them as members of the Quebec group. It yet remains to determine in this rej^ion the limits be- tween the Taconic and the First Graywacke. We now Mi .111 mm 604 THE TACONIC QUESTION IS GEOLOGY, [xr. know, moreover, from the discoveries of Dale, Dwight, and others, that still newer fossiliferous strata, of Ordo- vician age, are also included in this part of the Hudson valley ; and that we have, in fact, in this region, the three groups of rocks long since pointed out by Emmons. (§ 108.) The testimonj"^ of Logan is valuable as confirming that of Emmons, and of Hall, as to the existence of por- tions of the Second Graywacke series resting, not upon the Trenton limestone, but upon the older schistose rocks of the region ; and moreover, as showing the superposi- tion of the First Graywficke to the Taconian. § 117. The apparent absence of the characteristic lime- stone of the Trenton from the base of the Second Gray- wacke in this region may be due to a stratigraphical break and erosion at the close of the Trenton period, as we have seen in the Ottawa basin. Two other explana- tions are suggested by the thinning-out of the limestone- mass to the southeast, as already noticed (§ 111) ; one, that the region was beyond the Trenton sea, and the other that the sediments of this sea over the area in question were argillaceous beds, resembling rather the succeeding shales than the limestone deposited elsewhere. That this latter was the case in parts of the eastern region, will be shown in the sequel, but we shall there also find many evidences of movements in paleozoic times, subsequent to the depo- sition of the Trenton. I have elsewhere pointed out ("Azoic Rocks," page 123), besides the post-Trenton break in the Ottawa basin, the evidence 1.1 eastern North America of not less than five epochs, marked by movements of the strata, and by unconformities, subsequent to the deposition of the Tren- ton and Utiea divisions. Of these the earliest, and the only one which now concerns us, is that of which we see evidences in the unconformable superposition of Silurian beds over older strata to the north and east of the Hud- son valley. On St. Helen's Island, near Montreal, we find reposing on the eroded surface of the slightly inclined 00 Y. ^•"'• Dale, Dwiglit, ,trata, of Ordo- of the Hudson egion, the three t by Emmons. ,le as confirming xistence of p^^- esting, not upon r schistose rocks g the superposi- iiian. laracteristic lime- he Second Gray- a stratigraphical :renton period, as wo other explana- of the limestone- (§ 111) ; one, that and the other tbat V in question were 1 succeeding shales That this latter :"ion, will be shown nd many evidences ;quent to the depo- ,zoic Rocks," page 1 the Ottawa basin, ja of not less than i the strata, and by osition of the Tren- he earliest, and the hat of which we see •position of Silurian ,nd east of the Hud- near Montreal, we the slightly inclined XL] UPPEll TACONIG IX (JANADA. G05 Utica slates, a portion of fossiliferuus limestone associated with a doloniitic conglomerate. The fauna in the lime- stone is referred to the age of the Lower Helderberg; while the acc(jmpanying conglomerate contains forms which belong rather to the CI niton and Niagara divisions of the Silurian, and liolds at the same time pebbles of fossiliferous Trenton limestone, with others apparently of Potsdam sandstone, Utica shite, and red sandstone and shale, resembling those of the Medina ; the whole mingled with pebbles of Laurentian gneiss, and of igneous rocks, giving evidence of a period of disturbance, and considera- ble erosion of the older rocks. Other masses of similar conglomerate are found elsewhere in the vicinity, in ouo case holding Silurian fossils, resting on various members of the Champlain division, and on the Laurentian gneiss. Another mass of Lower Helderberg limestone is met with on the flanks of lieloeil Mountain, an eruptive mass which breaks through the Loraine shales in the Richelieu valley. In tlie distribution of these, and of similar areas of fossiliferous limestones, we have the evidences of a Silurian sea, which extended from the Helderberg region in New York, not only through the valleys of the Hudson and Lake Champlain to that of the St. Lawrence, but also through those of the Connecticut, the St. Fjancis and the Chaudidre, and thence to Gaspd ; depositing its sediments, with their characteristic fauna, unconformably over rocks of very different ages. We have similar evidence that the Chazy-Loraine or Ordovician sea had already, in like manner, extended over parts of this region, leaving its fossiliferous sediments spread unconformably over Cam- brian and pre-Cambrian strata. § 118. A section from Crown Point, New York, acrcKss the southern part of Lake Champlain eastward to Brid- port, Vermont, which was studied in detail by Wing and by Billings, presents, in its western portion, the whole succession of the Champlain subdivisions, from the Pots- dam to the Loraine shales. Farther eastward on this line, ■i '^■■l.i'^ !l t^ ■ :l 606 THE TACONIC QUESTION IN GEOLOGY. [Xl. '! I'' a great dislocation brings up the Red Sand-roclc, with Olenellus, causing it to overlie, in seeming conformity, the Loraine shales. This sand-rock is followed to the eastward by limestones holding the fauna of the Calci- ferous Sand-rock, with other forms like those of the Levis limestone. To these succeed other limestones, with an abundani Trenton fauna, interrupted by a second fault, which again brings up the Levis beds; the Sillery sand- stone being unrepresented, unless possibly by the Red Sand-rock to the we?t. For a summary of the observa- tions on this section, and reference to the original paper, see " Azoic Rocks," page 119. § 119. Other examples of an extension of the Ordo- vician sediments eastward are found in the province of Quebec, in various localities along the disturbed region northeastward from Lake Champlain. Lying alike among the uncrystalline strata of the Graywacke series, and the older crystalline rocks to the southeast of them, there are met with, in many localities, carbonaceous shaly beds, more or less calcareous, containing organic remains of Ordovician age. These strata, probably never very con- siderable in amount, have, however, rarely escaped erosion, except in localities where, as the result of the folds and dislocations already noticed, they have been protected by the overlying or adjacent older strata, beneath which they often seem to pass with an eastward dip. As studied at Farnham, in the province of Quebec, they thus appeared to be more ancient than the Graywacke series, and were described by Logan as portions of Potsdam rocks under- lying the Quebec group. The black slates of this locality, however, contain, according to Billings, besides unde- scribed graptolites, Ptilodictya and trilobites of the genera Ampyx, Dalmanites, Lichas, Triarthrtis, and Agnostus; and were hence referred by him to the Trenton or the Utica division of the New York system. Similar black slates appear, in like manner, to pass beneath the crystal- line schists which lie to the east of the Graywacke-belt in .OQY. I**' land-rock, with ing conformity, ollowed to the la of the Calci- ose of the Levis stones, with an a second fault, the Sillery sand- bly by the Red J of the observa- e original paper, ion of the Ordo- i the province of disturbed region ,ying alike among ike series, and the )f them, there are seous shaly beds, •ganic remains of y never very con- y escaped erosion, t of the folds and been protected by ,a, beneath which ddip. As studied they thus appeared ke series, and were tsdam rocks under- tes of this locality, ngs, besides unde- jbites of the genera us, and Aonostus ; ae Trenton or the em. Similar black aeneath the crystal- Gray wacke-belt in XI.] UlTEU T A CONIC IN CANADA. GOT this region, and were by Logan adduced as proofs of the view then maintained by the geological survey of Canada, tliat the crystalline locks in (juestiou were nothing more than portions of this same Graywacke in an altered con- dition. The fallacy of this view I have long since shown, and have pointed out the nature of the stratigraphical acci- dents by which this seeming inversion has been brought dbout. Selvvyn, of the geological survey of Canada, lias more recently furnished additional facts regarding the distribution of these fossiliferous shales ; outliers of which have been observed at various localities in eastern Canada, among Ihe crystalline schists, especially along the west side of a line of fault, with an upthrow on the east side, extending through Stukley and Ely. Similar fossiliferous beds are found in Tingwick and Arthabaska, and also near Richmond ; where a narrow belt of black shales, with Triarthrns and other organic forms, is found lying to the east of the crystalline schists of the region. The latter are a second time brought up on the eastern border of these shales, and soon pass beneath the argillitea of the Windsor basin (§ 81). In this connection it may be noticed that Dodge has found, still farther to the east- ward, in Penobscot County, Maine, black shales holding graptolites, which are regarded by him as species belonging to the Utica slate.* § 120. It was said at the commencement of this essay that the Upper Taconic rocks have been known both as the Hudson-River group and the Quebec group. This statement we have justified in the preceding pages, and are now prepared to state succinctly what has been the precise meaning attached to these two terms, which have been so conspicuous in the history of American geology. The Hudson-River group, by the admission of Vanuxera, who first proposed it, was a composite one, devised t include two, if not three divisions of strata, in part of * Amer. Jour. Science, ISSl, vol. sxii., p. 434. -0' m .1' , ll; a '■ G08 THE TACONIC yUIiSTlON IN GliOLOCiV. m disputed age, but at tlie same time embracing in it.s wppev portion the Loraine shales. As tiiis was the only j)art of the group of vviucii the fauna was known, the name of Loraine shales, in paleontological language, soon came to be regarded as the equivalent of lludson-Uiver group; and thus the fact of its heterogeneous character, clearly stated by Vanuxem, was lost sight of. Meanwhile, the name of Hudson-River group was applied stratigraphi- cally to the whole of the First Graywacke of Eaton, with its succeeding Sparry Lime-rock. This is seen from the language of James Hall, who, in 1857, wrote of the grap- tolites found in slates with tlie limestone of Pointe Levis, at that time assigned by Logan to this horizon, that they are met witli in " that part of the Hudson-Kiver group which is sometimes designated as Eaton's Sparry lime- stone, — being near the summit of the group." This was the Levis limestone of Logan.* § 121. The Red Sand-rock of Vermont was also, at tlio same time, regarded as either forming a part of the same group or as closely related to it. Thus Hall, in describing, in 1859, the trilobites of the genus Olenellus found in shales intercalated in the Red Sand-rock in Georgia, Vermont, which he then referred to this horizon, wrote, " I have the testimony of Sir William Logan, that the shales of this locality are in the upper part of the Hudson-River group, or forming a part of a series of strata which he is inclined to rank as a distinct group, above the Hudson-River group proper." f We have farther to mention in this connection the notion of Mather, who supposed that the crystalline rocks of western New England, including the crystal- line limestones, "and probably the associated micaceous gneiss, mica-slate, hornblende-slate, and hornblende-rocks . . . are nothing more than the rocks of the Champlain division greatly modified by metaraorphic agency." This * Report Geol. Survey of Canada, 1857, p. 117. t Twelfth Ann. Kep. Regents of the University of New York, 1859 ; cited by Barrande, Amer. Jour. Scl. (2), xxxl,, p. 213. OCiV. ^'^• ng in itH upper he only pttit (if n, the luuno vl e, soon camo to u-lliver group; laracter, clearly Meanwhile, the ed stratigrapUi- a of Eaton, with ia seen from the vote of the grap- of Pointo Levis, lorizon, that they dson-lUver group >n'8 Sparry lime- le group." This nt was also, at the part of the same lall, in describing, lus found in shales Gleorgia, Vermont, vrote, " I liave the the shales of this tdson-River group, hich he is inclined udson-River group iu this connection hat the crystalline uding the crystal- sociated micaceous d hornblende-rocks of the Champhvin [lie agency." This iltyof New York, 1859 -, p. 213. \i. Ul'PKIl TACO.sin IN CANADA. GOD view was udoptotl by Logan, and the similar crystalline rocks of the (r recn-Mountuin bolt in Canada were described as belonging io the iiltered IIudsoti-Kiver groU[). § 122. The Quebec group, which, in 18(J1, succeeded to the Iludson-Uivt-r grouj), iidieritod its traditions, witli a few exceptions. Its horizon being now cluinged fr(»ni above to below the Trenton limestone, it could, of course, no longer include within its limits the fauna of the Ijoraine shales, belonging to the Second (xraywacko. The greater antic^uity of the fauiui of the Red Sand-rock of Wn-mont having in the meantime been recognized, these rocks were assigned, under the name of Potsdam, to a position be- neath the so-called Quebec group. To this lower hori- zon, moreover, Logan, at the same time, referred certain black slates in Canada, which, though apparently under- lying the Graywaoke series, have since been found of Ordovician age (§ 119). The Quebec group, as at first defined, was nothing more uor less than the First Graywacke of Eaton, with its overlying Sparry Lime-rock ; which latter is really an upper member of that Graywacke series, and was included with it by Emmons in his Upper Taconic division. Emmons now read aright the relations of these rocks, and saw that the sections in which the limestones ai)pear to underlie the massive green sandstones give an inverted succession. Logan, however, though recognizing therein the existence, in many cases, of overturned anticlinals, hiverted the whole series, and regarded the basal or Sillery sandstone as the highest member, while the Levis lime- stones were made the lowest. § 123. This erroneous view as to the succession of the strata at Quebec, at first declared by Logan to be merely provisional, was the more acceptable to him for the reason that it could be m.ade to accord with the hypothesis that the adjacent crystalline schists were, as Mather had taught, the altered equivalents of what was now called the Quebec group. When, as is sometimes the case, the Sillery sand- f i;ii' ■ ! 1 I i 'tl ^? I ''■■m.l \ik< 1*1 V ( 13 IT fiii.il 610 THE TACONIC QUESTION IN GEOLOGY. [XI. stone was found alone (as long before described by Em- mons), resting upon the crystalline schists, the higher and softer members of the Gray vvacke series having disappeared, Logan supposed that these schists were no other than shales of the Sillery (and Lauzon) in an altered and so- oulled metamorphic condition, — which, according to his view of the succession, should underlie these sandstones. Hence it was that the Huronian rocks of the Notre Dame range (the prolongation of the Green Mountains) were by Logan called "Altered Quabec group," long after it had been shown by the present writer that fragments of these same eozoic rocks occur in conglomerates with the fossil iferous strata of the Levis division near Quebec. § 124. In like manner, when the Sillery sandstone was found, farther southward along the Graywacke-belt, to rest upon the Taconian marbles and slates, these were by Logan declared to be limestones and shales of the Levis division in an altered condition (§ 116). But this was not all : as the Levis beds, sometimes through inverted faults, and sometimes through dislocations, came to be placed beneath the Sillery, so the black Ordovician slates, whether in direct contact with the Cambrian or with the older rocks, were, as the result of similar accidents, made to underlie the more ancient groups of strata, and were believed by Logan to be older than these. In either case, his argument was the same : in the former, these Ordo- vician strata were Potsdam beds passing beneath the un- altered Quebec group ; and in the latter, they were the same beds underlying the altered strata of the same Quebec group. § 125. To complete this history, we must recall the fact that, not conicut with including in the newly organ- ized Quebec group, besides the Cambriiin Gray wacke with its limestones, the Taconian and the Huronian of the Atlantic belt, Logan proposed to extend it to Lake Su})e- rior. Assuming that the horizontal sandstones there over- lying unconformably the Keweenian or Copper-bearing ay. LXI. XL] THE KEWEENIAN SERIES. Gil •ibed by Ern- ie higber and T disappeared, ^ other than ,tered and so- jording to bis se sandstones, le Notre Dame .untains) were " long after it it fragments of erates with the jar Quebec. J sandstone was tywacke-belt, to J, these were by ,les of the Levis ,, But this was lirough inverted ons, came to be )rdovician slates, irian or with the accidents, made strata, and were In either case, ■mer, these Ordo- g beneath the uu- [er, they were the £ the same Quebec must recall the I the newly organ- |n Gray wacke with Huronian of the id it to Lake Supe- [dstones there over- lor Copper-bearuig series, were of the age of the Chazy or St. Peter's sand- stone of the upper Mississippi, Logan was led, in 1861, to assign tlie whole of this series of 20,000 feet or moie to the Quebec group, and thus to give it a position above the horizon of the fossiliferous Potsdam sandstone of Wis- consin and Minnesota ; which, as seen on the St. Croix River, and elsewhere in that region, is well known to overlie the Keweeniau unconformably, and is probably separated from it by a great interval of time. This view will be found represented on Logan's small map of Canada, dated 1864, and also in his larger map of 1866. The great Animilde or Tac(jniau series, — the relations of which in this region we have considered in §§ 89-90, and which had been previously described by Logan as the lower division of his Upper Copper-bearing series, — was not distinguished from it on the maps in question, but was now supposed to represent the Potsdam. § 125 A. [As to the history of our knowledge of the Upper Copper-bearing or Keweenian series of Lake Su- perior, it was by Houghton, in 1841, regarded as more ancient than the Potsdam, and by Logan, in 1846, as in- ferior to the horizontal sandstones of Sault Ste. Marie, then supposed by him to be Potsdam. Logan, in the report of the geological survey of Canada for that year, included, under the name of "Volcanic formations," two divisions, a lower one of dark-colored argillites and quartzites, which is seen in a nearly horizontal attitude on Thun- der Bay (where it was afterwards called the Animikie series by the present writer), and an upper division of sandstones, amygda^oids, and trappean rocks, regarded by Logan as equivalent to the series bearing native copper on the south shore, and elsewhere on the lake. Beneath this lower division on Thunder Bay were older crystalline rocks, then described as greenstones with epidotic rocks and chloritic slates, and noticed by Logan, in his report for 1846, as the "chloritic schists at the suuiniit of the older rocks upon which the Volcanic formations rest un- ^:% ,;'ii V' 41 j^ v,l rmff ]■ " ' , 1 \l"- 1 i^i ■ ' ■ ''e. 1 ■ ' I^i li ^1 1^1 m ^ WmA m : ■SS^^t. . ^ ^ ■-'■ n i ((■'' i ' ' i I'J".' T' I ('.I ' i ' 1 !-■■ 1 1! ri- j| if- ll ;fd E M im : , y mmE P4 THE TACONIC QUESTION IN GEOLOGY. [XI. aie by Wulcott iuclucled in his middle and upper divis- ions. Tlie first of these is the so-called Lower Potsdam of Billings, which is traced from the strait of Bellisle along the valleys of the St. Lawrence, Lake Cham- plain, and the Hudson, and thence southward along the great Appalachian valley, embracing a large part of tlie Upper Taconic of Emmons. The upper division of the Cambrian includes the typical Potsdam of the Adir- ondack area and of the upper Mississippi valley, in both of which regions the Middle Cambrian is unknown. The Middle and Upper Cambrian appear together in the sec- tions in Utah and Nevada, where the latter is succeeded, as in the more eastern areas, by rocks carrying the second fauna of Barrande, which Walcott, following Lapworth and the present writer, designates as Ordovician (§ 17). This, it will be remembered, is synonymous with the Lower Silurian of Murchison, and with the Upper Cam- brian of Sedgwick, which is thus distinct from and supe- rior to the Upper Cambrian of Walcott. § 137 B. [Ill the Grand Canon of the Colorado River 1000 feet of Upper Cambrian ov Potsdam strata, locally known as the Tonto group, rest uuconformably upon a great body of strata described by Powell as the Grand Canon series, divided by him into the Grand Canon and Chuar groups, and having an observed thickness of about 13,000 feet.* In like manner, a series of Cambrian sand- stones and limestones, about 900 feet in thickness, closely resembling those of the Tonto group, and affording a abundant Potsdam fauna, already made known by Shu- mard, occur in central Texas, where they are overlaid by Ordovician. Here, as in the Grand Canon of the Colo- rado, the Upper Cambrian strata rest unconfcjrniably upon a series of uncrystalline sandstones, shales, and lime- stones, several thousand feet in thickness, which are well seen in Llano county, and have -n by Walcott called the Llano group. These, according to him were pene- , * Powell, Amor. Jour. Science, 1883, xxvi., 437. 3GY. ^^'• i upper divis- ower Potsdam lit of BelUsle , Lake Cham- ard along the i-ge part of the )QV division of im of the Adir- valley, in both unknown. The ither in the sec- ter is succeeded, L-ying the second >wing Lapwovth dovician (§ l'^)- lymous with the the Upper Cam- it from and supe- le Colorado Hivev am strata, locally ..ably upon a great Ithe Grand Canon "anon and Chuar clcness of about ,f Cambrian sand- thickness, closely I, and affording a .e known by Shu- jy are overlaid by ^anon of the Colo- ^t unconformaijly E, shales, and lime- iss, which are well Iby Walcott called him were pene- Lxvi., 431. XI.] THE TACONIC HISTORY KEVIEVVED. 625 trated by granites befoie the deposition of the Pots- dam.* This great series of uncrystalline sediments, found alike in the Grand CaSon and in Texas, is by Walcott compared with the Keweenian series of Lake Superior, and regarded as belonging to a Keweenian area of conti- nental extent, over the upturned and eroded edges of which the Cambrian was laid down alike in Michigan and Minnesota, in Texas and in Arizona. Some evidences of organic remains have been observed by Walcott in these lower rocks in the Grand CaRon, and we have elsewhere noticed such evidences in tlie Keweenian of Lake Superior (ante, page 615). For the present, we agree with Pow- ell and with Walcott in regarding these lower rocks provisionally as pre-Cambrian.] f § 138. A similar great development of Cambrian rocks exists in mnthwestern Newfoundland, where, from his studies of their organic remains, the late Mr. Billings was led to admit a succession of over 9000 feet of paleozoic strata below the Trenton horizon. The subdivisions there recognized by him, in ascending order, were : 1. Lower Pots- dam ; 2. Upper Potsdam; 3. Lower Calciferous; 4. Upper Calciferous ; 5. Levis ; and 6. Phyllograptus beds. The second and third of these were regarded by Billings as the representatives of the Adirondack Potsdam and Calcifer- ous, while the Phyllograptus beds at the summit were con- sidered the equivalent of the Welsh Arenig, which belongs to the base of the Bala group, or the second fauna. It is evident, as Billings declared, that we have, in this great thickness in northwestern Newfoundland, a nuich more complete sequence than in the Adirondack region, where the Upper Potsdam, Calciferous, and Chazy subdivisions represent the whole succession from the ancient gneiss up to the Trenton limestone. • Walcott, ibid., xxviii., 431. t For these geneiiilizations by Walcott as to the American Cambrian, I am Indebted to a yet unpubiisherl paper read by him before the National Academy of Scioiirps at Washington, April 23^ 1880, and to his private communicatious. ¥ I. M i \\ i w .t^ i ,«) II . j i I ' i 626 THE TACONIC QUESTION IN GEOLOGY. § 139. Keeping in view the great development of the Cambrian alike in the Cordilleras and in Newfoundland, as compared with the Cambrian of the Adirondack and Mississippi areas, we are better prepared to understand the remarkable type assumed by this series in the Appalachian area, on the eastern margin of the American paleozoic basin, from near the Gulf of Mexico northeastward to the Gulf of St. Lawrence, and to Newfoundland, along the west- ern base of the Atlantic or Appalachian belt. These Cam- brian rocks throughout this extent, wherever preserved, are characterized by great thickness and considerable diversities in composition, due to the acfumulation of mechanical sediments <'"rived from tlie di. integration and decay of the various groups of pre-Cuinbrian rocks which made up the .-djacent eozoic land. To this, and to re- peated nujvemeuts of the land during and jifter the Cam- brian period, they owe their complex constitution, their great volume, tlieir disturbed and faulted condition, and tlieir unconformities. All of these characters serve to distinguish them widely from the horizontal and compara- tively thin quartzites and magnesian limestones, their representatives along the northern border of the great basin, as seen in the Adirondack and Mississippi areas. It is this Appalachian Cambrian, many thousand feet in thickness, which, as we have already seen, constitutes the First Graywacke and the Sparry Lime-rock of Eaton, the Upper Taconic of Emmons, the Quebec and Potsdam groups of Logan, and a large part of the original Hudson- River group. § 140. That the Levis limestones and Phyllograptus shales, found at the summit of this series, mark the begin- nings of the second fauna has already been noticed, as well as the fact that still higher strata, of Ordovician and Silurian ages, are found over portions of this Appalachian Cambrian series, among the strata of which they have sometimes been involved by subsequent movements. It will also be borne in mind : first, that this great mass of : w '4 3Y. IXI. XI.] THE TACONIC HISTORY REVIEWED. G27 praent of the ewfoundland, irondack and iiderstand the 3 Appalachian can paleozoic astward to the along the west- . ;. These Cam- (ver preserved, \ considerable 3cumulation of .utegratiou and tan rocks which this, and to re- [ i,i'ter theCam- nstitution, their i condition, and racters serve to tal and compara- limestones, their Lcr of the gveat ,Iississippi areas, thousand feet in a, constitutes the ,ck of Eaton, the |ec and Potsdam original Hudson- id Phyllograptus mark the begin- been noticed, as )f Ordovician and this Appalachian which they have it movements. It this great mass of 10,000 feet or more of diversified and folded Cambrian strata is exchanged in the Adirondack and Mississippi areas for a far more simple type of horizontal strata, but a few hundred feet in thickness; and, secondly, that ero- sion has removed this great series wholly or in part from over large portions of its original area. § 141. With these explanations before us, we are now prepared to consider the relations of the Cambrian and Ordovician series, in their two unlike types of the Appa- lachian and Adirondack areas, to the Lower Taconic limestones. It has already been shown that Emmons, in 1842, in his final report on the geology of the Northern district of New York, defined, with the present names, the lower subdivisions of the New York paleozoic system, from the Potsdam to the Oneida sandstone, both inclusive, to which he gave the collective appellation of the Cham- plain division. [He at the same time proposed for the Primitive Quartz-rock, the Primitive Lime-rock, and the Tran- sition Argillite of Eaton, together with the First or Transition Gray wacke — called by Emmons the Taconic slates — and the Sparry Lime-rock of Eaton, the general name of the Taconic system. The Taconic slates were then described by him as a great mass of argillites with interbedded limestones and coarse sandstones, limited on the east, in his original section, by the Sparry Lime-rock at the base of the Taconic hills, and on the west by " the Loraine or Hudson-River shales," bv which the Taconic slates were declared to be undoubtedly overlapped, al- though the line of junction on the west was said to be ob- scure. This intermediate mass, whose limits were thus clearly defined to the west of the Taconic hills in 1842, was farther said in 1846 to have an immense thickness, and, in the typical section in Rensselaer County, a breadth of not less than twenty miles. [All of these divisions from the Primitive Quartz-rock of Eaton to the Sparry Lime-rock, both included, were, I. ;■ t. I' ■ ; 1 f>' n '^i.. i! i I I. i,(!Mti i rb 0J8 THK TACONIC QUESTION IN GEOLOGY. vm by Emmons, in 1842, included in what he called the Taconic system, and described as "the rocks lying be- tween the upper members of the Champlain group and the Jloosic Mountain." They were then regarded " as inferior to the Potsdam sandstone, or as having been deposited at an earlier date than the lowest members of the New York Transition system." The precise relation of this system to the Silurian and Cambrian systems, and, indeed, the limits of these in England, were not at that date clearly defined, but Emmons then supposed that the Taconic rocks in part might "be equivalent to the Lower Cambrian of Sedg- wick," — "the upper portion being the lower part of the Silurian system,"* to which the Middle and Upper Cambrian of Sedgwick were then, on the authority of Muxchison, very generally referred.] § 142. In 1843 appeared the final report by Mather upon the geology of the Southern district of New York, in which he rejected entirely the notion of the Taconic system, and the whole teaching of Eaton, asserting that the Taconic was nothing more than a in<';Ii'ied funn of the Champlain division of Emmons. The Granular Quartz-rock of the Taconic ''c declared to be Potsdam; the Granular L^ne-rock, xiie ( ilciferous Sand-rock with the succeeding Chazy ami Ti_ ..ton limestones; while the overlying strata, including the Taconic slates or First Graywacke, were the Utica and Loraine shales. A simi- lar suggestion had been put forth by Messrs. H. D. and W. B. Rogers, in 1841, for the like rocks in New Jersey and Pennsylvania, and was cited by Mather in support of his view. When, later, in 1858, H. D. Rogers publislied his final report on the geology of Pennsylvania, the Lower Taconic rocks of Massachusetts had been by Emmons traced south westward through Pennsylvania, in the great Appalachian valley, and the adjacent and subordinate Lancaster valley. These rocks, under the names of * Eraiu'^ns, Geology of the Northern District of New Yorlc, 1842, pp. 1-iO, 144, 163. XL] lY. [XI. THE TACONIC HISTORY REVIEWED. 629 e called the ks lying be- yroup and the d " as inferior a deposited at the New York this system to Loed, the limits ilearly defined, Lc rocks in part brian of Sedg- lower part of .die and Uppe^' .16 authority of port by Mather t of New York, of the Taconic 1, asserting that o-Jl'-led form of The Graiiular to be Potsdam; Sand-rock with [tones; while the slates or First shales. A simi- ^essrs. H. D. ana n New Jersey and in support of his ers published lus vania, the Lower ,een by Emmons ania, in the great and subordinate [er the names of If New York, 1842, pp. Primal, Auroral, and Matinal, were now described by H. D. Rogers as local modifications of the Champlain series, — the great Auroral limestone being assumed to be the representative of the Calciferous, the Chazy, and the so- called Birdseye and Black-River subdivisions, while the Matinul slates were supposed to represent the upper part of the Trenton, with the Utica and the Loraine shales. For many details with regard to the facts noticed in this para- graph, and for other points in the Taconic history, the reader is referred to the author's volume on "Azoic Rocks." 8ee also ante, pp. 533-535. § 143. Coupled with this , . ow of Mather was that of a progressive alteration of these uncrystalline rocks of the Champlain division, supposed to be traced through the Taconic strata into the crystalline schists of western New England, designated by Mather as Metamorphic rocks; between which and the Taconic, it was said by him, "no well marked line of distinction can be drawn, as they blend into each other by insensible shades of difference." He was at length led to extend this same hypothesis to the more massive gneisses and crystalline limestones of southern New York, and to conclude that these also were, wholly or in great part, but altered rocks of the Cham- plain division, — a notion which has lately found an advo- cate in Dana, who has also revived Mather's view of the Champlain age of the Taconic quartz-rock and granular limestone, as will be noticed farther on. § 144. [As we have already seen, Eaton had long before announced the existence of a stratigraphic break at the base of his First Graywacke, — being the great group of strata called by Emmons the Taconic slatt in 1842, when he already recognized its distinctness iu)m the underlying portions of his Taconic system, and as- serted that it was '-the lower part of the Silurian," — that is to say, of the Silurian system as then defined by Murchison. This '-upper portion" of the Taconic sys- tem, including the First Graywacke and the Sparry lime- I !' ?i 1)30 THE TACONIC QUESTION IN GEOLOGY. (XL stone, Emmons had found to be fossiliferous, in 1844, and in 1846 declared it to be the stratigraphical equivalent of the Calciferous Sand-rock of the Champlain division, of which he regarded it as a great and "protean development," and "ncluded with it the Red Sand-rock of Vermont, which he supposed to represent the Potsdam. It was not, how- ever, until 1855 that Emmons gave to this paleozic fossili- ferous upper portion of his original Taconic system the name of Upper Taconic, but meanwhile the whole of what was afterwards called Lower Taconic, — including the Primitive Quartz-rock, the Primitive Lime-rock, and the Transition Argillite, — was assigned a position beneath the base of the New York system.] § 145. The above conclusion as to the age of the Red Sand-rock of Vermont was opposed by C. B. Adams and by W. B. Rogers. The former maintained, in 1846, after this announcement by Emmons, the opinion that this saiid-rpck was newer than the Champlain division, and refcred it to "the period Oj. the Medina sandstone and tlio Clintcn group," while W. B. Rogers, in 1851, discuss- ing the same subject, conceived that the reddish limestones which, near Burlington, Vermont, are associated with this sand-rock, were probably "a peculiar development of the upper p«jrtion of the Medina group." As regards the relations of this Red Sand-rock and its succeeding lime- stone to the granular quartz-rock and granular lime- rock of the Lower Taconic, Adams maintained that "the Tatojuc quartz-rock was probably but a metamorphic ecar Quebec, the by Logan as a lous Sand-rock. there disphiyed pparently super- mderlying fossil- ry Lime-rock of reversal of the there can be no .ally an inverted dest and not the .displayed. This th, in chapter \i. how Logan's view ue was made to chists which have ad representatives u the Sillery and Luuzon division. ^, soutliward into bles of the Lower these also in the le Levis limestone veady set forth in eologioal map ot ,hed in 186G, after ,ese rocks through into eas^^rn New imestoue in Massa- XI.] THE TACONIC IIISTOUY liEVIliWED. Olio chusetts is represented as an uninterrupted continuation of tlie Levis limestone from the province of Quebec, brought up along an anticlinal, and having on both sides overlying it, successively, ti>»- Lauzon and Sillery divis- ions, — these, on the west side of the anticlinal, having the ordinary type of the uncrystalline First Grayvvacke or Upper Taconic, but being represented on the east side by the crystalline schists of the Green Mountain range, their supposed e(iuivalents. Few will now question that Logan was wrong in this latter point, or will doubt the greater anti(iuity of these crystalline rocks. On the other hand, it is to be noted that, in thus asserting the infraposition of the Lov/er Taconic marbles to tiie First Graywacke or Upper Taconic series, Logan but confirmed the older ob irvations of Eaton and Ennnons, and only erred in having, by a false interpretation of the succession of the latter series near Quebec, assigned the Levis limestone to its base, by which he was led to confound ii vyith the Lower Taconic limestone. In either view, he placed the latter below the series of several thousand feet of sand- stones, conglomerates, and shales, which constitute the First Graywacke of Eaton and the Upper Taconic of Emmons. § 151. We have already seen that Emmons, as early as 1846, had recognized the fossiliferous character of the First Graywacke, which he afterwards called Upper Taconic ; that he described and figured, in 184G, trilobitic forms found therein, and did not hesitate, in 1860, to declare that it corresponded with the Primordial zone of Barrande.* Thus it happened that Barrande, Marcou, and after him Perry, assumed the Taconic system to be the equivalent of the Primordial zone or Cambrian of Great Britain, Bohemia, and Spain, — they having failed to recognize the distinction which Emmons had made, * See, In this connection, Barrande and Marcou on the Primordial Fauna and the Taconic System ; Froc. Boston Soc. Nat. Hist., Dec, 1860, vol. vii., pp. 369-382. m IMAGE EVALUATION TEST TARGET (MT-3) < ^ 1.0 I.I u US ^ 1^ III 2-0 1.8 1.25 1 u |l.6 ^ 6" — ► Photographic Sciences Corporation ^^ # iV :\ \ k ^ 23 WIST MAIN STREIT WEBSTIR.N.Y. 14580 (716) 872-4S03 f ^Jb 636 THE TACONIC QUESTION IN GEOLOGY. tXI. as early as 1842, between the lower and upper divisions of his original Taconic, when he referred the upper por- tion to wliat he then called the Silurian system. In 1867, J. B. Perry described the Taconic system of Vermont as composed of three parts : 1. Lower, consisting of quartz- ites, marbles, and talcoid schists, the Lower Taconic of Emmons; 2 and 3. Middle and Upper, including the un- crystalline fossiliferous Scranton and Georgia slates, and the overlying Red Sand-rock, which he regarded as the equivalent of Potsdam. The succeeding graywacke, con- stituting a great part of the Upper Taconic of Emmons, was by Perry supposed to be separated by an unconformity from the Red Sand-rock, and he was disposed to divide it from the Taconic and connect it with the Champlain division.* § 152. Still more recently Marcou has given us his own later views of these rocks in Vermont. The true or typical Taconic is, according to him, the Upper Taconic of Emmons, and rests unconformably upon the Lower Taconic. This upper series he divides into four parts, in ascending order, designated the St. Albans, Georgia, Phillipsburg, and Swanton groups. In these are found, besides the Primordial fauna, fossils of i'le second fauna in included limestones, a fact which he explains as indi- cating centres of creation in which the forms of the second fauna first made their appearance ; the whole of these being, according to him, below the horizon of the Red Sand-rock, which he supposes to overlie unconformably the Upper Taconic.f That the forms of the second fauna found in portions of this region belong to a lower hori- zon than the Potsdam, is in discordance alike with the facts of paleontology and of stratigraphy, and is opposed to the conclusions of all other observers in that region, including alike Emmons, Logan, and Perry. Marcou's * The Red Sand-rock of Vermont, etc., J. B. Perry ; Proc. Bost. See. Nat. Hist., 1867, vol. xi. t Marcou, Bull. Soc. G^ol. de France 1880 (3), Ix., pp. lcS-46. LOGY. [XI. vXL] THE TACONIC HISTORY KEVIEW3D. 63T upper divisions the upper por- stem. In 1867, 1 of Vermont as isting of quartz- ovver Taconic of icluding the un- ;orgia slates, and regarded as the y graywacke, con- mic of Emmons, r an unconformity Lisposed to divide h the Champlain has given us his out. The true or le Upper Taconic upon the Lower 8 into four parts, Albans, Georgia, these are foiuid, ciie second fauna 3 explains as indi- 'orms of the second he whole of these orizon of the Red •lie un conformably f the second fauna g to a lower hori- ice alike with the hy, and is opposed ers in that region, Perry. Marcou's Iperry ; Proc. Boat. Soc. , ix., pp. 18-46. conclusions would seem to be based on some of the fre- quent cases of inversion of strata, or of dislocation and upthrow, to wbich we have elsewhere alluded, and which led Logan to place the Levis limestone near Quebec at the base of his Quebec group, and to represent the Taconic marbles of southern Vermont as passing below the crystalline schists of the Green Mountain range. It should, however, here be said, at the same time, that in a disturbed region like eastern Vermont, where areas of the higher rocks of the second fauna exist, and have probably at one time been more widely spread than now, it is not impossible that there may be outliers of a sand- stone of Oneida or Medina age, such as in Pennsylvania we have described as overlying unconformably Lower Taconic rocks, and also that such higher sandstones may have been confounded with the older Cambrian or Potsdam sandstone, and thus afford a seeming justifica- tion for the strange hypothesis advanced by Marcou, that the whole of the Appalachian Cambrian in Vermont is older than the Lower Potsdam. [The late discovery in the Green-Pond Mountain range, in New Jer.sey, in close association with older sediments, of •Silurian lime- stones and Devonian sandstones, as mentioned in § 148, is significant in this connection.] § 153. The studies of the last few years have thrown much light on the character of the lower portions of the Cambrian in its development to the east and southeast of the Adirondack area. It has been noticed tliat the Red Sand-rock and its accompanying slates and limestones near Burlington, Vermont, referred by Emmons to the Potsdam, but by Adams and W. B. Rogers to the Medina, and by Logan to the summit of the Hudson-River groUp, were subsequently by Billings called Lower Potsdam, to indicate that the fauna of these rocks belongs to a some- what lower horizon than the typical Potsdam of tlie New York system. The later studies of Logan in western Vermont, as given by him in 1863, showed that these :!:>sr' i:l:^'iil fl-'i n 638 THE TACONIC QUESTION IN GEOLOGY. tXI. ancient rocks are brought up by a north and south dislo- cation, with an upthrow on the east, from beneath rocks of Trenton, of Chazy, or of Levis age, which latter here occupy their natural position at the summit of the Upper Taconic or First Graywacke group.* Billings, also, in 1868, as already pointed out, had shown that farther southward in Vermont the Red Sand-rock, or Lower Potsdam, is in like manner brought up by a dislocation, 80 as to overlie on the east the Loraine shales. § 154. It was now clear to all, that much of what had been called Hudson-River group to the east of the Hud- son valley and of Lake Champlain, consisted, not, as taught by Mather and his followers, of disturbed and altered strata newer than the Trenton limestone, and of the age of the Loraine shales, but of older rocks, carrying, in part, at least, the forms of the first fauna. We have already seen (§ 112) how, in view of these facts. Hall expressed his opinion, in 1862, as to the relations of these newer strata to the older ones. In 1877, he returned to the subject, and, after retracing the history of investiga- tion, concluded that "we now know approximately the limits between the newer and the older formations, and there is now no longer any question that the newer series, or the rocks above the Trenton limestone, do occupy both sides of the Hudson River for nearly one hundred miles, and continue along the valley for many miles farther towards Lake Champlain. The term Hudson-River group has, therefore, a definite signification, from abso- lute knowledge of superposition and fossil remains. The error lay in extending the term to rocks on the eastward, at a time when their fossil contents had not been studied, and were, in fact, unknown, and their geological position had not been determined by critical examination." f The distinction between the two had however been clearly pointed out by Emmons as early as 1842 (awfe, p. 586.) We * Geology of Canada, cliap. xxii., pp. 844-800. , t Hall, Proc. Amer. Assoc. Adv. Science, 1877, p. 263. XL] THE TACONIC HISTORY REVIEWED. 639 have already shown in §§ 13-14 how Vanuxem had de- vised this term to include, besides the true Loraine shales, other disturbed and apparently non-fossiliferous rocks of controverted age, which he supposed might be included with the former, and thus introduced much of that con- fusion which has prevailed in the use of the name of Hudson-River group as the equivalent to Loraine shales. § 155. The eastern limit of the rocks of the second fauna, along the Hudson valley, being defined as stated by Hall, and as already shown by him for that region on Logan's geological map previously published, it was im- portant to determine the age of the uncrystalline rocks along their eastern border, and to decide whether these were (as mapped by Logan) portions of the so-called Quebec group, or of the still older Potsdam which had been found in this position at several points in Vermont. Nothing has contributed more to the solution of this problem than the careful studies of Mr. S. W. Ford, who, in 1871, discovered the existence of fossiliferous rocks of this lower horizon at Troy, New York, and, following up his investigations, showed that these strata, containing an abundant fauna of Lower Potsdam age (corresponding to the Olenellus slates of Georgia, Vermont, and to the beds at Bic, Quebec, and at the Strait of Bellisle, in Labrador), are at Troy broug]it up on the eastern side of a fault, against the Loraine shales.* Continuing his studies. Ford has recently traced these Lower Potsdam rocks, under similar conditions, through various parts of Columbia and Dutchess Counties, the stratigraphical break and the up- throv of the Cambrian strata on its eastern side being well defined. He does not attempt to estimate the thick- ness of this series of Cambrian sandstones, shales, con- glomerates, and limestones, but says that it " is manifestly very great in eastern New York,"t Emmons having already in 1846 declared its volume to be probably equal * Amer. Jour. Science, 1873, vl., p. 135. t Amer. Jour Science, 1884, xxviii., pp. 35 and 206. >M :':■: ■^.ir I!-' . .'^- 1 f: !<■, 1 i: ■>■!' 1 I i ■ rn;: l-iH 640 THE TACONIC QUESTION IN GEOLOGY. [XI. to thai of all the members of the New York system in their ordinary development (ante, p. 586). § 156. It is hardly necessary to mention that this series of Cambrian fossiliferous rocks, traced by Ford through Rensselaer, Columbia, and part of Dutchess Coun-. ties, along the eastern side of a belt of Loraine shales, is a part of the First Graywacke belt of Eaton, tlie Upper Taconic of Emmons, which Logan, after his ex- amination of the region with Hall, in 1863, described and subsequently mapped as Quebec group. These ob- servers, as has been already stated (§ 115), and as may be seen on Logan's map of 1866, then traced a narrow buL persistent belt of Loraine shales along the eastern side of the Hudson, from Washington County southward to a point a little above Hyde Park, where they found the boundary between these shales and the older group to cross to the west side of the Hudson. The accuracy of this delineation is confirmed by Ford, who, while remark- ing that the distribution of the upper rocks might entitle them to be called the Hudson-River group, suggests, in view of the perplexities which have attended its use, that it would be better " to discard altogether the designation, and go back to the old term, Loraine shales." Ford farther speaks of the "great dislocation," which, at so many points from western Vermont to the Hudson in Dutchess County, brings up the Cambrian rocks against newer strata of Ordovician age. A reference to the sec- tions of Logan and Billings, already cited, will, however, show the existence, not of a single dislocation, but of parallel dislocations, with upthrows on the east side, towards the barrier of older rocks. Of such parallel faults we find, in fact, repeated examples, not only east of the Hudson, but farther southward in many places, along the eastern border of the Appalachian valley, as already pointed out, in § 101. § 157. The one continuous break, with an upthrow, on the south and east, of 7000 feet, extending from Gasp^ to lY. IXI. XI.] THE TACONIC HISTORY REVIEWED. 641 )rk system in ion that this iced by Ford )utchess Coun-, Loraine shales, of Eaton, the n, after his ex- 1863, described Lip. These ob- 5), and as may ed a narrow but J eascern side of southward to a they found the older group to The accuracy of 10, while remark- pks might entitle roup, suggests, in [nded its use, that the designation, , shales." Ford ,n," which, at so b the Hudson in [ian rocks against grence to the sec- ;ed, will, however, islocation, but of fn the east side, ,0f such parallel s, not only east of tany places, along valley, as already Ith an upthrow, on ling from Gasp6 to Alabama, imagined by Logan, was required in his struc- tural scheme, because he had assumed the Levis limestone, (which near Quebeo is brought to adjoin the Loraine shales), to occupy a position at the base of his Quebec group, and to have been originally buried 7000 feet beneath the Loraine shales, in a great conformable series. The strata along the west side of these dislocations in Canada and in Vermont are, according to Logan, either Levis, Chazy, Trenton, or Loraine, the Lower Potsdam being on the east side. In a section described by Bill- ings, and already noticed (§ 148), where the first disloca- tion brings up the Lower Potsdam — which is successively overlaid by Calciferous, Levis, Chazy, and Trenton — against the Loraine, a second parallel fault, a little farther to the east, brings up the Levis against the Trenton. We see, from the late studies of Ford, that the great belt along the eastern border of the Loraine shales, which Logan described and mapped as Quebec group, is iu large part Lower Potsdam. The whole series must now be farther studied in the present light : we must know the real thickness of the Cambrian in the region in question ; the interval therein which separates the Lower Potsdam from the Levis fauna; and how much of the Quebec group of Logan is to be included in the Potsdam. § 158. As regards the relations of the Cambrian and Ordovician rocks over this area, we have already shown that there is every reason to believe that there exists a stratigraphical break between them (as is alsb the case between the Lower Taconic and Cambrian) and farther, that the lower members of the Ordovician series (tlie limestones of the Trenton group) thin out and present, irregularities to the south and east. Although to Hall and Logan it appeared that the line between the Loraine shales and the inferior series passed from the east to the west bank of the Hudson near Hyde Park in Dutchess County, subsequent studies * have shown the existence of * Amer. Jour. Science, xvii., 57. U 'f: ■■! it mm 642 THE TACONIO QUESTION IN GEOLOGY. [XI. the higher strata farther southward, on the east bank. Dale, in 1877, found fossils of the Loraine period in shales at Poughkeepsie, and Dwight soon after detected abun- dant forms of Trenton age in the limestone of the Wap- pinger valley, a little farther south, as well {is at Newburg on the west bank of the Hudson. These discoveries were soon followed by that of a remarkable fauna of Calcifer- ous age in other limestones in the Wappinger valley, thus showing the presence here, as in Vermont, to the east of the outcrop of the Cambrian, of strata carrying the fossils of the Calciferous, the Trenton, and the Loraine subdi- visions. These observations by Dwight were made in 1877-1880,* and joined to those of Dale, and those of Ford, show the existence, in what has there been called both Hudson-River group and Quebec group, of fossil- iferous strata ranging from the Lower Potsdam to the Loraine, both inclusive, — a result identical to that already arrived at in Canada for the area which had been succes- sively mapped as Hudson-Rivrr group and Quebec group. § 159. Having thus recalled the latest results of paleontological research among the so-called Upper Taconic, and shown the association of areas of Ordovi- cian rocks with the predominant Cambrian, we may pro- ceed to notice the views of Prof. J. D. Dana on the Taconic question. He, in 1872 and 1873, published an extended series of papers on the rocks of the Taconic range, as seen in Berkshire County, Massachusetts, and reasoning from the organic forms found in association with similar limestones in Vermont, reached the conclu- sion that the Stockbridge limestone " is mainly Trenton," the overlying schists being of the Hudson-River group.f This latter statement, supported by a stratigraphical argu- ment, may be found in his paper on The Geological Age of the Taconic System. § 160. [In the paper just named (communicated to * Amer. Jour. Science, xvii., 389; xix., 50; xxi., 78; and xxvii., 249. t Ibid., vL, 274. OGY. the east bank, period in shales detected abun- le of the Wap- L as at Newburg discoveries were iina of Calcifer- nger valley, thus t, to the east of .rrying the fossils e Loraine subdi- ht were made in ale, and those of there been called 3 group, of fossil- ,r Potsdam to the [cal to that already 1 had been succes- land Quebec group, latest results ot so-called Upper ,f areas of Ordovi- brian, we may pvo- D. Dana on the x873, published an sks of the Tacomc Massachusetts, and und in association reached the conclu- is mainly Trenton, ddson-River group-t stratigraphical argu- The Geological Age (communicated to ^.,18; andxxvll.,249. XI.] THE TACONIO HISTORY REVIEWED. 643 the Geological Society of London in 1882),* Dana pro- poses to limit the question to the Taconic hills, and the area originally described by Emmons. He claims that the " true original Taconic schists " are those of the Ta- conic range, extending north and south along the boundary between Massachusetts and New York, including the counties of Rensselaer and Columbia in the latter State, to which he adds Dutchess County on the south. In the centre of the range are, according to him, these "Taconic schists," having on the east the Stockbridge limestone (the latter being three times repeated, with intervening Granular Quartz-rock and the Magnesian slates of Em- mons), and on the west the Sparry limestone or Sparry Lime-rock of Eaton, all the strata having a general eastern dip. Dana declares that these three rocks — by which he means the Stockbridge limestone, the Magnesian slate, and the Sparry Lime-rock, neglecting the Granular Quartz-rock — "are all that need be considered," and that the only question is whether these limestones are the same rock, repeated, with alterations in character, to the eastward, or whether the Sparry Lime-rock, which seems to dip beneath all the others, is a newer rock or an older rock than they. Emmons, in 1842, was perplexed by the continuous eastern dip of the strata across a great breadth of country, and expressed doubts on this point, which he was, however, enabled to solve before 1846, and to assure himself that the position of the Sparry limestone was younger than the Stockbridge limestone, or the Magnesian slate which overlies this last, while Professor Dana con- tinues to cherish the contrary opinion. That the Sparry Lime-rock is not only younger than the Stockbridge or Lower Taconic limestone, but belongs at the summit of the First Graywacke, had been clearly pointed out by Eaton, in 1832. § 160 A. [In an ideal section given in 1846 to show * Quar. Jour. (Jeol. Soc, xxxviii., 397, and, in abstract, Amer. Jour. Science, xxiv., 291. ji M r- II i h 644 THE TACONIC QUESTION IN GEOLOGY. PL their supposed order of deposition, Emmons thus arranges the members of the Taconic system: — 1. Granular Quartz-rock ; 2. Stockbridge limestone ; 8. Magnesian slate ; 4. Sparry limestone ; 5. Roofing slate ; 6. Coarse sandstones ; 7. Taconio slate ; 8. Black slate. In a com- panion section, showing the apparent succession of these in the Taconic region, from east to west, he gives : 1. 2. 3, successive alternations of the Granular Quartz-rock, Stockbridge limestone, and Magnesian slate; 4. Sparry limestone, followed by the higher members already noticed, and unconforraably overlaid on the west by the Loraine shales. These numbers, 5-8, we are told, "refer to the Taconic slate in its subordinate beds." * This name of '* Taconic slate " was, in fact, already employed by Em- mons, in 1842, to designate the whole group of strata lying west of the Sparry Lime-rock, and between it and the Loraine shales (^ante, p. 586). The name of "Taconic schists," employed by Dana, serves only to confuse his readers, and was not used by Emmons, who called the schistose strata of the Lower Taconic simply talcose slate, or Magnesian slate, and gave to the great mass oi sedi- mentary strata of the Upper Taconic the collective name of the Taconic slate. The various subdivisions of this Taconic slate group are given by Emmons farther on (loc. cit., pp. 66-67) as coarse greenish sandstones, gray sand- stones, red and chocolate-colored slates, green and black flinty slates, blue compact limestones, and gray silicious limestones, all of them lying to the we^t of " the great mass of the Sparry limestone." The order of these is variable, and the observer " will, in the space of fifteen or twenty miles, pass several times over the same beds, which are brought up by many successive uplifts " with a seem- ing thickness of 25,000 feet (loc. cit., p. 67). The nature of these movements of dislocation, by which the subdivisions of the Taconic slates are thus repeated, is farther shown by an ideal section in 1855. At the same • Agriculture of New York, i., pp. 60-61. XL] THE TACONIC HISTORY REVIEWED. 645 )GY. l^'"'* thus arranges ^1. Granular 3. Maguesiau ite ; 6. Coarse ,te. In a com- ession of these 3 gives : !■ 2. o, r Quartz-rock, ate; 4. Sparry already noticed, by the Loraine I, "refer to the This name of aployed by Em- ip of strata lying ween it and the me of "Taconic y to confuse his . who called the nply talcose slate, eat mass ol sedi- coUective name ^divisions of this iis farther on (loc. stones, gray sand- , green and black >nd gray silicious Lt of "the great order of these is space of fifteen or 5 same beds, which lifts" with a seem- lit., p. 67). The ion, by which the thus repeated, is 155. At the same ). 60-61. time the real order of succession in the Upper Taconic was declared to be, — greenish chloritic sandstones at the base, followed upward by a great mass of various colored sUvtes and sandstones, and, towards the top, by the Sparry limestone, with quartzose and conglomerate beds, black shaly limestone, and fine black slates.* § 161. [Nothing of all tiiis can be gathered from Dana's statements. In his latest communication on the subject, read to the American Association for the Ad- vancement of Science, in August, 1885,f he refers to Emmons's description of the Taconic system in 1855, wherein, he would have us believe, the Sparry Lime-rock is made a part of the Lower Tacc^^ic. By referring thereto,! we, however, find it to consist of: A, Granular Quartz-rock, with repeated interstratifications of so-called talcose slate ; B, Stockbridge limestone, and, C, overlying talcose or Magnesian slate, with included roofing-slate, 2000 feet thick. These "form by themselves a distinct physical group," in the Taconic range, about 5000 feet thick, and Emmons adds : " the sequence of the Lower Taconic rocks, which has been stated and illustrated in the foregoing pages, is essentially the same from Maine to Georgia." No mention is there made of " the Sparry limestone, with its associated slates," which Dana seems to say are included by Emmons in his Lower Taconic, and the only apparent ground for this interpolation is the statement of Emmons that the Stockbridge limestone " is seamy and sparry," or, as he elsewhere says, " occasionally sparry," a fact which, he tells us, had led others to mistake it for the Sparry limeatone of Eaton (awie, p. 585). No place is left for the Sparry limestone in the Lower Taconic, and in the Upper Taconic this, as well as the other limestone- masses and fossiliferous slates, is placed towards the sum- mit, and not at the base. This is in complete accord with • American Geology, il., p. 49, 13. t Amer. Jour. Science, 1886, xxxi., p. 241. t American Geology, ii., pp. 15-20. 646 THE TACONIC QUESTION IN GEOLOGY. [XI. ■'■■■{ 11 m,' I :i:W the remarkable section at Quebec, as interpreted by Bill- ings and myself, as well as by James Hall, who, in 1857, spoke of the graptolitic shales of that vicinity as being near the Sparry Lime-rock of Eaton, towards the summit of the Hudson-River group, as it was then called (^ante^ p. 587). Tiie horizon of this was declared by Billings to be considerably higher than that of the New York Cal- ciferous, and at the base of the second fauna of Barraude (aw^e, p. 625). The recently announced discovery by Messrs. Ford and Dwight, of organic forms, believed by them to be of Trenton age, in the Sparry limestone found in the Upper Taconic area, in Canaan, New York,* is only another instance of the fact, so often insisted upon in these pages, of the recurrejice of Ordovician strata at many points along this great Gray wacke belt from Quebec to Pennsylvania.] § 161 A. [That the Sparry limestone was regarded by Emmons as related to the Taconic slate group, or rather a subordinate part thereof, is evident from his descrip- tions in 1846. After noting that this limestone, while generally persistent in its extent throughout the counties of Dutchess, Columbia, Rensselaer, and Washington, in New York, seems in parts of its distribution to be "en- gulfed, pinched out, or lost," for short distances, he farther tells us that, though the principal mass of this limestone occurs on the eastern border of the Taconic slate group, similar masses, often thinner, are found farther westward in the sections, and that, while some of these " are, undoubtedly, mere repetitions of the same mass " of Sparry limestone, others are distinct. He conjectures, therefore, that the production of limestones of this type " occurred at intervals during the whole period of the de- position of the Taconic slate," thus clearly showing that, in his opinion, it belongs to the Taconic slate group or Upper Taconic, and not, as Dana imagines, to the Lower Taconic. Emmons at the same time informs us that he * Amer. Jour. Science, 1886, zxxi., p. 240. XI.] THE TACONIC HISTORY REVIEWED. 647 had found organic remains (including trilobites and graptolites) in three of the subordinate nienibors of tlie Taconio shite group, namely, the green sandstones, the green slates, and the black slates, and remarks, with regard to the Sparry limestone, " no fossils have yet been dis- covered in this rock, though it must be confessed sufficient examination has not been made for microscopic bi- valves." * [The conjecture of Emmons as to the recurrence of sim- ilar limestones at different periods during the deposition of the Taconic slate group is so far true that there are many bands of limestones, both pure and magnesian, among the shales in the upper portion of the group. This is well shown in the section at Pointe Levis, near Quebec, where numerous bands of this kind were mapped by Logan in 1861. Of these repeated interstratifications of pure limestones, dolomites, sandstones, and argillaceous shales of Pointe Levis the author had already written in 1856: "Both limestones and dolomites are very irregular and interrupted in their distribution, the beds sometimes attaining a considerable volume while at other times they thin out or are replaced by sandstones." Some of these were then described as forming masses many feet in thickness, of pure limestone, without visible marks of stratification, and without organic remains, and were compared to travertines, while others were granular and fossiliferous, more or less magnesian, frequently conglom- erate, and passing into dolomites and dolomitic sandstones. In this section "other organic forms, obscure and un- determined, occur in the calcareous beds both above and below " the belt of graptolitic shales.f § 162. The different views with regard to the geolog- ical horizon of the Lower Taconic or Stockbridge lime- stones of Emmons — the Granular Lime-rock of Eaton — may be resumed as follows : — * Agriculture of New York, vol. I., pp. 68-74. t Azoic Rocks, pp. 101-104, 106, 133. 648 THE TACONIC QUESTION IN GEOLOGY. [XI. l|.« f^'^^m :^^4 ,■ ■ » ' .v^'HSfll^l ^kjt ^c^^^^tl 1^ PI Iff 5 '' 1 4 ( ■ mam 11 I. That they are pre-Cambrian, and occupy a position below the Potsdam sandstone or Red Sand-rock, and the Quebec group of Logan, which together constitute the First or Cambrian Graywacke of Eaton and the Upper Taconic of Emmons, as shown in the table, § 18. (Eaton, Emmons, Perry. Marcou.) II. That, although lying beneath the greater part of this Graywacke series, they are not distinct therefrom, but are the altered representative of the Levis limestone or Sparry Lime-rock, imagined by Logan to lie between the Red Sand-rock below and the chief part of the Quebec group above. (Logan, in his geological map of 1866.) III. That they are the altered representatives of the whole of the limestones which, in the New York system as seen in the Adirondack area, appear between the Pots- dam sandstone and the Utica slate. (Mather, H. D. and W. B. Rogers, J. D. Dana.) IV. Allied to the last is the view expressed by Wing, in 1875, that they include the representatives of the lime- stones of the Potsdam and Quebec groups of Logan, together with the Trenton and the Loraine or Hudson- Ri^^er group, or, in other words, the whole of the Cham- plain division of the New York system, from the Potsdam to the base of the Oneida. V. That they belong to a horizon above the Champlain division, and are true Silurian and Devonian. (C. B. Adams, Ed. Hitchcock, W. B. Rogers.) § 163. We have already briefly set i'orth the arguments on which these various and contradictory hypotheses have been ? ised. While the fifth supposes the Lower Taconic limestone to hold a position above the Oneida sandstone, and consequently superior to the Second Gray- wacke, the third was devised at a time before the existence of the First Graywacke (maintained by Eaton and Emmons, but denied by Mather) had been again brought into favor uy the conversion of Logan to the teaching of Emmons, and by his farther admission that the Lower Taconic limestones iGY. CZX. XI.] THE TACONIC HISTORY REVIEWED. 649 ipy a position ■rock, and the jonstitute the nd the Upper § 18. (Eaton, rreater part of net therefrom, .evis limestone to lie between b of the Quebec ap of 1866.) ntatives of the sw York system tween the Pots- ither, H. D. and ressed by Wing, ives of the lime- •oups of Logan, aine or Hudson- le of the Cham- ■om the Potsdam the Champlain svonian. (C. B. bh the arguments ftory hypotheses Lses the Lower bove the Oneida Lhe Second Gray- fore the existence [ton and Emmons, [rought into favor Ig of Emmons, and Faconic limestones in Vermont and Massachusetts are inferior to a great mass of sandstones, conglomerates, and shales many thou- sand feet in thickness, constituting what he called the Lauzon and Sillery divisions of the Quebec group. § 164. It was not until after his change of view as to the geological horizon of this great sedimentary or Gray- wacke series, or, in other words, after he had recognized the fact that its place is below and not above the Trenton limestone, that Logan began to examine the Lower Taconic rocks in western New England. Having then, by a mis- conception, placed the Levis or Sparry Lime-rock at tho base instead of the summit of the Graywacke, and still holding to the notion of Mather that the crystalline rocks along the eastern border of the great Appalachian valley are but a portion of the paleozoic strata in a so-called metamorphic condition, Logan was led to look upon the Lower Taconic limestone as an altered representative of the Levis limestone, and its underlying quartzite as Potsdam ; the immediately overlying schists, and the suc- ceeding sandstones, conglomerates, and shales of the Gray- wacke series, being referred to the Lauzor and Sillery divisions of his Quebec group. Hence the wide difference between the view of Logan, given under IL, and that of Mather and his followers, which we have numbered HI. While both would place the Lower Taconic limestones above the Potsdam and below the Oneida, Mather imag- ined the slates and sandstones overlying them to be Ordo- vician and Silurian (that is, Utica, Loraine, and Oneida) or the Second Graywacke of Eaton. Logan, on the other hand, conceived the same overlying beds, as seen by him in Vermont, Massachusetts, and New York, to belong to the Cambrian or First Graywacke. The error of Mather and of H. D. Rogers was that both failed to recognize the distinctness of this great series of sandstones, conglomer- ates, and slates, which are so conspicuous in the Appala- chian valley, and confounded them with the Second Gray- wacke. This error it was which completely misled the i r ;| ■' I 1 iti i ■i' ■ \ 1 1 t 5 :J:! ti \ 1 1 ^lif 650 THE TACONIO QUESTION IN GEOLOGY. [XL 'hi mm geological survey of Canada up to 1861, and continues to obscure the subject in the minds of many American geo- logists to the present time. § 165. It should be remembered that, as already pointed out in chapters ii. and iii., the overlying Graywacke or Upper Taconic does not include the schistose rocks imme- diately above the Lower Taconic limestone, but that a considerable amount of crystalline schists and argillites occurs, both interstratified with and overlying this lime- stone, and forming an integral part of the Lower Taconic series. We have, moreover, set forth in chapter v. evi- uencfcs of the distinction between the Upper and Lower Taconic, ax^d have shown that the latter is not limited to the great Apptvlachian valley, which confines the former, but is met with in more or less interrupted belts lying upon the crystalline rocks of the Atlantic region south and east of the great valley, from New Brunswick to Georgia. Thus in North Carolina not less than four dis- tinct and separate parallel bands of the Lower Taconic are met with between that of the great v illey and the overlying tertiary strata of the coast, while similar narrow bands of the same rocks are found in southern New York and New Jersey, lying upon the ancient gneisses. With none of these Lower Taconic belts outside of the great valley, so far ag 1^; known, is the Upper Taconic to be found, its absence being due either to erosion, or, more probably, as suggested by Emmons, to the elevation of these areas above the sea during Cambria', time. § 166. On the other hand, it has been shown in chap- ter vi. that what Mathev regarded as a continu<'.tion of the great Graywacke series from the east of the Hudson extends south-westv; ard across Orange County and, ac- cording to Horton., there rests, with a high eastern dip, on the northwest side of the gneissic belt of the Highlands. From central "^'ermont, northeastward along the great valley, to the St. Lawrence below Quel^ec, the Lower Taconic is not known, and the Upper Taconic or Gray- XI.] THE TACONIO HISTORY REVIEWED. 651 iontin\ies to lerican geo- jady pointed raywacke or rocks imme- , but tliat a ind avgilUtes ng this lirae- ower Taconic tiapter v. evi- 31 and Lower aot limited to es the former, ed belts lying e region south Brunswick to than four dis- Lower Taconic V lUey and the i similar narrow liern New York nieisses. With de of the great Taconic to be irosion, or, more ihe elevation of . time. shown in chap- continu.'.tion of of the Hudson County and, ac- ]h eastern dip, on I the Highlands, .along the gi-eat |el>ec, the Lower 'aconic or Gray- waeke series rests directly upon older crystalline schists, as in Orange County, New York. The same condition of things is again seen in Newfoundland. These facts, already given in detail, serve to show the distinctness and independence of the crystalline Lower Taconic from the uncrystalline Upper Taconic or Cambrian series, which two were probably separated by a considerable interval of time, corresponding to the stratigraphical break, long since pointed out by Eaton, at the base of the First or Transition Graywatjke. § 167. The student wLo refers to Dana's paper of 1882, already noticed, on "The Age of the Taconic System," will obtain no light on the question of this Graywacke series, nor, indeed, any evidence that the author has ever seriously studied the literature of the subject, or compre- hended its relation to the complex problem before us. He will get no notion of the two opposing vie'vs as to this series of rocks, or its position as above or below the Trenton limestone, or even of its existence as a great suc- cession of uncrystalline sediments, many thousand feet in thickness, and distinct from the Lower '"^.aconic limestones, as maintained alike by Eaton, by Emmons, by Mather, and by Logan, and as set forth in the preceding chapters. § 168. The hypothesis of Mather and H. D. Rogers as to the Lovver Taconic rocks was devised at a time when the progress of geology in New York had made known, in the Northern district of that State, a great series of nearly horizontal fossiliferous strata resting upon the up- turned granitoid gneiss of the Adirondacks and includ- ing certaixi familiar subdivisions of the paleozoic, from the Potsdam sandstone upwards. The relations and suc- cession of these various rocks were simple and evident. To the east and southeast of this region, however, beyond Lake Champlain and the Hudson River, there were found other crystalline rocks, unlike the ancient gneiss, and other uncrystalline sediments very different in physical character and in stratigraphical attitude from the paleozoic strata of ^fil 652 THE TACONIO QUESTION IN GEOLOGY. [XL the Northern district of New York. The question then arose as to the correlation of these unlike rocks in the two regions. Amos Eaton, by a grand generalization, had already arrived at a system of classification in which he recognized the existence in the eastern or Appalachian region of types of Primitive crystalline rocks other than the granitoid gneiss, and of great masses of sedimentary strata to which nothing similar was found in the contem- porary series in the Adirondack region. § 169. Rejecting the teachings of Eaton, and falling back on the metamorphic doctrine, which was then so gen- erally received, Mather maintained, in 1843, that whatever to the east of the Hudson differed lithologically from the ancient gneiss, on the one hand, and from the paleozoic rocks of the New York system, as seen in the Adirondack region, on the other, could be nothing else than these same paleozoic rocks folded and subjected to successive stages of so-called metamorphism, as seen in the Lower Taconic quartzites and marbles and the crystalline schists which accompany them, as well as those others that succeed them farther to the east. All of these were, according to Mather, nothing but the more or less altered equivalents of the members of the New York system, from the Potsdam sandstone to the Loraine shales, both inclusive ; while the great Graywacke belt, extending along the east side of the Hudson from Dutchess County northward through Vermont was declared to be, not, as maintained by Eaton, older than the Trenton limestone, but newer than the Loraine shales. § 170. The considerations which lent probability to tliis scheme were, first, the general resemblance of this Graywacke series to the Oneida, Clinton, and Medina subdivisions of the New York s} ^tem, to which it was by Mather referred ; and, secondly, the fact that the argillites with unctuous schists, granular limestones, and granular quartzite, which he agreed with Eaton and Emmons in placing below the adjacent Graywacke, presented a certain GY. XL] THE TACONIO HISTORY REVIEWED. 658 question then ; rocks in the generalization, ition in which r Appalachian ;ks other than )£ sedimentary in the contem- )n, and falling ras then so gen- ,, that whatever jically from the n the paleozoic the Adirondack else than these id to successive n in the Lower -ystalline schists lers that succeed jre, according to jred equivalents :om the Potsdam usive ; while the the east side of thward through .ained by Eaton, newer than the probability to mblance of this ;on, and Medina which it was by _;hat the argillites les, and granular and Emmons in •esented a certain resemblance to the Loraine and Utica shales, the Trenton and Chazy limestones, the so-called Calciferous Sand-rock, and the underlying Potsdam sandstone. This general parallelism from the top of the Graywacke downward, which, to the mind of Eaton, suggested only the great law of cycles in sedimentation (since generally recognized), was accepted by H. D. Rogers and by Mather as a proof of identity. In fact, the Lower Taconic, as seen along the Appalachian region, in its regular succession of gran- ular quartzites with granular limestones and intervening and overlying soft schists and argillites, presents, notwith- standing its many mineralogical differences, its crystalline character, and its great thickness, a general parallelism to the Champlain division, like that so often remarked in groups of sedimentary strata at very various geological horizons. It is thus, in certain respects, more like the Adirondack Cambrian and Ordovician, with which it has been confounded, than their Appalachian representatives. These resemblances were coupled with the fact that along the base of the South Mountain, in Pennsylvania, this suc- cession is found lying between the ancient granitoid gneiss beneath, and the Oneida sandstone above, precisely as the Potsdam-Loraine succession in northern New York inter- venes between the same gneiss and the same sandstone. § 171. It was not, therefore, surprising that the geolo- gists then engaged in the study of Pennsylvania, New Jersey, and southern New York, should have accepted this plausible and, at first sight, natural explanation of the apparent lithological parallelism presented between these regions and northern New York, or that Mather endeavored to extend it to the rocks east of the Hudson. This attempt led him to assign to the great G;aywacke series which we now know to be of Cambrxcin age, a position above the Loraine shales, or, in other words, to thus to mistake the First for the Second Graywacke of confound it with the Oneida, Medina, and Clinton subdi- visions of northern New York and of Pennsylvania, and I ,1 if • i I - ii • 654 THE TACONIC QUESTION IN GEOLOGY. CZI. m mm H mm fBBmBBsSB m Itfll Eaton, and, in fact, to deny the existence of the former as a great series lying above the Lower Taconic and below the horizon of the Trenton limestone. The two brothers Rogers, with Mather, forty years since, reasoning from the paleozoic succession as displayed in the Adirondack area, were not prepared to admit that, in a region so near as the great Appalachian valley, the paleozoic sediments beneath the Trenton horizon could assume a type so unlike the well known Potsdam and Calciferous subdivis- ions of the Northern district of New York, or that these subdivisions could be represented in the Appalachian area by the vast and lithologically unlike series of the First Graywacke, which Eaton had already, ten years before, assigned to its true position below the horizon of the Trenton limestone. Hence came the great mistake in American stratigraphy, the denial by Mather and his followers of the distinctness of the First Graywacke of Eaton, and the assertion of its identity with the Second Graywacke of the same author. So long as this false position was maintained, there was a plausible argument to be made for the original hypothesis of the brothers Rogers and Mather as to the age of the Lower Taconic series ; but with the recognition of the correctness of Eaton's view of the First Graywacke, the fallacy of this hypothesis became obvious, and those who would still advocate it can only do so by rejecting the results alike of stratigraphical and paleontological study for the last gen- eration. IX. — THE METAMORPHIC HYPOTHESIS. § 172. The absence from the granular quartz-rock, the granular marbles and their intercalated and conformably overlying schists and argillites of the Lower Taconic series, of the organic remains of the various members of the Champlain division, or, indeed, of any organic form save the peculiar Scolithus of the granular quartz-rock already noticed (§ 23), was explained by those who maintained the former as c and below two brothers ling from the L-ondack area, on so near as )ic sediments ne a type so irons subdivis- or that these palachian area 8 of the First . years before, lorizon of the ;at mistake in ather and his Graywacke of ith the Second g as this false sible argument »f the brothers Lower Taconic correctness of fallacy of this 'ho would still results alike of lor the last gen- Ihesis. luartz-rock, the id conformably ^ower Taconic I members of the Irani c form save rtz-rock already ^ho maintained xi.i THE METAMORPHIC HYPOTHESIS. 655 the paleozoic age of the series by the convenient hypoth- esis of a chemical change, attended by crystallization or so-called metamorphism, which was supposed to have effaced the original characters of the sediments, and oblit- erated their organic remains. In accordance with this hypothesis, it was believed that great series of strata might, within short distances, assume a new aspect, not through any original differences in the sediments, but from transformations wrought in these after deposition, in virtue of which fossiliferous and earthy limestones, losing all traces of their organic remains, could be con- verted into granular limestones containing, instead, only crystalline silicates — while ordinary sandstones and argil- lites might become micaceous, chloritic, or hornblendio schists, and even gneisses and granite-like rocks. § 173. These views, a development of the Huttonian school in geology, were, as is well known to students, accepted a generation since by a large number of geolo- gists, both in Europe and America, and were carried to an extreme in America. Mather, in his final renort on the geology of the Southern district of New Yor.., declared that " the Taconic rocks are of the same age with those of the Champlain division, but modified by metamorphic agency and by the intrusion of plutonic rocks." They were, however, designated by him as " imperfectly Meta- morphic rocks," while the various crystalline schists of New York and western New England, included by him in his group of proper Metamorphic rocks, were declared to be the same series in a still more highly altered condition (§ 121). Respecting these, he asserted that where the Taconic and Metamorphic rocks come together, " no well marked line of distinction can be drawn, as they pass into each other by insensible shades of difference." Mather was disposed to admit, in addition to these, an older or so-called Primary series of crystalline rocks in the High- lands of the Hudson, but, in the course of his report, ended by declaring that the Primary limestones of south- 656 THE TACONIO QUESTION IN GEOLOGY. [XL em New York and northern New Jersey, with their asso- ciated granitic and hornblendic rocks, were nothing more than modifications of the members of the Champlain divis- ion. He had been led to believe that the Primary lime- stones in question " can be easily traced through all the changes from a fossiliferous to a crystalline white lime- stone, containing crystallized minerals and plumbago." From the interstratification of these crystalline lime- stones, supposed by him to be paleozoic, with gneissic and hornblendic rocks, he was brought to maintain the paleo- zoic age of these, and thus to doubt whether a portion, at least, of what he had called Primary gneiss was not also paleozoic. § 174. Apart from the crystalline rocks of the High- land or South Mountain belt, whose primary character was in part questioned by Mather, the great area of crys- talline rocks lying to the south and east of this range in New York, comprising those of Westchester and New York Counties, and embracing Manhattan Island, was by him included, with the adjacent rocks of western New England, in his Metamorphic series, and declared to be "nothing more than the rocks of the Champlain division, modified greatly by metamorphic agencies and by the intrusion of granitic and trappean aggregates." In this area of southern New York he noticed hornblendic rocks, gneiss, mica-schists, and crystalline limestones, besides granite, syenite, and serpentine, the latter three being regarded by him as intrusive rocks.* § 175. The doctrine of the Metamorphic school of forty years since, as then resumed and formulated by Mather, was briefly as follows: The different groups of crystal- line stratified rocks in southeastern New York and west- ern New England (with the doubtful exception of the ♦Tor the details of these views see Mather's Geology of the Southern District of New York, 1843, paaaim. A summary of Mather's somewhat diffuse statements will be found in the author's volume on Azoic Bocks, pp, 38-42. Y. [XI. XI.] THE METAMORPHIC HYPOTHESIS. 657 1 their asso- )thing more nplain divis- limary linie- 3Ugh all the white lime- plumbago." talline lime- i gneissic and in the paleo- • a portion, at was not also of the High- lary character , area of crys- this range in 3ter and New Jsland, was by western New ieclared to be plain division, a and by the es." In this blendio rocks, tones, besides r three being school of forty ed by Mather, ups of crystal- ork andwest- ception of the .y of the Southern -lather's somewhat le on Azoic Rocks, gneissic belt which he had designated Primary), including the Lower Taconic series, the series of micaceous gneisses and mica-schists, as well as the massive granitoid and hornblendio gneisses with tlieir crystalline limestones, all belong to one and the same geological period, and are contemporaneous in age with tb^ paleozoic rocks of the Champlain division of northern ^low York, from the Pots- dam sandstone to the Loraine shales, both inclusive. These various and unlike, though coiitiguous groups of crystalline rocks were, according to Mather, all produced from the same uncrystalline Cambrian and Ordovician sediments, through a mysterious process of transforma- tion, by what he called " metamorphic agencies," and the intrusion of igneous rocks, in which category he included not only the interbedded serpentines, but apparently, under the name of granites, much of the granitic gneiss, which characterizes large areas of the region, as well as the abundant endogenous granitic veins, — true intrusive or exotic granites being rare in the region. In Mather's cosmogony there was nothing in the geological sequence, at least in northeastern America, between the New York paleozoic series, as seen in the Adirondack area, and the fundamental Laurentian gneiss which there underlies it. Consequently all crystalline rocks which could not be referred to the latter were, unless plutonic, the result of some unexplained transformation of this lower part (ji. the paleozoic column, known as the Champlain division. § 176. This hypothesis, extravagant as it now seems, was, during the next few years, accepted by many geolog- ical students on the authority of Mather and the brothers H. D. and W. B. Rogers. These latter, in 1846, extended this view of Mather to the White Mountains of New Hampshire, and suggested that the gneissic, hornblendic, and micaceous rocks of this series, since named Montal- ban, instead of belonging, as hitherto believed, to the "so-called Primary periods of geological time," were prob- ably altered paleozoic strata of Silurian age, including the I ..m, ^ ^1'i|' M n >(ii ; : 1' ill ! 058 THE TACONIC QUESTION IN OEOLOOV. IXI. Oiiekla, Medina, and Clinton subdivisions of tho New Yorlc system. These observers then proceeded to name many species f)f characteristic organic forms of the Silu- rian i)eriod, 'which they thought to recognize in certain crystalline aggregates in the mica-schists of the region. In 1847, however, the same observers announced that they no longer considered these forms of organic origin, * and, while they did not then formally retract their opin- ion as to the paleozoic ago of tho gneisses and mica-schists of the White Mountains, are known, from their subse- quent writings, to have abandoned it as unfounded, although it was for some years afterward maintained, with some variations, by Logan, Lesley, and the present writer.! § 177. As regards the ancient crystalline series of the Highlands of the Hudson and of New Jersey, which differs in lithological characters from the last, we find that H. D. Rogers, while he did not accept the notion of Nuttall and of Mather that its gneisses are altered paleo- zoic sediments, im" 'fined the crystalline limestones, whicli are really interstratified with them, to be portions of a younger limestone, altered by supposed igneous agencies. In the words of Lesley, Rogers, while maintaining the Primary age of the Highland gneisses, " mistook the crys- talline limestone engaged among the Highlands for meta- morphosed synclinal outliers of No. II., as at Franklin," in New Jersey, whereas Cook has since shown that the hori- zontal strata of this later period overlie the upturned crystalline limestones of Franklin.:): As a consequence of this, H. D. Rogers was quoted by Mather as support- ing the extreme notions of metamorphism maintained l)v Nuttall in 1824, which Mather himself accepted, and which, as I have elsewhere said, " were adopted by H. D. *Amer. Jour. Science [2], L, 411, and v., 116. t See, for historical notes. Hunt, Amer. Jour. Science, vol. 1., 84; also Azoic Rocks, pp. 62, 181, 182. X Lesley, Amer. Jour. Science, 1865, xxxix., 222. QY. ixi. XI.] THE METAMORPHIC HVPOTIIESIS. O.V.) II ; of the New iiled to name J of t\ie Silu- ize iu certain )f the region, uounced that ganic origin,* vet thoir opin- iid mica-schists xn their suhse- as unfounded, rd maintained, ,nd the present ne series of the ' Jersey, which ^e last, we find pt the notion of re altered paleo- imestones, which he portions of a gueous agencies, 'maintaining the mistook the crys- [hlands for nieta- i at Franklin," iu ^vn that the hovi- [ie the upturned u a consequence [ather as support- Im maintained V)y klf accepted, and \dopted by H. D. IciencC; vol.1.. 84; also R«)gers, as far as regards the crystalline limestones of the Highlands in New Jersey,"* — while he soon after ap- plied the same doctrine, in its fullest extent, to the great gneissic series of the White Mountains. § 178. To sum up in a few words the views of the Metamorphic school forty years since (1840-184(3), we find that 11. I), and W. B. Rogers then maintained the paleozoic age of the Lower Taconic series, of the White Mountain gneisses and mica-schists, and also of the crys- talline limestones found among the gneisses of the New York and New Jersey Highlands, though admitting the primary age of these Highland gneisses. Mather, again, while holding, in like manner, to the paleozoic age of the Lower Taconic, was not acquainted with the White Mountain series, but maintained that the whole o'" the gneisses, mica-schists, and crystalline limestones of south- eastern New York, with the possible exception of the Highland belt, were paleozoic, and of one age with the Taconic series. It is worthy of note that on the geologi- cal map of the State of New York, published in 1842, ''by legislative authority," of which the Southern district was prepared by Mather himself, there is no distinction of color between the gneissic rocks of the Highlands and those adjacent to them on the south and east, described by him in his final report, in the following year, as meta- morphic paleozoic strata. The serpentine of the region, as seen in Staten Island, is colored on the map like the adjacent intrusive triassic diabase,! but no attempt is there made to designate other eruptive rocks than these. § 179. In opposition to the views of this Metamorphic school, there were not wanting some, like Emmons and Charles T. Jackson, who maintained the Primitive age of the whole, or a part, of these crystalline rocks of New England, though recognizing, as Eaton had done, their * Hunt, Azoic Rocks, p. 41. t See, for details witli regard to this and the other serpentines of the region, ante, pp. 435-442. iU 1 1 III, , I.. . , ^h I II 'i Ih 6(30 THE TACONIC QUI«TtON IN QEOLOCJY. [XL lithological distinctness from the gneiss of the Adiron- ducks and of the Ilighhinds of the Iludsun. Already, moreover, in 1824, Bigsby had discovered, around Luke Superior and beyond, the existence of two series of crys- tal lino rocks, and distinguished the younger of these as beh)nging to the Transition series. More than twenty years hiter, the geoh)gical survey of Canada, while adopt- ing for the crystalline rocks of New England, and their extension into ('anada, the hypothesis of their i)aleozoic age, re-examined these Transition crystalline schists of Bigsby, as seen both on Lakes Superior and Huron, and on the upper Ottawa, and described them as forming a distinct group between the base of the paleozoic series and the ancient gneiss, upon which it was found to rest unconformably. This intermediate series, first described in 1847, was by the present writer designated, in 1855, by the name of ;iuronian, — the underlying gneissio series having, in 1854, received the name of Laurontian. With the Huronian, as we have endeavored to show (^ante, pp. 414, 581), have since been included, in the region of the great lakes and elsewhere, areas of Taconian rocks. § 180. In 1858 appeared the final Report of H. D. Rogers on the geology of Pennsylvania, in which we find no recognition of the extreme doctrines of metamorphism maintained by Mather in 1843, and by W. B. Rogers and himself in 1846. Not having come to an understanding of the question of the P'irst Graywacke, H. D. Rogers regarded the Lower Taconic series in Pennsylvania as an altered form of the Champlain division, and considered the granular quartz-rock with Scolithus to be the equiva- lent of the New York Potsdam sandstone.* The charac- teristic crystalline rocks of western New England and southeastern New York, described by Mather as altered paleozoic, pass beneath the mesozoic sandstone in New * For Lesley's doubts as to the precise equivalence of the Primal quartzite of Pennsylvania and the New York Potsdam, see Aiuer. Jour. Science, 18d5, xxxix., 223. I Y. pa. the AtUron- IX. Alremly, wound Lake nieft of cvys- i- of tliese as than twenty while tttlopt- ind, and then- lieir paleozoic ine sciiista of id Huron, and I as forming a ,aleozoic series found to rest first described ;etl, in 1855, by gneissio series rentian. With show iante, pp. le region of the ^an rocks, iport of H. D. which we find metamorphism B. Rogers and understanding „ H. D. Rogers nsylvania as an and considered be the equiva- , * The charac- w England and lather as altered ndstone in New Lnce of the Trimal kam, see Aiuer. Jour. XL] THE METAMOIIPHIC HVPOTHK8I8. fiOl Jersey and re-appear in sontheastern Pennsylvania. Those rocks were now, in 1858, describeil by II. 1). Ilogcis us forming two greiit gronps, an older or so-called llvjinzoio gneiss system, and a younger one of crystalline scliists, which he called Azoic and placed beneath the horizon of the Scolithus sandstone. The views of H. D. Rogers, in 1858, with r»'gard to the crystalline rocks of the Atlantic belt, were thns, as I have elsewhere said, "a retnrn to those held by Eaton and by Emmons, but were in direct 02)position to that maintained by Mather, which bad been adopted at that time by Logan and by the jjresent writer" (ante, p. 40G), and, so far as regards the White Moun- tains, were nuiintained by the Messrs. Rogers themselves, in 1846. § 181. Henry D. Rogers died in 1867, but his venera- ble brother, William li. Rogers, survived till 1882, and fully shared the views set forth by the former in 1858, as to the pre-paleozo'ic age of the great groups of crystalline rocks. His careful and extended studies in Virginia dur- ing many years had convinced him of the fallacy of the metamorphic hypothesis of Mather. In a sketch of the geology of that state, contributed by him as late as 1878 to James Macfarlane's "Geological Railroad Guide," Rogers makes it plain that the crystalline rocks of tliat region are all pre-paleozoic, and older than what he calls the Primal or Potsdam group. This he describes as lying on the western slope, and in the west-flanking hills of the Blue Ridge, " often by inversion dipping to the southeast, in seeming conformity, beneath the older rocks of the Blue Ridge, but often, also, resting unconformably upon or against them." These older rocks, he tells us, "com- prise masses referable probably to Huronian and Lauren- tian age," and, farther, he informs us that the letters, A, B, C, and D, used in his tabular view, " mark four rather distinct groups of Archean rocks found in Virginia, of which the first three may probably be referred to the Laureutian, Huronian, and Montalban periods, respec- n it: 662 THE TACONIC QUESTION IN GEOLOGY. IXI. tively, and the fourth to an intermediate stage, — the Norian or Upper Laurentian." § 182. It should here be remarked that this Primal group of the valley of Virginia, also called by Rogers Lower Cambrian, is no other than the base of the Lower Taconic series, which he continued to regard as in some sense the representative of the Cambrian Potsdam of the Adirondack region. Li this connection, as showing the relations of this group to the crystalline rocks, and the apparent inverted succession, I venture to make the fol- lowing extracts from a letter from W. B. Rogers, written to me in 1877, for publication in the volume on Azoic Rocks, after an examination with him of some forty unpublished transverse sections, made across the Blue Ridge during his geological survey of Virginia. In many of these sections " illustrating the position of the Lower Cambrian (our Primal conglomerate, etc.), in their con- tact with the crystalline and metamorphic rocks of the Blue Ridge in Virginia," " the unconformity of the Cam- brian upon and against these crystalline and metamor- phic rocks is unmistakable and conspicuous; the lower members of the Primal being seen to rest upon the slope of the Ridge, with northwest uiidulating dips, on the edges of the steeply southeastward-dipping older rocks. lu other cases, the Primal beds, thrown into southeast dips in the hills which flank the Blue Ridge, are made to underlie, with more or less approximation to conformity, the older rocks forming the central mass of the mountain." Here follow details aa to localities, for which the reader is referred to the letter as published. * § 183. While, therefore, tlie brothers Rogers held, and odiers still hold, to the paleozoic age of the Lower Taconic rocks, the view put forward by Mather, that the great region of gneisses and crystalline schists with lime- stones, lying to tiie east of these, consists of more highly altered paleozoic strata, had become discredited. It was, ' * Hunt, Azoic liocks, p. 198. •GY. IXI. XI.] THE METAMORPHIC HYPOTHESIS. 663 stage, — the t this Primal ed by Rogers of the Lower rd as in some otsdam of the s showing the rocks, and the make the fol- Rogers, written lume on Azoic of some forty- cross the Blue rinia. In many n of the Lower .), in their con- YLG rocks of the lity of the Cam- e and metamor- ,ous; the lower upon the slope g dips, on the older rocks. In , southeast dips re, are made to fi to conformity, if the mountain." vhich the reader JRogers held, and of the Lower ("Mather, that the schists with \wc- of more highly n-edited. It was, as we have seen, abandoned by H. D. Rogers for Pennsyl- vania, in 1858, and by W. B. Rogers for Virginia, where he recognized in the pre-Taconian rocks the same great divisions which I had elsewhere pointed out. The history of the studies of Tlionias Macfarlane, ard my own, wliich showed conclusively the pre-paleozoic age of the exten- sion of the New England crystalline schists i'.ito the Province of Quebec, has already been told elsewhere.* § 184. It was, therefore, with some surprise that geo- logical stu'.lents found J. D. Dana, in 1880, attemi)ting to resuscitate, in its completeness, the discarded view of Mather. In an elaborate paper on " Tlie Geological Rela- tions of the Limestone Belts of Westchester County, New York," which appeared that year, Dana, following up the reasoning already noticed (§ 161), by which he sought to sustain the paleozoic age of the Lower Taconic rocks, proceeds to assume that the crystalline marbles enclosed in the gneisses, as well as the gneisses and crystalline schists of the region named, are altered rocks of paleozoic age. To quote his conclusions : " The limestone of West- chester County and of New York Island, and the con- formably associated metamorphic rocks, are of Lower Silurian age," and, farther, " the limestone and the con- formably associated rocks of the Green Mountain region, from Vermont to New York Island, are of Lower Silurian age." f His argument in favor of these assumptions appears to be, briefly, this : — That the crystalline lime- stones of the gneissic series, the granular Lower Taconic marbles, and the fossiliferous Cambrian and Ordovician limestones found among the uncrystalline sediments of the Appalachian valley, along the western flank of the crystalline helt north of the Highlands, are but three dif- ferent conditions of one and the same calcareous series, and, lience, that the great area of crystalline rocks south of the narrow range of the Highlands (of which ho * Hunt, Azoic Rocks, pp. 182-188, and ante, pp. 406, 407. t Amer. Jour. Science, 18S0, xx., 455. 664 THE TACONIC QUESTION IN GEOLOGY. [XI. Vif'" admits the eozoic age) consists of paleozoic strata, Cam- brian or Ordovician in age. § 185. Dana, having announced his conclusions as above, adds : " The evidence which has been adduced, though then but partly discerned, led Profs. W. B. and K, D. Rogers, and Prof. W. W. Mather, nearly to i:.e results here reached." In support of this assertion, he refers to Mather's report of 1843, in which, as we have seen, the hypothesis was advanced, and also under the head of " Professors Rogers," to a paper by them in 1841, in the Proceedings of the American Philosophical Society. as well as to a statement in the American Journal of Science for 1872 (vol. iv., p.age 363). Tliis the reader will find to be nothing more than Dana's assertion that the Messrs. Rogers, in that same paper of 1841, main- tained the Champlain age of the Lower Taconic series, — a view which, as we all are aware, one of them, some years later, abandoned for that of its Devonian age. These eminent geologists did, for a time, put forward the view (afterwards relinquished) that the gnel^^sic series of the White Mountains consists of altered Silurian (Oneida- Clinton strata), and Mather, in his argument, made the most of the error of H. D. Rogers, who mistook, in 1840, certain interstratified crystalline limestones among the Primary gneisses of New Jersey for superincumbent limestones in an altered condition; but Dana fails to show that the Messrs. Rogers ever maintained the paleozoic age of the great series of crystalline rocks in southeastern New York, as he would have his readers infer. When, in 1858, H. D. Rogers had occasion, in his final report on the geology of Pennsylvania, to describe the continuation of these same rocks into that State, he distinctly assigned them to a horizon below the base of his paleozoic series, proposing, at the same time, a Hypozoic and an Azoic sj-stem to include them. § 186. The Highland range en the east side of the Hudson traverses Putman County, and, passing southwest- Y. t^- strata, Cam- iclusions as en adduced, 3. W. B. and learly to tLe assertion, he 1, as we have 30 under the ;hem in 1841, ,hical Society. ,n Journal of us the reader assertion that if 1811, maui- jonic series, — L)f them, some Devonian age. it forward the eissic series of urian (Oneida- ent, made the Istook, in 1840, es among the uperincumbent a, fails to show le paleozoic age n southeastern Ifer. When, in il report on the jontinuation of [inctly assigned [aleozoic series, and an Azoic ist side of the tsing southwest- XI.] THE METAMOKPHIC HYPOTHESIS. 665 ward to the river, occupies but a small area in the north- west corner of Westchester County. Along its southeast base, at Annsville, and at Oregon, is met a narrow belt of scarcely crystalline limestone, accompanied by an argillite or talcoid slate, and resting unconformably upon the ancient gneiss. This belt, apparently a Lower Taconic outlier, is regarded by Dana as partially altered Lower Silurian, and "the grade of metamorphism " is declared by him to become more intense to the south and east, giving rise to the whole gueissic area of Westchester and New York Counties. The gneisses and conformably interstratified crystalline limestones of this large area are, as we have seen, supposed by Dana to be metamorphosed Lower Silurian, though they are really undistinguishable from the rocks of the adjacent Highland range, which he admits to be Archeun or Primary. In support of his startling proposition, Dana might be expected to point out some distinctions between the rocks of the two areas. He begins by suggesting certain differences as to more or less micaceous or hornblendic gneisses in the two regions in question, but confesses that "there are gradations between tut two, in both respects, which make the appli- cation of a lithological test very perplexing," * and admits that " the lithological evidence of diversity of age is weak," a criticism which is equally applicable to Dana's stratigraphical argument. I am familiar with the rocks of many parts of West- chester County, and since the publication of Dana's paper in 1880 have taken repeated opportunities to examine the rocks called by him Metamorphic Lower Silurian, in various localities, as at Sing Sing, Tarrytown, Yonkers, Spuyten Duyvil, and Kingsbridge, along the Hudson. I have also studied the same rocks farther to the east, along the River Bronx and the Harlem Railroad, to Pleasant- vale, as well as between this line and the Hudson, and have crossed eastward to Long Island Sound, and examined * Amer. Jour. Science, 1880, xx., 373. ( (_i iiii II ' wtk \ 11 w 1! ti' . 1 1 '1 i i i. me THE TACONIC QUESTION IN GEOLOGY. [XI, (Vii i4ti 1 1' the exposures on the shore at and near New Roehelle. Being already familiar with the Laurentian rocks through- out Canada, as well as in parts of the Adirondacks, and in the Highlands from Putman County, New York, through New Jersey and Pennsylvania to the Schuykill, and beyond, I do not hesitate to say that the gneisses and their associated crystalline limestones of Dana's so-called Metamorphic Lower Silurian, in Westchester County, cannot be distinguished from the typical Laurentian. I believe that tlie judgment of an impartial observer would be that the notion of any difference between the Lauren- tian gneisses and limestones of the areas mentioned, and the gneisses and their interstratified limestones of West- chester County, has no foundation in fiict. § 187. Passing now from Westchester County to the adjacent Manhattan Island, the same Laurentian gneiss is seen in its northern portion, between Seventh and Eighth Avenues, especially in a cutting at One Hundred and Fortv-fifth Street, and thence in a ridge some distance farther south, the strata being nearly vertical and of grayish hornblendic gnei&s, and a band of crystalline lime- stone appearing a little farther to the east, on Harlem River. A quarter of a mile to the west of tl is ridge, in Mount St. Vincent, is seen a distinct type of highly mica- ceous gneiss, and mica-schists, and similar rocks are exposed at intervals in the western part of the island, as far south as Fifty-ninth Street. Farther eastward, in tiie southern part of Central Park, just above Fifty-ninth Street, the numerous rock-exposures are all of similar mica-schists, and micaceous gneisses, often at moderate angles. They include endogenous granitic veins, occa- sionally presenting in their structure a marked bilateral symmetry, ond sometimes tranverse, but at other times interbedded. Several perched blocks here found are of similar endo-renous granite, and are apparently boulders of decomposition, left in the sub-aerial decay of the rocks of the region. These micaceous rocks are unlike those of Y. [XI XI.] THE METAMORPHIC HYPOTHESIS. 667 \v Rochelle. ;ks througli- acks, and in jrk, through luykill, autl Tueisses and Ill's so-called ter County, lurentian. I server woukl the Lauren- jiitioned, and nes of West- lounty to the itian gneiss is h and Eighth Hundred and some distance rtical and of ystalline lime- ,t, on Harlem tV.is ridge, in f highly mica- ar rocks are the island, as Isi-ward, in the re Fifty-ninth [all of similar li at moderate [c veins, occa- irked bilateral tt other times found are of [ently boulders ly of the rocks inlike those of Laurentian areas, but, on the contrary, closely resemble those of the White Mountains, and of Pliiladelphia, which I have called Montaibau, and are like tlie younger gneissic series of the Alps and the Scottisli Highlands. I there- fore, as long ago as 1871, * noticed these rocks as belong- ing to this younger series, and have since expressed the opinion that the Laurentian "of Manhattan Island appears to be overlaid in parts by areas of younger gneisses and mica-scl lists, the remaining portions of a mantle of Mont- alban." f It is, however, by an error for which I am not responsible, that in James Macfarlane's " Geological Rail- road Guide," in 1878, the Montaibau of Manhattan Island has been represented as extending upward along the Hudson River Railroad by Spuyten Duyvil, Yonkers, Tarrytown, and Sing Sing, as far as Croton, before meet- ing the Laurentian of the Highlands. There appears, nevertheless, to be an outlier of Montalban rocks at Cruger's Station, just above Croton, and there may be others in various parts of Westcliester County. § 188. It has been deemed necessary to notice thus at length, in this coiniection, Dana's resuscitation of the ancient views of Mather, for two reasons : first, because therein, both the Lower Taconic rocks and various crys- talline rocks just noticed, are supposed by him to be con- tiguous portions of the same Cambrian and Ordovician (Lower Silurian) sediments in different stages of trans- formation ; and secondly, because the manner in wliich the names of the brothers Rogers are cited by Dana in conjunction with that of Mather is such as to lead the reader to the false conclusion that those eminent geologists supported Mather's hypothesis of 1843 as to the Cambrian and Ordovician age of these same crys- talline rocks, as well as that of the Lower Taconic series ; which latter view, as we have shown, W. B. * Hunt. President's Address before the Amer. Assoc. Adv. Science, 1871; in Cliem. and Geol. Essays, pp. 24S and 197. t Smithsonian lieport for 187S, Progress of Geology. !}!»-* 668 THE TACONIC QUESTION IN GEOLOGY. [XI. Rogers repudiated a few years afterwards, in 1851, and again in 1860. § 189. The rise and fall of the doctrine of regional metaniorphism, which is but an extravagant development of the Huttonian hypothesis of the origin of crystalline rocks, forms a curious chapter in the history of geology. I have elsewhere related the early application of this doctrine to the crystalline rocks of Mont Blanc by Ber- trand, about 1797, and its subsequent restatement by Kef- erstein in 1834, until it was taken up and popularized by Lyell, Murchison, and various continental geologists, so that the view became generally accepted that the gneisses and mica-schists of the Alps an- but altered secondary and tertiary strata. The story of the refutation of this hypothesis for the Alps by the studies of Favre, Pillet, Gastaldi, and others, has also been told.* A similar view was extended to crystalline rocks in other parts of continental Europe, in the British Islands, and in eastern North America, save that to all of these a paleozoic age was generally assigned. The opinions of Mather on this subject were adopted by Logan and others, including the present writer. Tlie brothers Rogers, in 1846, advanced a similar view for the rocks of the White Mountains, but abandoned it before 1858. It was not until 1870 and 1871 that the present writer, rejecting entirely the views of this school, asserted the pre-Cambrian age of all the great areas of crystalline rocks, alike in North America and in Europe. Nearly coinciding in time with this, came the independent action of numerous continental geolo- gists, including those already named, and the result has been such an advance of the views of the new school that, in 1881, Callaway could say that " every case of supposed metamorphic Cambrian and Silurian has been invalidated by recent researches," and in 1883 Bonney wrote that the hitherto accredited "instances of metamorphism in * Amer. Jour. Science, 1872, iii., 9, and Chem. and Geol. Essays, pp. 338-342 and 347, 348. Also, ante, Essay X., part iv. . . ,, . ,, . , XI.] THE METAMOllPHIC HYPOTHESIS. 669 Geol. Essays, pp. Wales, and especially in Anglesey, in Cornwall, in Leices- tershire, and in Worcestershire, have utterly broken down on careful study;" * as had already been the case in the Alps and in North America. § 190. The last stronghold of the metamorphic school in the British Islands was in the northwest of Scotland, where Cambrian and Ordcnician fossiliferous sandstones, limestones, and shales, resting upon the ancient granitoid gneisses to the west, are, towards the east, overlaid, in apparent conformity, by a great series of unlike gneisses and mica-schists, which form the Scottish Highuinds, and were declared by Murchison and Archibald Geikie, from their studies, to consist of still newer rocks in a so-called metamorphic condition. The structure of this north- western part of Scotland was, in fact, according to their teaching, the procise counterpart of that of New England, as formerly taught by Mather and his followers, and still supported by Dana. The late Professor Nicol, however, constantly opposed this view of the structure of the Highlands maintained by Murchison and by Geikie, while the present writer, from his lithological studies of the Highland rocks, declared in 1871 his conviction that the upper gneisses of " the Scottish Highlands will be found ... to belong to a period anterior to the deposition of the Cambrian sediments, and will correspond with the newer gneissic series of our Appalachian region," f then descri- "ed as the White Mountain series, — an opinion which was reiterated, after farther examination of speci- mens of the rocks, 'ia a communication in 1881 to the Geological Society of London, when these Highland gneis- ses were designated as Montalban.J § 191. The studies by Hicks of the geology of parts of this region, from 1878, and the later and independent * Callaway, Geological Magazine, Sept. 1881, p. 423, and Bonney, ibid., Nov., 1883, p. 507. t Hunt, President's Address before the Amer. Assoc. Adv. Science, 1871, and Chem. and Geol. Essays, p. 272. } Froc. Geol. Soc. London, in Geological Magazine, 1882, ix., 39. \r. ,i! I I 670 THE TACONIC QUESTION IN GEOLOGY. [XT. I t; *fcb f ^1 ones of Callaway and of Lapworth in other districts, had already, in the beginning of 1883,* shown the fallacy of the views maintained by Murchison and Geikie as to the geological structure of the Highlands. The united testi- mony of these observers made it clear that in the region in question were portions of two gneissic series, — an older or granitoid gneiss, like that of the western coast, and a younger, very distinct in type, which has been variously designated as Upper Pebidian, Grampian, and Caledonian, and is that described by me in 1871, and again in 1881, as of the White Mountain or Montalban type. This, the younger gneissic series of Murchison and Geikie, was clearly established to be of great thickness, and older than the fossiliferous Cambrian, which it is brought to overlie by a series of great folds, overturned to the west, and accompanied by parallel faults, with upthrows on the east side, as shown by Hicks in Ross and Inverness shires, as well as by Callaway in Assynt, and by Lapworth in Eriboll (ante., p. 423). § 192. The concordant and independent results of the eminent observers just named having thus demonstrated the fallacy of the view of Murchison and Geikie that the gneiss, which in the Highlands overlies the fossiliferous strata, is a still younger paleozoic series in an ?ltered con- dition, the geological survey of Great Britain, of which Geikie is now director, undertook, in 1883 and 1884, a re- examination of the region in question. The result of this has completely disproved the former statements of Mur- chison and Geikie, and has confirmed those of the new school. Geikie, in a note recently published,! tells us * Hicks, Qiwr. Geol. Jour., 1878, xxxiv., 816; Geol. Mag., 1880, vi.; also Quar. Geol. Jour., 1883 (with appended notes by Bonuey), in ab- stract in Geol. Mag., Marcli, 188o, x., p. 137. Callaway, ibid., x., pp. 139 and 330; and Lapworth, ibid., x., pp. 120, 192, 337; also Callaway on Progressive Metamorphism, ibid., May, 1884; and summaries in accounts of the Progress of Geology in the Reports of the SiuitUsouiau Instituticu for 1882 and 1883. t Nature, Nov. 13, 1884, xxxi., 22-35. .OGY. [XI. r districts, had the fallacy of eikie as to the he united testi- ; in the region jries, — an older rn coast, and a been variously md Caledonian, gain in 1881, as ^pe. This, the id Geikie, was iiess, and older t is brought to led to the west, .ipthrows on the [nverness shires, by Lapworth in it results of the 18 demonstrated Geikie that the the fossiliferous an plteredcon- ritain, of which and 1884, a re- le result of this ements of Mur- lose of the rew ished,t tells us eol. Mag., 1880, vi.; by Bonney), in ab- ilaway, ibid., x., pp. n ; also Callaway on mmaries In accounts tbsoniau Institution XI.] THE METAMOUPHIC IlYl'OTIIESIS. G71 that he has "found the evidence altogether overwhelming against the upward succession, which Murchison believed to exist in Eriboll, from the base of the Silurian strata in- to an upper conformable series of schists and gneisses," and adds: "Tliat there is no longer any evidence of a regular conformable passage from fossiliferous Silurian quartzites, shales, and limestones, upwards into crystiilline schists, v/hich were supposed to be metamorphosed Silu- rian sediments, must be frankly admitted." The same conclusions are also reached by Geikie from the re-exami- nation of the similar sections in Ross-shire, previously described by himself in accordance with the views of Murchison. The preliminary rei)ort of the surveyors, Messrs. Peach and Home, which is subjoined to the director's note, shows the same structure as was already described by the late ob-servers, namely, overturned folds and great faults, with lateral thrusts westward, by which the gneisses are made to overlie the fossiliferous strata, — the horizontal displacement of the gneisses to the west, which are superimposed on the Cambrian rocks, being, in some cases, according to Geikie, not less than ten miles. [Judd, who, in 1885, reviewed before the British Asso- ciation the early work of NicoU in this region, writes that in a paper by the latter, published by the Journal of the Geological Society in 1861, he "must be admitted to have established the main facts concerning the geology of tiie Highlands as accepted by all geologists at the present day." He adds that the conclusions arrived at by Nicoll in early as 1860, those of the later investigations named, previous to 1883, and those of the British geological survey in 1883 and 1884, "are in all their main features absolutely identical, and the Murchison ian theory of the Highland succession is now, by universal consent, aban- doned.*] § 193. Geikie notices the distinction between the older or granitoid gneiss, portions of which also appear in the .1 : * Nature, xxxii., 455, 45G. I i m '■ .1'; 672 THE TACONIO QUESTION IN GEOLOGY. [xr. i 9 M f ! fe ^9 '!.. II JTT IJH HflPPHPIIIIH [9^B ' i' 1 '"'>'9| ' %M '^4 ' ' ' ' Highlands, and the upper gneissic and mica-schist series, the pre-paleozoic age of which was sliown by the observa- tions alike of Hicks, of Callaway, and of Lapworth. He calls attention to the laminated and schistose structure developed by the great pressure and friction along the lines of movement in gneissic and hciiblendic rocks, and also to similar changes produced by the same agency in detrital rocks, such as arkose. Both of these structural alterations are apparently included by Gcikio under the head of what he calls a "regional metamorphism," — a misapplication of the term likely to confuse the reader, since local structural changes, induced by mechanical movements in ancient crystalline rocks, have nothing in common with that mysterious process which has been supposed by the metamorphic school to generate similar crystalline rocks from uncrystalline sediments. As re- gards the changes wrought by the same agency on detrital masses, it may be repeated that "the resemblance between primitive crystalline rocks and what we know to be detri- tal rocks compressed, recemented, and often exhibiting interstitial minerals of secondary origin, is too slight and snperficial to deceive the critical student in lithology, and disappears under microscopical investigation" (^ante, p. 108.) § 194. We have already elsewhere in this essay (§ 135) referred to the local development of crystalline silicates in sedimentary rocks by infiltration, and have in another place considered the relation of such a process to the question of the origin of Primitive crystalline rocks. These we believe to have been formed anterior to the existence of detrital sediments, and by a process which excludes alike all so-called metamorphic, metasomatic, and plutonic hypotheses of their origin. At the same time, we reject the Wernerian or chaotic hypothesis; and its modification by De la Beche and Daubr^e, which we have called thermochaotic, in favor of a new aqueous or neptunian hypothesis, which supposes the elements of Y. XL] THE TACONIC SERIES. G73 chist series, the observa- (WortU. He vjo structure u along the c rocks, and le agency in se structural, ie under the rphism," — a e the reader, f mechanical e nothing in leh has been lerate similar 3nts. As re- icy on detrital lance between )\v to be detri- ,en exhibiting ;oo slight and lithology, and on" (ante, p. essay (§ 135) ^ne silicates in \iQ in another Lrocess to the ktallin.e rocks, literior to the {process which metasoniatic, At the same lypothesis; and |v6e, which we Iw aqueous or elements of tlicse rocks to Imve been dissolved, and bronc;ht to tlu' surface IVdiii a disintofifiuted layer of Ij^mu'ous basic rock, the sn[H'rtiL'ial and last-solidiliud poilioii of u cooHmlj globe, through tlu* action of circulating waters. The soluble and insoluhle products of the sub-uerial decay alike of igneous and aqueous rocks are, however, con- ceived to have intervened in the process, especially during the period of the later crystalline or Transition rocks. This exi)lanation of their genesis, which we have called the crenitic hypothesis, is discussed at length in Essay \. of the present vohuue. IX. — CONCLUSIONS. § 195. The task attempted in the preceding chapters, of discussing the history of the Taconic question, has involved a review of much of the work done in American geology for more than sixty years, going back to the labors of Eaton, and even to those of Maclure. Of tiie somewhat extensive literature * of the subject I have made use, so far as has seemed of importance, in the con- troversies whicli have arisen on this question, and have supplemented the researches of various investigators by personal observations extending over a wider field and a greater number of years than those of any of my prede- cessors. From all of tliese sources, I have here sought to bring togetlier whatever has appeared to be of value for the elucidation of the important problems before us. In the following sections, the conclusions which have already been set forth at length are summed up. § 196. There exists in eastern North America a great group of stratified rocks, consisting of quartzites, lime- * Dana, in the Amer. Jour. Science for 1880, xix., 163, has given "a list of tlie principal papers" on the Taconic System, in wliich, wliile pro- fessing to bring togetlier those adverse to the pre-Cainbrian age of the Taconiau, he omits all reference to the opinions of Ailams, of Kd. Ilitch- coclc, and the later conclusions of \\. B. Rogers as lo tlie (Upper) Silu- rian or Devonian age of the Taconian limestones. 'J'he list is in other respects very incomplete, and serves to mislead the student. 074 THE TACONIC QUESTION IN GEOLOGY, rcr. ( ' it ^ stones, argillites, and soft crystalline schists, which hiivo together a thickne.ss of 4000 feet or more, and are found resting unconforniably ui)()n various niore ancient crystal- line rocks, from tlie Laurentian to the Montalban inclu- sive. This series, called Transition by Maclure, includes the Primitive Quartz-rock, the Primitive Lime-rock, and the Transition Argillite of Eaton, and is the Lower Taconic of Emmons, and the Itacolumitic group of Lieber. The series, which I have preferred to cull Taco- nian, is essentially one of Transition crystalline rocks. The (juartzites, which predominate in the lower portion, contain much detrital matter, and are sometimes conglom- erates. They are, however, often vitreous or granular, the latter variety being sometimes flexible and clastic, and constituting what is called elastic sandstone or itacolu- mite. These quartzites, like the limestones of the series, often contain an indigenous micaceous substance, which is in most cases a hydrous muscovitic mica, related to sericite or to damourite. A similar mineral predominates in certain layers of soft unctuous lustrous schists, which, from their aspect, have been called talcoid or magnesian, and are found intercalated alike v ^^^ the quartzites and the limestones of the series. TliC latter, often more or less magnesian, are generally finely granular, and yield marbles for statuary and for architecture. They are often variegated in color or banded with green or gray, consti- tuting cipolins. The mineralogy of the limestones and their associated crystalline schists has been noticed in §§ 51, 65, 68, 76-79, and farther, on page 184 of the present volume, and it has been shown that the Taconian is an important ore-bearing horizon, including, besides great deposits of magnetite, and of hematite, others of siderite and of pyrite. Both of the latter species, by epi- genesis, give rise to hydrous iron ores, which, throughout the Appalachian region, characterize the outcrops of the series, and are generally imbedded in clays, the result of the sub-aerial decay of the enclosing schists, which, it may OY. P*'' 3, which \iivve lul are found icient crystal- iitalbivn iiiclu- eluve, indudes /ime-rock, ami is tho Lower itic gioup of d to call Taco- ^•stallino rocks, lower portion, itimes congh)ni- ov granular, tho lud elastic, and tone or itacolu- les of the series, ,ubstance, which mica, related to -al predominates 9 schists, which, d or magnesian, 3 quartzites and r, often more or uular, and yield They are often or gray, consti- limestones and been noticed in 5age 1B4 of the at the Taconiau .eluding, besides matite, others of sv species, by epi- ■hich, throughout outcrops of the xys, the result of Lsts, which, it may XI.J TllK TACONIC SERIES. fi76 theuco 1)0 conjectured, inchido, in nuiny cases, huge pro- portions of a ields|)iitl,ic mineral. The argillites of the Tai!onian, often yichliug rooluig-shiteH, are intcrstratiru'd with more! or less silicious beiU, and o(!cur chielly in tlie upl)er part of the series. § 11)7. These Taconian rocks are not confined to the Appahichian vall(;y. Extending southward therefrom, tliey are traced in Pennsylvania ah)ng the eastern base of tho lilue Kidge into North Carolina, and are found in outliers to the east, over the Atlantic belt from (Jeorgia to New Brunswick and Nova Si'otia. To the west of tlie great valle}', tliey are known to underlie the eastern part of the paleozoic basin, and appear in eroded anticlinals from beneath tiie coal-measures, alike in Alabama and Pennsylvania, where they are overlair('visions of facts veriiied by subsequent observation or experiment, and (iiitliy upon their simplicity, or rather their reducing power," — btalio, in The Con- cepts and Theories of Modern Physics, p. 85. I 'I .Hi hi! I 'I 680 THE TACONIC QUESTION IN GEOLOGY. m J Taconian rocks are in parts overlaid by mesozoic and by tertiary strata. § 203. As regards the existence in oth^r lands of a similar series of roclKy of Vormont, 030. Adiromliick :Muuiitaiii8, 403, 0'22. Ae^lritf, .11!). Ai'riiil ili'iuiiliitiiiii, 27.5. Agftti's, l'l;i> fair ou, 74. Agrlciillte, ;it!ii. Alubiiiim, (;o()lo(,'y of, 357, noO, 503. AlbercKo, -liHi. Alburtl, cniptivo rdck-.snlt, 00, Alblte, arlitlnial fcinuiillon of 157, 605. Ali'xaiidriii, sidiiidl of, 'J. Alkaliiif) Hilk'.ilii) in toniiatlou of oxyda, ir>0, 181, 2»i). Alkaliiio waters, 218, note, AUaiiito gr(iu|i of tilllcati'S, 346. AUiiian, protoplasm, IH. Ali>s. — Apuaii, 473, 477, JS3, 084; East- ern, 4fi.'> et .1(7/. ; LiRurian, 475 ; Mari- time, 473, 47."> ; Wt'Storn, 45H et .lei/., 48.') ; general nooloKy of, 457 rt ««(/., 083 et aeq.; (iastaldi on, 45H et arij., 4.81; I Gerlacli on, 4,")9, 407 ; Von Hauer on, 45!l, 40.'), 471, 481 ; Lory on, 4(!0 el seq.; Staplf on, 470 et get/. Altered (.'liamplain division, 008, 0.55, O.'i0, 057 ; (.)uol)cc (,'roup, 400, 410, 010; Ulld- Bon-KiviT group, JOG, 000. Alundna, relal ions of, 100, 300. Alundnoiis silicates, 37, IL'0, 1,53, 100, 183; • dissooiiUion of, 148, 1,")0, 240. Andantlioiil silicate from Portsmouth, . K. I., 19.5, ,?,5!t. Amorphite, order of, Weisbach, .381. Amphibole, 147, 310; its relations to woUastonite and jiyroxene, 290, 330. Amphiniorphic rocks, 485, 487. Aniygdaloids of Faroe Islands, l.'!5. Analcito, artlticial formation of, 157, 505. Anaxagorns, 4, 7. Ancient gneiss. See Gneiss, older or cen- tral. 689 Andaliislto, 3t.fl, 412, 478. Angli'ney, geology of, 1 11-1, 417. Anindkie group of rooks, 411, 578 ct $cq.f Oil, 01,'!. Antliiaeoids, tribe of, .380, Antliraciie of Pennsylvania, Oil, 521 ; of Itliode Nland, 105, X,{), Antlipiity of rock-doeay, 2,50, 200. Apatite, In <-'anada, 225 el sii/.; veins of descritpi'd, 1/32 ; in Norway, 2.10. Apennlnei, geology of, l,"(> it w/.; .lor- vis on, t77 ; (ia^*laldi on, 48;!. Appalaelii;in valley, linionilis of, 201 et nei/., 5,15 ; Taeoniaii in, ,5,"i0 ; Cambriau of, ,580, 020. Apuan Alps. See .Mps, Apuan. Arab pliysicians, 10. Arago, tile inttM-steiiar medium, 03. Areluean or ICo/.oie roeks, 01, 402. Ardennes, gi^ology of, A'2i ; wlietstonea of, 425, .t.ile. Ardennile, a vanadosilicate, 347. Arenig roeks, t!25. Arfvedsonite, .'!4!t. Argiliite, Transition of Katon, 519; it und(Tlios tlie First Graywacko, 587; in Minnesota, ,578, ,580. Argilloids, tribe of, 318 ; table of, 370. Aristotle, 1, 2, :i. Arizona, geology of, 014, 024 et seq. ArsenopyriloiilH, tribe of, 378. Arvonian roeks, 409; AVales, 418 c^ .'"''7. ,' Pennsylvania, 540 (7 seq.; Wiseonsin, 540; Mlsso\irl, 103,409; Atlantlecoast, 408 e^ .'I''/., 547; Scotland, 424 ; possibly in the Alps, 515, nule. Ashburner, C. A., anthracite seams, 611. Asphaltolds, tribe of, 380. Astronomy, its object, 27. Astrophyllite, a zireonle niiea, 355. Atlantic belt, decayed rocks of, 250 el aeq,; Messrs. Uogers on, 544, 001. ) ;; 1 \ ,J t :ti ■ 1 ■ ( t1 f I i •!«3 mW- mm ■ WS^*'' ' t , 'i ^H MLJ ' v« 1 iy --'-9 j M ^H&< *l ' rf ' w tjpif^ G90 INDEX. Atmosphere, cliomical and geological relations, of, 30 et scq. ; composition ami weight of, 31 ; Hecular changes in, 34; relations of to climate, 43 e< sej.; to a cooling planet, 47. Atomic formulas, i.'U2 ; notation, 302 et seq,, 31!) ; weiglits, table of, 320 ; sym- bols, ibid.! volumes (see Molecular volumes). Auerbachite, .300. Augite. See Pyroxene. Auriferous gravels of California, 272. Auroral lime; one of Hogers,532 et seq. ; thickness oi, 037, 041. 557 ; in Alabama, 557 J is Tacoiiiaw, 532, 535; Lingula of, 582. Axinite, 138, 347. Azoic series of Uogers, 405, 544, 6C1 ; of Foster and Whitney, 405. Bac.inqtonitk, 340. Bacon, Francis, on activity of matter, 20, note. Bailey, L. W , geology of New Bruns- wick, 407. Bandeil structure in eruptive rocks, 201, 210 it seq. ; in veins, 224 et seq. See Lamination and Veinstones. Barabno Kiver, Wisconsin, 54(5. Earlier, G. F., life, lit. JJarranile, .1., the Taconic system, 635. Barrois, Ch., geology of Spain, 085. Barsowite, 142. Basic rocV.s, secretions of, 134, 135 1< seq., 220, ,307. Basalt, Mutton on, 74 ; suggested origin of, 116, 207; of Colorado, 135; Bun- sen's normal, 129, 189, 212 ; Durocher's, 212. Bastard, unt., conglomer.ates of, 577. Bauxite, 376 ; its relation to corundum, 604. Becker, O. F., cited, 294 ; law of cooling, 245, note. Bccraft's -Mountain, 032. Belir, Arno, dextrose, 504. BeloBil :\Iountain, 005, 632. Belt, Thos., eruptive quartz, 95; min- eral veins ibid.; death of, ibid,, note; displacement of decayed rocks, 275, note. Beroldinger, granite, 82. Borthier, rock-tlecay, 31, Berkshire Co., Mass., geology of, 653 et seq. Beryl, its fusion, 290, note; analyses of, 347 ; change to kaolin, 370. Berzelius, artificial formation of zeolites, 155 et seq. ; chemical system of miner- alogy, 282 et seq. ; dynanuds, 13, Bieilese, Italy, ancient gneiss of, 402 ; pietre verdi of, ibid.; younger gneiss of, 403 ; syenite of, ibid. ; serpentine of, 496. Bigsby, J., Huronian, 407 ; two series of crystalline rocks, 600. Billings, E., Lower I'otsdam, 618, 624,038 ; Cambrian of Nowfounilland, 025 ; Levis limestone, 618, 633 ; Ordovician of Farn- ham, 606 ; section to Bridport, Vt., 605. Biotic (Biotics), 17, 18, 28; relatio'- to organograt)hy, 2»<5. Biology, its object, 17. Biophysiology, 21. Birdseye limestone of Eaton, 526. Bischof, G., metasomatism, 83, 103, 200, 498. Bismuthic oxyd, 306 ; silicates, 322, 365. Bismutoferritc!, .360. Blair Co., Ptnn., geological section In, 537. Blake, W. P., calcareous, veins, 229; rock-decay, 248. Biake, T. M.. and Johnson, kaolin, 369. Blue Rl^lge, decayed rocks of, 251, 258; geology of, 656, .559 et seq. Boltonitfa, 507. .SVe Chrysolite, Bomlvshell ore, its origin, 202. Bonney, T. G., serjientines as igneous, 4i0etseq.; Ihorzolite, 508 ; metasoma- tism, 407, note; metamorphism, Ofis ; serpentines of Cornwall, Eng., 449 ; of Italy, 452, 495 ; of Scotland, 510 ; ser- pentine breccias, 453. Bornemann, geology of Sardinia, 476. Borates, water inrfused, 220. Borosalinoids, tribe of, 380. Boric oxy, ;is9, 087 ; in mineralogical dasr-ification, 285 et scf/., 318, ,'189; of prinievii! earth, 114, 117; its relation to physics, 15 et seq,, 390. Chester Co., Penn,, serpentines of, 437 et seq.; limestones of, 519, Cl.ickis, I'enn., section at, 644. Chlorids, sub-order of, 380. ih f! 692 INDEX. Chlorites, 355 et seq. Chlorine in silicates, 293, 303, 324, 341, 342, 344. (;hoii(lr.,.'-te, 145, 328, 507. Chromic minerals, 436, 508. Chrysolite, 92, 100, 101, 145; sub-aerial decay of, 505 ; relation of to serpen- tine, 450, 503, 505 ; formed from serpen- tine, 500, 513; by igneous fusion, 209, 219, 220, 333, 506 ; by aqueous action, 220, ,504, ,507 et set}., 513, 510 ; roclia of Norway, 508 ; of Nortli Carolina, ibid., 560 ; of the Pyrenees, .TOS ; of New Zea- land, ihiff. ; of Mt. M.a, .509 ; of Saxony, 479 ; in dolerites, 211, 506 ; analyses of, 212 ; in lav^vs, 213. Chrysotile, .324, 4.53. Church, A. H., density of zircons, 366. Cicero on phy.siology, 2. Cipolliiio marbles, ,5.54, 674. Clays, genesis of. 254, 308. ,370 et aeq. Clarke, P. W., cosmic evolution, 47, 55. Classes in mineralogy defined, .382. Clerk-Maxwell, dissociation, ,54. Cleve, geology of Porto Rico, etc., 683. Cleveland, P., Werner's mineral system, 280, note. Clitford, W. K., dynamics, 13 ; molecular motion, 15. Coal-seams displaced, 511. Cobalt in iron-ores, .5.36. .5,53, 569, Cobalt-ammonias, 386. 392 Cobequld Hills, geology of, 573. Coldbnt k rocks, 407 et seq. Colloids, 314, 316 et seq. : relation of to life, 18 et seq. ; sohibility of, 168 ; changed to crystalloids, 375 ; igneous and aqueous, 362, 374 ft neq., 383 ; limits of species in, 398. See Porodini and Porodic. Colloidal rocks, 192 ; liquids, 221. Colorado, Table Mt., 135 ; granitic veins of, 223 : Grand Caflon of the. 624. Conglomerate, in ancient rocks, 110, 183, 2.54 et seq., 479, 577. Contraction of cooling globe, 241. Connecticut, rock-decay In, 248 ; granite in, 211, no^e. Concretionary rocks, 222V/ seq., 234. Condensation in mineral species, 166, 285 et seq., 305, 390 et seq., 391, 687 ; relation Of to hardness, 285, 299, .304, 395; to chemioul indifference, 286, 299, 304, 395, 687. Cook, G. H., 618; geology of New Jersoy* 670, 590 et seq., 658. Copper, ores of in Blue Ridge, 268 et seq.; mesozoic, 260; theory of its con- centration and depo'dtion, 259 et seq,; native with zeolites, 139; in Keweenian, 260. Corniferous limestone of Eaton, 527. Coronite, a mag-iesian touriualine, 162, 350. Corrugation of crystalline strata. 111, 179, 241, 243. Corundum, 100 ; genesis of, 239, 240 ; artificial production of, 300 (note), 504. Cordier, origin of lime-carbonate, 36. Corsica, rock-deoay in. 276 ; serpentinea of, 474 ; granites of, 475. Cornwall, Eng,, cryst.nlline rocks of, 449 j serpentines of. ■)'"<, 449. 510. Cornwall, Penn., iron-ores of. 5.''6 (note), 650 et seq. ; serpentines of, 442. Cosmic evolntinn, 47, ,5.-;, .56. 59 ; dust, 61. Cossa, It.ilian serpentines, 4,54, 484. Cosmos, the. 26, vote: Humboldt's, 22. " Cotgrave, physiology, 3. Coticulite, Rt^nard on, 425. note. Crenitic hypnihesis of crystalline rocks, 132 et seq., 175. 216 et seq,, 199, 2.38, 241, 673 ; action, chances by, in plutonio rocks, 216 ; in crenitic rocks, 186, 217 ; compared with eliquation, 217. Crediier. H.. 405 : geology of Saxony, 256, 479 ; of Lake Superior, 581. Crinoids, silicntes in, 193 et seq. Crosby, W. O., geology of West Indies, 681 et seq. Cross and Hildebrand, zeolites, 135 e/ seq. Crystalline admixtures, 294,296, ,304. 342. Crystalline rocks defined, 191 ; origin of, 68 et seq. : various hypotheses regard- ing, 82 et seq.. 104 ; a new hypothesis, 112 et seq. ; three propositions relating to, 125; universality of, 107, 110; au- thor's early studies of. 112; succession in time, 106, 678, OSS ; .secular changes in, 187; great groups of, 1S4 ; inclined strata anan roclis, 91 ; erup- tive granites, 90; a heated ocean, j /«V/. ; eruptive limestones, 90, 228; pseudomor- phism, 100, 101, note; pinite, in?., 10.'); his .System of Mineralogy, 282 ; .adopts and then rejects tlie >;atural-IIistovy method, 282 et seq. ; atomic volumes, 303, note ; metamorphism. 063 ; grades in, 065, 678 ; crystalline rocks of .soiilh- eastera New York, 663 ; his arguments examiued, 665 ; liis citations of Messrs. Rogers, 664; Taconian rocks,612 ef seq. ; Taconic literature, 673, note; Gray- ■wacke series, 651. Dana, J. F. and S. L., rock-decay, 247. l>arton, N. H., Green-Pond Mountain, X. J., 591. Darwin, Charles, laminated rocks, 201 ; rock-decay, 250. Daubr^e, A., Hutton's system, 70; the origin of crystalline rocks, 78, 86 ; the primeval ocean, 79 ; decomiiositicm of glass, 147 et seq. ; alteration of bricks, 152 ; artificial production of pyro.xene, quartz, and mica, 149 et seq. ; miner- alogy of thermal springs, 150 et seq. Davidson, Thos., Rosmini, 16, note. Dawboii, J. W., on silicates in organic forms, 193 et .teq. : Kozoiin, 231, 573, ,575, 676 ; a fossil sponge, 678 ; organic forms from Hastings rocks, 575 et seq. ; Eozoic rocks, 402 ; Cobequid Hills, 573 ; Keweenian, 012. Decay of rocks, 31 et seq., 127, 246 et seq., 677; Hoosac Mountain, 2.'i6 ; Connecti- cut, 249 ; Pennsylv.inia, 251 ; Georgia, 258; Blue Ridge, 251, 258; Corsica and Norway, 276 ; limestones, 249, 250, 205, porphyry, 209 ; serpentines, 208,441; dolerite, 271 ; auriferous gravels, i;7- ; preliminary to erosion, 252, 277 ; bould- ers formed by, 183, 247 et seq., 254, 257, 272, 276, 278. Deerfield, Ma.ss., diabase of, 138, 1.30. Delaware \Vater-(iap, sections at, 538, Delesse, A., ou crystalline admixtures, 204; origin of serpentines; liis earlier and later views, 431 et seq. ; luetamor- pliisin, 432. De Luc, aqueous origin of rocks, 69. De la Beohe, therniochaotic hyiiothesis of crystalline rocks, 77 et seq. ; Hut- ton's system, ibid.; serpentine, 428. Denudation of decayed rocks, 251 et «('(/., 274, 277. Derby, O. A., geology of Brazil, 681. Descartes, plenum of, 57, 02. Detrital rocks, alteration of, 108, fi72,C8s. Devllle, II. Salnte-Claire, dissociation, 5.1 ; soda-dolomite, 171 ; artificial produc- tion of zeolites, etc., 156 ; dissociation of aluminous silicates, ilnd.; crystalli- zation of amorphnus matters, 173. Devonian age of Taconian, supposed, 031. Dextrose, hydrous and anhydrous, .504. Diagencsis, 105; its importance, 173. Diabase, mesozoic, 121.211, wo^', ,338, 440. Dlau'onds, .503, .504, Osl. Diaii.'ore, 377; arlilieial production of, 504. Dichrotte-gneiss, 412, 478, 482. See L> lite. Dieulefait, serpentine, 474, 501 ; ophites of the Pyrenees., 502, note ; views of T. Sterry Hunt on serpentines, ibid. Dillsburg, Pi'nn., iron ores of, 560 tt seq. Diinetian rocks, 417, 419. Dioritic group (lluroniau), of Kominger, 579, .581. Dipyre. .301. SMet seq. Disintegration of rocks, 273, 277. Dissociation, its universal application, 48, 53 ; of b,;iicates, 148, 150, 240. Dislocations of Cambrian strata, 639 et seq. Dolomite, its origin and formation, 171 et seq. Dolerite, decayed, 271; chrysolltio, 211 et seq., 512; analysci of, 212; mica- ceous banded, 211, note. Dorset Mountain, Vt., 031. 1 ' t I \ G94 INDEX. Douglas, James, elemental matter, 49, Hole. l^raiiur, J. W., Arab physicians, 10. J)u tango, GlossariuMi, 7. l)ucktown, Teuu., metalliferous veins of, V£i. I>iilulh, noritesof, 5S0. ]>uinas, J. li., elemental matter, 49, 56 ; niolucular volumes, 302. Dumoiit, geyscrian deposits, 96. Duncan, P. M., interstellar medium, 43. Duroeher, J., two igneous terrestrial niugiuas, 207 ; Comparative Petralogy, 20S ; ulii|uation, 208, 214 <-/ se0. Fluorhydric acid, action of on silicates, 214, 087. Fontiiine, W. M., section in Virginia, 550. Fontaincbleau sandstone, .104. Forcellinus. See Faceiolatus. Ford, S. W., Cambrian in eastern New York, 039. Forchhammer, kaolin, 360. Formulas, chemical, and notation, 302, 312. Fossil searwaters, 253. Fouque and .^lichcl I.(5vy, artificial pro- duction of crystalline silicates, 209, 219, 375. Fouque, analysis of rocks of Santorin, 213. Fournet, rock-decay, 247. Frankfort slati's, 625. FrapoUi, sepiolite, 448. Frazor, I'ersifor, linionitcs, 264 ; geology of Pennsylvania, 439, 540, 590, note ; iron- ores of Dillsburg, 551. Freiesleben, sepiolite, 448. Fremy, artificial production of quartz, 149. Frt5my and Fell, artiflci.al production of connidum, ;!(li», note. Frenirh llroail Kiver, N.C., section on, 559. Frit/.selie, gaylussito, 171. Friel(t.; rosso and verde, 491,492; of Montul'errato, 495. Galen on lliiipoorates, 8, note. Galestro, 400. Galenoids, tribe of, 378. Garrigou, serpentine, .'502, note. Garnet, 340 ; associated with prehnite, 121 ; manganesian, of coticulite, 425, note. Giiataldi, IJart., his geological work, 458 et seq. ; list of piil)lication», ihitl., note ; serpentine-breccia, 453; antiiiuity of Italian serpentines, i'lG ; (.'■>->''"f!''''''' sections by, 459 ; his two groups of crystalline roeks, 4()0 et seq. ; studies in tins liiellesi!,4G2 ; older and younger gneisses, 4(i4 ; pietre verdi, 403 ; its thickness, 405; granitic rocks and por- phyries, 471 ; geology of the Alps and Apennines, 483. Gastaldite, 310 ; relation of to jadeite, 348. Gamliii, fused alumina, 300, note. Gaylussito, formation of, 171. Geikie, Alex., rock-decay, 271 ; crystal- line schists of Scotl.ind, 009, 671 ; meta- mori)liisni, 072. Geikie, .Jas., serpentines, 510. Gems, order of, 281, 315. Gen(.'tic history of crystalline rocks. See 190 et seq. Genth, F. A., metasomatism, 100, 102; pseudoniorphisni, 101 ; diilienon, ibid. Genoa, serpentines of, 452, 485.. Geodiferous I,ime-rock of Katon, 627. Geogeny detlned, 28. Geognn.-y detlned, 27. Geological map of New York, O.^O : sur- vey of New York, 521 et seq. ; Railroad Guitle, .T. Macfarlane's, 601, 007 ; Text- book, Katon's, 518. Geometrical chemistry of II. Wurtz, 395. I 1 1 ! a 696 INDEX. twi* Georgia, State of, Taconian in, 6C3 ; de- cayed gueiiis in, 258 ; Stone Mountain in, 258, 274, note. Georgitt, Vermont, slates of, RM. Geui'giuu, proiioseil division of Cambrian, 528. Gerliardt, Ch., lioniologous or progres- sive series, 289 ; atomic volumes of native oxyds, 376 et seq. Gerlacli, geological sections in Italy, 468 et seq., 473. Geyserian deposits, 96. Gibbs, Wolcott, complex inorganic acids, 387 et seq. ; and Genth on cobalt-ammo- nias, 386. Glan, P., interstellar ether, 64, Glauvil, Scepsis Scieiitilica, 5. Gl.icial drift, origin of, 251 et seq. Glaciation, J. F. Campbell on, 46. GlauzscUief er of the Alps, 464, 467 et seq., 473, 683. Glass, decomposition of by hot water, 147. Glauoonite, history of, 196 et seq, ; analy- ses of, 198 ; origin of, 309, 333. Glaucophane, 310. Glinkite, 509. Gneiss, Werner on, 74 ; Hutton on, 75 ; eruptive, 81, 201 et scg., 403; aqueous, 82, 131: from limestone, 102 el seq.; rela- tion of to granitic veins, 124, 125, 236, 241, 243 ; older, of Xorth America, 107, 404, 412, 665 ; younger of do., 406, 423, 437,442,4806/ seq. ; older, or central, of Alps, 459, 462, 465, 472, 481, 683 ; younger, or recent, of, 46"., 404, 466, 471, 472, 482 ; apparent unconformity of faese, 463, 469, 481 ; Gastiildi on the two gneisses, 464 ; Bojian and llorcynian, 481 ; Lew- Isian. 417. Goessmann, C. A., dolomites of Syracuse, 446. Goroeix, geology of Brazil, 680, Gosselet, rocks of the Ardennes, 421. Gower, 7, vote ; delinition of physic, 2. Graham, Thos., colloids, 19 ; pectlsation, 315, 322, 323, note. Grampian rocks, 423, 670. Grand Caflon group, 614, 624. Granite, Werner, Saussure, and Hutton on, 72; Beroldingenon, 82; the primitive substratum, 80 et seq. eruptive, 90 et seq., 204 ; derived from limestone, 102 et seq. ; aqueous origin of, 116, 131, 177 et seq., 204, 242; water in, 96, 217; banded structure in, 211, no/e ,• of Alps, 471 ; of Klba and Corsica, 475, Granitic aura, 128 ; "juice," 127; vein- stones, 72, 123e< seq., 125; in various localities, 223, 226 ct seq., 657,666; in serpentine, 438 ; in basic rocks, 122, 137 ; genesis of minerals in, 309. Granitone, 451. Granulitc, 411, 439; of Saxony, 202,478. Graptolitic shales of Levis, 595, 608, 625 647 ; of Teach Bottom, I'enn., 690. Gras, Soipion, on serpentine, 428, 432, note. Graves Mountain, Georgia, 563. Graphite in Laurentian, 1U9 ; In Taco- nian, 561, 573, 580, 682. Gray Sandstone of Knimons, 522, 524. Granular Quartz-rock of Eaton. Sec Primitive Quartz-rock. Granular Lime-rock of Katon. See Primi- tive Lime-rock. Graywacke. See First Graywacke anil Second Graywacke. Greenstone (lluronian) seriis, 404, 406, 411 ; group of Lake Superior, 404, 611 ; of Cornwall, Eng., 450. See liurouion and Pietre verdi. Green Mountains, 408, 400, 410, 594, 635 ; serpentines of, 429, 436 ; their pre-Oam- brian age, 619, 620. Greeu-Pond Mountain, 590, 591, 632, 637. Grenvillo series, 412, 413,425. Groton, Conn., banded granites of, 211. Grove, W., interstellar ether, 62. Oaillemin, on pholerite, 367, .'168. GUmbel, geology of Bavaria, 481 et seq., 684. Gypsum with serpentine, 4*4, 448, 467 et seq. ; in Onondaga group, 440 ; in Cal- ciferous Sand-rock, 618. IlAciii':, Arnold, Cambrian in Nevada, 623. Ilaidinger, W., pseudomorphism, 83 ; dol- omite, 172; nietasonmtosis, 200 ; trans- lator of Mobs, 280. Hall, James, on limestones at Port Henry, Kew York, 2.'!8 ; supposed Huronian in Wisconsin, 54i> ; position of Sparry Lime-rock, 587, 608; Hudson-Kiver group, 601 ef ,ieq. Halley, the earth's interior-, 86, note. Halletlinta rocks, 409 et seq., 418, 515, note, 546 ; unconformable to gneiss, 479. Sea Arvunian. ETDEX. 697 :st Graywacke and inbriaii in Nevada, jmorphlsm, 83 ; dol- inatosis, 200 ; trans- ones at Port Henry, poseil Iluronlan in losition of Sparry m ; Iludson-Rlver tcrior, 86, note. «.iei7.,418,5ir),»io^e, 3 to gneiss, 479. .Sec Halloyslto, ir>2, 372. HalKiy, d'Oniallus d*. Sec Omallus d'. Huluidutuo, Older of, 380. Ilanitditi!, a iiiitivo silicato, 104, 334. llardnoss, riMaiinii of to cUoiiiical con- densation, 2W, '^Ki, ;i(l4, 305. Ilarkiiuss, geolotsy of Wiili'S, 410. llartl, C. F., rock-ducay, 1!4»< ; Cambrian in Niiw Uruiiswick, 407, 023. Harrington, II, J., ndncral volns in »Iount Itoyal, 137 ; aimtite, 232, 230. Hastings series of roeks, 414, 574 et acq., 675. Hauer, F, von, geology of Kastern Alps, 458e^sefy.,4G5, 471; two gnoissic series, 405, J81. Hawi'S, (J. W., feldspars, 339; veneritc, 358, note. Hdboit, crystallino rocks, 85; liis plu- tonic views, 430. Hebrides, gneiss of, 417. Heddle, analysis of pllolito, .300. Hegel, clieniisni, 15, 307. Hehnliollz, clieniieal eliaiigc, 15. llelderberg limestones, 521, G0,'», 032. Heniatili! in aniygdaloids, etc., 138; in dolcmitu of Syracuse, 415 ; inXaconian, 530; brown (.scr- I.inionile). Hercyniaii gneiss, 4!sl ; prindlivo clay- slate, iliUL, 084. Hicks, IT., geology of Wales, 410, 418 ; of Scottish Higlilands. 009(7 sciy.; Tebidi- au, 417, 422 ; Candjriiin, 528. Highlands of the Hudson, 403, 405, 655 et acq., 658, 600 ; are pre-Canibrian, 666 et acq. High liock, apatite-niino of, £26. Hildebrand and Cross, zeolites of Colo- rado, 1.35. Hinrichs, G., elmnontal ni.atter, 49. Hippocrates, nature, 8; medicine, 9; Galen on, 8, note. Hitchcock, Edward, on serpentine, 428 ; geology of Verninnt, 031. Hitchcock, C. H., crystalline rocks, 94 ; Hiironiiin, 410 ; Taconian, 031. Hobokeii, N. ,T., serpentines of, 439, 441. Hidxheiid, Wales, geology of, 410. Hcdlando. geology of Corsica, 473. Homologous series, 2HG, 289, 394. Homologies in niineral species, 289, 290, 291,304, 387 ('«,sr7., 390. Hoosac Mountain, Mass., decay i! rocks in, 256; tunnel in, ihiil. Hopkins, W., tho earth's interior, 115. Hornblende, Its decay, 32. See Amphi- bolo. Hot springs, zeolites, etc., formed in, 150 et aeq. Hot water, action of, on glas«!, 147 ; on bricks, 1,52 ; on feldspars, pyroxene, etc., 148 ; on locks, 022. Houghton, !>., Taconi'in of Lako Supe- rior, 579. 075. Hudson slates of Mather, .523, ,'■.24. Hudson-Hiver groupof \'aiuixem, ,524 et »('7.,531, 007 t-< st-r/.; its two divisions, 524,002; dames Hall on, 5s7, 001, 008 ; Logan on,00K,034, (i40 ; farther defined, 038, 04(1, 070 ; altered, 400, 4(19. Hughes, 1). T., auriferous gravels, 273. Hughes, T. McK., ge(.logy of Wales, 417. Humboldt, A., the unity <.•" nature, 22; physiography, iliid.; interstellar medi- um, 03 ; origin of nebula' 0.j ; serpen- tine, 427. Hummel, congloinerates in Sweden, 479. Hunt, K. 1?., terrestrial atmosphere, 44. Hunt, T. Sterry, Siemens on his studies of Newton, 51 ; views on limonites noticed, 205 ; views on serpentine defended, 502, 503, note; address at Priestley's grave, .'■)5 et !ry of, 611 ; sup- posed organisms of, 015 ; probably pre- Cambrlan, 021, 025; congb>morati'H in, 254, 415, OU; coutoundfd by some with Huroniiui, 612. Keweenaw and Kcweoiiawlan, 415, 614. Kimb.iU, J. P., Laureutian, 405 ; lluro- niau, 681. Kinetics, 12. King, Clarence, crystalline rocks, 02 ; volcanic agency, ibid. King and Itowuey, metasomatism, 90, 497 et acq. King's Mountain, S. C, rocks of, 559, 561, 604 ; Lieber on, 500 et seq. Kingston series, 408. Kishacoquillas Valley, geology of, 530 et seq., 540. Klttatinny Mountain, Tenn., rocks of, 531 ct seq., 538, 510 ; uncouformity to argil- lites, 632, 540. Kittell, inclined crystalline strata, 111. Kjorulf, geology of Norway, 508, 685. Klopstock, nature, 23, itnte. Koenig, astrophylliti', 355. Kopp, A., crystalline rocks, 93. Krablite, a supposed feldspar, 294, 338, 340. Kuhlmann, F., water-glass, 149. Kunz, G. F., orthoclase veius in diabase, 121. Kyauite, 101, 300, 503. LADRADOiti.VN, 413. See Norian. Labradorite, 338, 371 ; rocks, 404, 413, 580. Lako St. John, Ordovician of, 599. Lake Sui)erior, geology of, 205, 578 et seq., 010 (t seq, Lamination in rocks, from movement: exoplutonic view, 81, 201 et seg,, 210 ; endoplutonic view, 200, 205. Lambertville, K. J., banded diabase of, 211, note, Lapie-lazuli,342. Lapworlh, C, Ordovician, 528; Scottish Highlands, 070. Laumontito, veins of, l."5. Laurent, Aug., water in fused borates, 220; constitution of silicates, 297; di- visibility of molecules, (//((/. Laurentian series, 404, 412, COO : its divis- ions, 413 ; limestones, 109, 220, 238, 4.35, 658 et seq., CGO;; serpentines, 109, 332, 435 ; in the Adiroudacks, 403 ; south- eastern New York, 403, 405, 650 et seq., 658, ClO, 665 et seq.; South Mountain, I'enn., 257, 549; Buck Uidge, renn., 437, 550 ; North Carolina, 500 ; Colorado, 223 ; Canada, 223 et seq., 403 et seq., 412 ; Alps, 402, 472, 479; Ilavaria, 482. ice Gneiss, older, of Alps. Laun'Mtidi'S, 404 ; not the nucleus of the continent, 500. I.au7.on rocks, 506, 6.34. l.avas, water in, 06 el seq., 222, 245, note. Lavoisier, eloniental nuittor, 49, 57. Lazulile, BG'S. Lead, iiialato of, its cry3talliz,ation, 170. Leboiir, on Carrara marbles, 477, noti. LeCoute, ,Ios., gold gravels of C.difornl.l, 272. Lehman, on prindtive rocks, 68 et seq. Leliinann, .Job., lamiuiitcil rocks, 202. Leibnitz, primeval eartli, ri8, 70. Le lloyerand Dumas, mulecidar volumes, 302. Lesley, J. P., limonltes in Pennsylvania, 203 et neq.; iron-ores in do., 5.50 ; iron Manufacturer's (inide, ilnil.; Potsdam sandstone, 600, iin/e ; relations of tho Second (iraywaeke, 632, nnle ; on H. I). Kogers, 658. Leucite, artificial production, 219 ; change to analcite, 371. Levant Siimlstone, 5,^4. Levis limest -no, 608, 047. Sec Sparry Limc-roek. Lewisian gneiss, 417. Lewis, 11. C, laminated rocks, 203; ter- tiary 'imonites, i;63, note. Lherzoiite rocks, 500 ; in North Carolina, 507 ; Norway, Pyrenees, and elsewhere, 608 ; their stratified character, iOitl, See Chrysolite. Lieber, Oscar, serpentine and steatite, 429 ; Itacolumite rocks, ,"65 et .feq., 680. Liguria, serpentines of, 452, 485 et seq,, 405 et seq. Lime-carbonate, origin of, 178, 239, 2.53 ; solubility of, 108 ; replacement by iron- carbonate, 266. Limestone, decay of, 249, 271 ; with ser- pentines, 4.'!5, 462, 501, 573; si^)posed metasomatic changes of, 102 ; changed to granite and gneiss, 103, 408; of Clielmsfoni, Mass., 230; St. John, N. B., ,572; Kockland, Me., iliid.; Port Henry, N. Y., 237 ; of Stockbridge {see Stockbridge limestone) ; Auroral (see I I # 111 . •1: 700 INDEX. Auroral Hmostone) ; eruptive (weEnip- tlve rocks). liiiuoniios, tertiary, 203, notf ; Taconlan, I'd, Ii.'i.'), 5:Wi, 5,">5; not from inaiinetito nor lu'iiiatitc, 601) ; from Bitlorltu, atjl, 4H1, 07.1, nso ; pyrites, lifi!», L'Cl, no", r,m ; Bcrpnntiiie, 2f3M ; contiaotion in forma- tion of, lliiL' ; hollow I, assoH uf, ihid. ; of Appalaclii.iu Valley, 1:01, 1!C4 ; Stati'u Islanil.aW ; .South Carolina, 509 ; Nova Scot.ia, .573 ; ^Aliohigaii, 5S0 ; oonnlomer- ntos of, fi"). T'iugula of Auroral limestone, 582. hinmeus, orystallino ruclis, O'J. IJiMKMiiaiin, zircon, l!!4, noti'. Ijinvlllc Mountains, X. C, 053, 550. Littrc, Dictiiiiiary, 7, 8. l.lanlicris, Wales, slati^s «♦, 420. Llano ffroup of Texas, 021, 024. Llyn Padarn, WaloP, coiigloniurates of, 420. Locke, John, natural p.illosophy, 3. Lockyer, .J. X., cosmic evolution, 48; solar physit"), r>X Logan, \V. K., I.aiirontlan, 2.1R ; Potsdam group, O0!t, Oil, Oi:i, 670 ; Hed Sand-rock of Vermont, 008 ; (,»neboc group, 507 603, 031, 070 ; altered Quebec (jroup, 40C, 410, 610 ; altered IIiidson-Kiver (.'rouft, 400, 409 ; geology of Hu. MuinphreiuiigdB Litkc,litnc8toiin!i of, 031, Menage. I>ictiunuaire Ktyinnloglque, 7. Meneglilnl, geology of the Alps, 473. Mental pliysiokigy, 6, Menevliin rocks, 407, 573, 677, 018, G'.'S. JSIeuoiuinoo UlBtrict, Mich., geology of, blH, 6«0, 075. Mesozoii! in I'onn., W5, 549. MetageneHls, 14. Molallates, oriler of, 3;;o, 378 ; volumes of, ;i79 et siq. Metallic all.. 72, en.iimtiDii in, 209. Metalliferous I line-rock of Katon, 520, 521 i veins, igneous origlu of, 74, 00. See Veinstones, Metallometallates, sub-ortler of, o78. MotuUoids, tribe of, 378. Metamorphosis, hypothesis of, 'r,, 00, P2, 104, 107, 403, 0.14 et mi/., Cis, C8K ; T?onU and Lyell on, 82, 83: Delesso 0)i, 4;i2; Mather on, 629, O.W ff mq.; falla- wayand Honnoy on, 008; Dana, . I. I)., on grades in, 005, 07H ; rise and fall of, 108,008; inechaidca!, 20.'), 072; in New England and New York, 055 et seq.i Scottish Highlands, 009 et seq,; de- fined as pseudomorphism on a broad scale, 100, 200. Metasoniatosis, hypothesis of, 84, 105. 497 et seq.; its difflrulties, 108; two schools of, 98 et seq., 101 ; of pUUonic rocks, 99 et seq.; of limestones, 102 et seq.'; Haidlnger and Rischof on, 200 ; Pum- . pellyon,498; Dana, J. I).,on, 101,7m?e,' of granite to limestone, 101, 7io^', 498 ; limestone to granite, 102 et seq., 498 ; serpentine, to limestone, 498 ; grfvnite to serpen'Jne, 431 ; limestone to ser- pentine, 102 ; limestone to petrosilex, 103, 498; limestone to hematite, 103, 499 ; corundum, 100 ; chrysolite, 101. Methylosis, 498. Meunier, Stanislas, sources of carbonic dioxyd,40; planetary atmospheres, ibid. Micas, 353, 3.'56, 308 ; artificial production of, 149 ; origin of, 160 ; table of musco- vitic, ibid. Mica-schist (Montalbani series, 411, 413, 423, 403, 460, 472, 479 et seq., 482; of Michigan, 580 et seq. ; New York city, 606 et seq. ; Alps, 401, CS.T ; Saxony, 202, 47H ; liavaria, 4H1, .Ij'ic tiuuiss, younger, and .Montalbau. Michel Levy a;id Fou(|Ui*, formation of ■lllcatOH, 209, 219, 375. Michigan, geology of, 579 ct seq., 075. Mimetic resemblances in minerals, 108, 31H. Blinas Geraes, Brazil, geology of, 604, 080. Minnesota, rock-decay in, 270 ; geology of, 579 et seq. Mincralogical evolution, 088. .Mineralogy dellned, 2.'>; basis of classlfi' cation in, lt;5, 107; scope of, 287,398; natural-history ni,,i;-od In, 2>^0, 2H3 j cliemical method in, 2m2 ; natural sys' torn of, 279, 2M7, 297 ; history of its develnpmenl, 2Hl et neq.; Objects and Method of .'Mineraloj.'y, 2m5 ; character' iatlo In, 313 ; classes and orders in, table of, .'tK2 ; tribes in, 314 et .seq., 321 ; Manual of, pniposed, ;t;iK ; families and geni^ru in, Ohh ; binomial l.atiu nomen- clature In, ibid. Minerals seeroted from basic rocks, 134, i;!8 et seq., 220, .'i07. :\lissis8'ppi valley, Cambrian of. Oil, 023, 021. Mittelgebirge, Saxony, 202, 478. JIolis, System of Minrralogy, 280, 313} visit to Kdinburgh, 2h(). Molecular weights, 2>>0, 3t<3 ; relations o£ to density, 384 et .fcij., 392; volumes, 291, 302, 304, 319, 376 et seq., 379, 384, 391 et seq. Molecules, indefinite divisibility of, 297. Monocraterion, 542. Montalban series, 18.3, 400, 411, 413, 481, 057 ; thickness of, 412 ; serpentine and chrysolite rocks in, 507, 500 ; veinstones of, 223 ; conglomerates in, 255 ; in New llampsliire, 057; New York, 007; Blue Kidge, 258 ('< .leq. ; Michigan, nHOet seq. ; Scottish Highlands, 423, 009 et seq.; Ireland, 423; Saxony, 202, 255,479; Ba- varia, 482: St. C}othard,47l ; Alps, 472. See Gneisses, recent, cuid Mica-schists. Montarville, cbrysolitic dolerite of, 210, 605 ; its banded character, l.O ; analy- ses of, 212. Monteferrato, Italy, 8erpen*innu, etc., of, idO et seq., iW>. Moorr Ancient Mineralogy, 307, note, Morlot, Von, dolomites, 172. I 702 INDEX. I Moro, liftzfftro, rruptlvo rocks, 00. Mutt, •!. \j., terrcHtrlal onrl)(>ii, .'tn. Mount Ida, (Irucco, chryHoUtu-rocka of, BOO. Mount Uoyrtl, rnnaila, orllioolndo vcinii In, 137 ; ImiwU' I ilolcrllo of, 'JIO. Amount Sorrel, V.wg., rockn "f, 4'JI. MurcliiHon, K., Slliiriim of, (I2(, C2H; his vlow of Su()ttlriil.; Southern District, 523; gneisses of, 403, 650 H Seq. Now Hampshire, geology of, 410, 057. Newfoundland, Cambrian of, 026 ; Taco- nian in, 571. New Hoeheiio, N.Y., serpentine nf . 4.^.^ New Zealand, dunlto, 608. See Lberzo' lite. Niagara limestone, 520. Nlcoll, geology of Scottish Ilighlandi, 600,071. Nlidiic oxyd in silicates, .100, ;m. NIpigon grouii of Lake Superior, 678. Nittany Valley, Pcnn., 531. Nordunskiilld on geological climate, 45 ; on roek-tlecay, 240. Norlan series, 17s, 404, 413, 580. North Carolina, Kmmons on geology of, 558 et Hiq., 5(i2; Kerr on do., 5,')8, WiO, 501 ; H. Wurtz on, 509 ; Laureniian in, 500 ; Taconian In, .Ifil et seq,, 5o:t. North America, pre-Cambrlnn of, 402 el seq. ; paleozoic history of, 015 et seq. Northbrldge, Mass., metalliferous veins of, 123. Notre Dame Mountains, 610. Notation, cbondcal fornndas in, 302, 312, 319. Novara, Italy, serpentines of, 490. Nova Scotia, Taconiai\ In, 573. Nuttall, gneisslc rocks, 403; metamor- phism, 658. On.siniAN, 221, 362, 375. Ocean, primitive, its nature, 36, 177, 180, 253 ; its temperature, 77, 79, 00. Ocrstedlte, .367. Oken, Physiophllosophy of, 23; his influ- ence, 24, note. Oldhamia, 421. Olivine. See Chrysolite. Omaiius d'Halloy, eruptive rocks, 90 ; crystallophyllian rocks, 80. Oneida sandstone and conglomerate, 620, 523, 525, 531, 540. Onondaga salt group, 527 ; serpentines of, 443 et seq. ; gypsums of, 444. Ontario division of New York rocks, 522. Opaloids, tribe of, 370. Ophicalcite, 99, 439, 402, 483, 485, 403, 501, 602. See Serpentines. INDKX. ioa [y of, fl:> it mn., (ill!, CC7. Ornanoni'iiy ili'llni'cl, 17. OrlKkany HiinilitioiK', 527. Ortliocliisc, It^-. ili'oay, ;i2, 247, 240; veins in liiuiiust^ 121 (7 «(Y/., in iloleritc, t:i7 ; 111 inotiiUiferouH lotius, 123 ; witli zoo- liti'g, 120 it Mn/, Ortliofelsilc, 5t(l. Si'f rrtrosilox. Ottawa l)aHin, i;22 ; unoonfonuity iu, 59'J ; UlieisH, 412, 11.1. Oxycollolils, Irilie of, 370. Oxyiliiti's, onler of, ;)'jo, ;i7fi. Oxyduinantoiils, tribe of, 37C ; volumes of, ibitl. OxyiiH, native sources of, 1,50, 181,239; a.tsoclati'tl Willi silieatuM, IHI, 239. Oxygen, (llslriliiilion of in space, 41 ; re- lation to cmtionio ilioxyil, 35; libera- tion of in organic processes, 33, 3U. Oxyphylloids, trilte of. .'!7C. OxyspatUoiils, tril)o of, 37C. P = UNIT WKinnT, 303 et scq., 301 ; bow calculated, .'?2fi. ralagfinito, 120, 1.52, 150; artificial pro- duction of, l;!0; eliango to a zeolite, 120, 374 ; silicated protoplasm, 1«8, 3G2, 374. Paleotrochis of Kniiiions, 501. Paleozoic history of Xorth America, C15 et scq. Pallas, crystalline rocks, CO ; rock-decay, 247. Pantanelll, phthanitcs, 490, note. Paragonite, 101, 102, .357. Pargasite, 310, 345. Paris basin, nepiolites of, 448. Parmentier, niolybilates, 388, note. Parophite, 1(M. Passaiiiaquoildy Bay, petrosilex of, 647. Passau, Uavaria, kaolin of, 371. Patrin, serpentine, 426. Pebidian series, 417 et seq., 422 et seq., 449; upper, 423, 670. Pe( llHntloii, of riralmn), .115, ,122. I'rTtiillti', I'tymcdogy of, ;\St, xnle, I'eetolitiu silicatun, 13N, 140, 147, 178; list of, ir,. I'ec|ollioidi<, trilie of, ;!15 ; table of, 323. I'elljitl, Italian serpentines, 1,51, 474, 484, 4.'(I. 4HH. Penilirokcsliire, geology of, 41C, renlleld, lieryl,;H7. I'ennsj ivaiila. geology of, 5,10 ft *eq. Peradanmntoids, Ir'ljo of, 318; tublo of, .'Km. Pcrlite, 221, .102, .175. Porpiiylloids, trilic of, .118 ; table of, ,K1R. Perry, •!. h.. limestone veins, 230 ; Ued Sand-roek, 0;>0 ; Taconian, iliul. I'eisilieates, sub-order of, 306 et ten,; table of, 401. Peispallioids, trilie of, 318, .106. lVr/.e(ditoids, :117, .10.5. IVtalite, 2:10, 231, 340, 0H7. I'etralogy, IMiikerton's, 427, note; com- parative of Iiiiroclicr, L'liH, 214 et siq. Petrosilex series ^.Vrvoldanl, 408, 410, 41H ; ill Wiscon.Hin, 410, 540; Missouri, 10.1, 200, 4!iH; I'ennsylvania, 510 ; Massachu- setts, 40.S (7 mil. , New Hruiiswick, 400, 540, rA' ; Lake Superior, 410; Wales, 4Ih; 111 Scotland. 424 i Alps, 515, iio^-; Sweden, 419, 470 ; conglomerates of, 41!i, 420. Phillips, John, two igneous magmas, 87, 207. Phillips, J. Arthur, rocks of Cornwall, 4.50. Philosophical Society of Cambridge, 61, 07. Phiisphates, complex, 388. Pliospiiotungstates, ;ihO, 387, 302. Piiosplioiiiolyliilates, 3S0. I'hiderite, its liistory, 367. Phoiiolite, (u-igin of, 218. Phthanite, 488, 400, iiore. Phyllograptus shales, 625, C2C. Phylluid type, 310, 374. Physic, defined, 1-11 ; general, 17. Physics. Si.'V I'liysio. Physician, 7, lo; defined, 7, vote. Physiology, general, 2-7, 11, 12, 18, 21 ; of Ionian school, 4 ; mental, 5, vote; ani- mal, 11 t< stry./ vegetable, 25 ; mineral, 26, 28, 29. Physiography, scope of, 17, 22, 27. Picotite, 148, 508. Pi -olite, 324. I <'l 704 INDEX. Pietre verill (greenstone) growp, 411, 45G, 459 et acq., 4G3 et sirj.. 472, 477, 514 ; thickness of, 405 ; UastiilJi on, 464, 483 ; Neri on 4C.'3,4r>4; is Iluroni.an, D82. See Huroni.'in diul Ureuustone. rilolite, 360. Pinit3 and related species, 163 et seq,, 1C4, note, 301, 303. riniti)ids, tribe of, 317 ; table of, 361. Pinkerton's PetralDgy, 427, note. Pitchstond, 221, 302, 375. Planets, alnmsphores of, 46, 47. Playfair, John, exposition of Ilutton, 68, 74, 75 j biography of, sO, note. Plenum of Descartes, 57, 62. Plonibiiires, thermal spring of, 150 et seq, Plutonic substratum, 115, 117, 118, 132 et seq., 175, 179 ; secular changes in, 131, 178, 207, 214, 218, 244 ; Prestwich on, 118, note ; rocks, relations of to metamor- phic, 432. Plutonists, 74, 75, 217. See Eudoplutonic and Kxoplutouic. Pollucito, 331. Polycarbonates, 289, 290, 298, 304, 386. Polymerism in mineral species, 285, 286, 289, 377, 394. See Condensation, en. Polysilicates, 286, 289, 298, 3U}, 393. Porodini, Ureithaupt's order of, 383. Porodinous (Porodic) species, 383, 687. ■See Colloids. Port Henry, N. Y., limestones of, 237. Portsmouth, B. I., amiunthoid silicate of, 195, 359. Potsdam, sandstone, 521, 526, 534, 577,611, C24 ; deposition of, 618 ; Lesley ou, 660, note; Lower, 618,624,637. Poussin, De la Vallee, and RiSnard, geol- ogy of the Ardennes, 422. Po.vell, J. W., Grand Cafiou group, 624. Prato. See Monteferrato. Prehnite, 143 ; is an adamantoid, 347. Pre-Cambrlan rocks, history of, 402 et seq. ; in Nortli America, 402 ; Europe, 416 et seq. ; literature of, 113, note. Pressure, its elfect on solution, 222. Prestwich, T., on a solid earth, 1!S. Priestley, J., address at grave of, 49, C5, 396. Pr-mal series of Rogers, 533, 542 et seq. ; thickness of, 542, .543 ; slates of, 544,547, 548 ; iron-ores of, 650 ; in Virginia, 602. Primary plutonic rock. See Plutoflo substratum. Prime, F., geology of Pennsylvania, 540 etseq. Primitive Clay-slate, 481, 482, 684. Primitive Lime-rock of Katon, 519, 529, 554. See Taconian, limestones of. Primitive Quartz-rock of Katon, 519, 529, 554. See Taconian, quartzites of. Primitive rocks, 69, 70, 190, 402, 530; Lehman on, 69 ; Werner on, 70. Primitive schists of Lory, 406. Principia, Newton's, 58 et seq., 6J, note, 62, note. Protadamantoids, tribe of, 315 ; table of, 3211. Protogine of Alps, 46G. Protoperauamantoids, tribe of, 317; table of, 345. Protoperphylloids, tribe of, 317; table of, 354. Protoperspathoids, tribe of, 310 ; table of, 337. Protophylloids, tribe of, 31G; table of, S31. Protospatholds, tribe of, 315; table of, 327. Protosilicates, sub-order of, 305, r07, 314; table of, 399. Protojiersilicates, sub-order of, 305, 307, 316 ; table of, 400. Protoplasm, dettned, 18 ; mineral, 188, 374. Prntoxyd-silicates, table of, 145. ProuKtoids, tribe of, 379. Pseudomorphism, 83, li'O, 101, note, 303. See Metasomatism and Metamorpbism, Psychology, limits of, 18. Pulaski shales, 520, 524 et seq, Pumpelly, metasoiiiatosis, 103, 498 ; zeo- lites, etc., of Lake Superior, 143 ; rock- decay, 249. 209, 274 ; erosion, 275. Pyrallolite, 333. Pyrenees, serpentines of, 483, 502, note, Pyricaustates, order of, 380. Pyritoids, tribe of, 378. Pyrites, associated with limonite, 259, 261, 555. Pyrognondc minerals, 96. Pyropliyllite, 309, 425, ?jo/e; of Taconian, 561, 503. Pyroxene, 290, .328, ,330, 393; formation of, 148, 220 ; aluminous, 213, 310. Pyroxenite veinstone, 225, 227. Pythagoras, school of, 3. QuANTiVALENT ratios of silicates, 310. INDEX- 705 Pennsylvania, 540 i of, 315 ; table of, itU limonite, 269, 0, 393; formation us, 213, 310. 225, 227. of silicates, 310. Quartz, varieties of, 138 ; volume of, 377 ; artificial formation of, 148, 149, 157, 216, 218, 219; fusion of, 209, note; fluorhy- dric acid on, 687 ; supposed eruptive, 95, 96, 429 ; -rock. See Primitive Quartz- rock. Quartzlto group of Uominger, 580. Quenstedt, primeval atmosphere, 114. Quebec group, 518, 585 et seq., 5!)6 et seq,, 603, 609, 617 ; Logan on, G34 ; inverted by him, 596, 634 ; altered, of Logan, 406, 410, 610. See First Graywacke, Hudson- River group, and Taconic slates. Quebec city, geology of, Emmons on, 584, 585 ; Logan on, 594 ; section at, 596, 634. Rammelsbero, chemical mineralogy, 282 ; studies of tourmalines, lii2, 350 et seq. ; micas, 359. Kay, John, natural systems in classifica- tion, 284. Beading, Penn., iron-ores of, 551. Ked Sand-rook of Vermont, 593, 608, 630, 638. Refrigeration of the earth, 114, 117 ; Ssemann on, 47. Rensselaer Co., N. Y., laconic slates of, 586, 627. Renevier, amorphous chabazite, 154 ; geology of the Simplon, 466. R^nard, geology of the Ardennes, 422; coticulite, 425, note. R^uard and ^Murray, zeolites in deep-sea ooze, 154. Rensselaerite, 333. Resinoids, tribe of, 380. Betinalite, 332, 435, note, Reusch, crystalline rocks, 93 ; erosion in Korway ami Corsica, 276 ; and Brogger, apatite of Norway, 236 ; Iherzolite, 508. Reynolds, O., physiology, 5, note. Rhode Island, Taconlan in, 257; new silicate from anthracite of, 195, 360. Riohthofen, aerial erosion, 275. Rideau, apatite mining district, 224. Rifts or divisional planes in rocks, 274, note. Rigaud Mountain, Quebe'5, boulders of decay on, 271. Rivot, geology of Lake Superior, 612. Robb, Ch., garnet with prehnlte, 121. Rock-basins, origin of, 252. Rock-salt, supposed eruptive, 90 ; of the Alps, 4G8. Rocks, crystalline (see Crystalline rocks); indigenous, 72 ; exotic, 73 ; endogenous ibid. ; igneous, eliqualion in, 189, 208 et seq., 245 ; secular changes in, 187, 253 ; solid, intruded, 73, 204, 512 ef seq. ; dutri- tal, changes In, 108, o72, tm. Rock-decay, sub-aerial, 246 et aeq., 277, 308 ; Its antiquity, 236, 270 ; its chem- istry, 31 et seq., 250 ; relations to erosion, 251, 271, 274 et seq. Rogers, Henry D., calcareous veins, 229; crystalline rocks, 405 et seq. ; eruptive quartz, iron-ore, etc., 95, 429; llypo- zoic and Azoic rocks, 405, 661 ; the White Mountains, 658, 663 ; geology of Pennsylvania, 533, 543; his strati- graphical divisions, 553; Taconic, 628. Rogers, William B., rock-decay, 258 ; formation of iron-carbonate, 266 ; Red Sand-rock of Veru)ont, 630 ; Taconlan, 628, 631, 664 ; crystalline rocks In Vir- ginia, 661 ; geology of the Blue Ridge, 662. Rogers, Messrs., cited by Dana, 664. Rolljstiin Hill, Mass., gneiss of, 274, note. Romlngur, geology of Michigan, 579 et seq. Roofing-slates, Lower Taconic, 525, 538, 555, 571, 578, 580, 589, 592, 675 (see Argil- lite, Transition) ; Upper, 555, 592. Roscoe, H. £., a new inorganic chemis- try, 389, Rosenbusch, Iherzolite, 508. Rosinini, philosophy of nature, 16, note. Rougemont. Quebec, dolerite of, 210, 506. Royal Institution, London, 53. Russia, diamond region of, 565, 680. Rutile, 377 ; of Taconlan, 503, 568. S.«:mann, a cooling earth. 47. SatTord, geology of Tennessee, 559. Saint David's, Wales, rocks of, 416. Saint Gothard, Mount, geology of, 73, 204, 470; tunnel of, 470; gneisses of, 470, 471, 613 ; serpentine of, 511 et seq. Saint Helen's Island, Montreal, geology of, 604, 631. Saint John, N. B., Taconlan of, 672 ; Lower Cambrian of (St. John group), 623 (see Menoviaii). Saint Ours, alkaline water of, 218. Saint Peter's sandstone, 611, 623. Salinoid type, 380. San Domingo, geology of, 683. Santorin, lavas of, 213. Sapoulte in zeolitlc rocks, 139. 706 INDEX. !!v ' 1 ' 1 •I Sardinia, geology of, 475. Sauer, cunglouiarates in crystalline rocks, 183, 255, 479. Saiissure, granites, 71, 72; serpentines, 427. Siiussurite, 299, 301, 348, 451. Savi, igneous limestones, 228, 477; sor- penti'.ies or ophiolites, 428, 451. Saxony, granulites of, 202, 255, 478; dichroite-gneiss of, 478; gabbro of, ibid. ; serpentines of, 479. Scnpolites, 300 at seq., 340 et seq. ; inter- mediate species of, 295, 342. Scandinavia, gneisses of, 404. Schelliug, 23, note; life in matter, 21, note. Scblel, progressive or homologous series in chemistry, 289. Scheerei, ^yrognoiuic minerals, 96; water in granites, 96; hydrous iolites, 143; crystalline admixtures, 294 ; serpen- tines, 504. Schorlite, a tourmaline, 351. Schulten, l>e, artificial production of zeolites, etc., 157 et seq. Scolithus linearis, 534, 542, 554, note, 567, 615, 676 ; Canadensis, 554, note ; of the Medina, 634 ; of Hastings rocks, 576 ; sandstone, 414, 544 et seq., 548, 554, 660. Scotland, ancient gneiss of, 404 ; decay of rocks in, 271 ; Highlands of, 423, 424 ; liistory of their crystalline rocks, 669 et seq. Scott, Walter, the word mediciner, 10. Scrope, Poulett, New Theory of the Earth, 81, 201 ; hydrothernial hypothe- sis, 81, 96; granites, 81 ; volcanoes, 96, 201 ; Igneous rocks, 210, 222 ; lamina- tion by movement, 81, 201. Sea-water, sovirce of magnesia in, 180, 182 ; removal of from, 177, 182 ; fossil, 253. Second Oraywacke, 520, 622, 520, 630, 684, {■*5, 620, 677 ; its relation to Transition Argillite, 632; overlies First Gray- wacke, 686, 698, 600, 627 ; overlies Cam- brian and Ordovician, 599; confounded with First Graywacke, 698, 649, 654. Secondary rocks of Werner, 70, 190. Secretion of minerals from basic rocks, 134, 135 et seq., 220, 307. Secular variation In composition of rocks, 113, 187, 214, 216, 253, 678. Sedgwick, rocks of Wales, 416 j Cambrian divisions of, 624, 628. Sella, Qulntlno, geology of Jjombardy. 462, 496. Selwyn, A. R. C, pre-Cambrian rocks, 424 ; Ordovician in province of Quebec, 607. Senarmont, H. de, serpentines, 428, note ; artlflcial production of corundum and diat^^ore, 501. Sepiolite of lertiary, 119. 185, 196, 448. Sericite-schists, 161, 467, 474, 683. Serpentines, 426 «t seq. ; intrusive, 95, 427, 428, 429,431, 452, 456, 469, 483, 486, 495,497, 509 ; metasoniatic, 431, 450, 452, 498 (see Metasomatism); metamorphic, 486, 499 ; hydroplutonic, 97, 487, 492, 500; in- digenous, 430, 431, 433, 448, 486, 501 et seq., 503, 51? ; decay of, 268, 441 ; dehy- dration of, 606 ; chrysolite from, 500, 513 (see Chrysolite) ; stratigraphical relations of, 427, 428, 483 et seq., 492, 509, 510, 511 et seq.; grauulite with, 439 ; granite vein in, 438 ; conglomerates and breccias of, 453, 493, 495 ; veins of, 453, 609; Laurentian, 332, 435; Huro- nian, 436 ; Montalban, 437 ; Taconian, 442, 673 ; Silurian, 185, iiSet seq.; sup- posed tertiary, 455 et seq., 486, 515 ; of North A merica, 434 et seq. ; New Ro- chelle, N. Y., 435 ; New York City, 439 ; Staten Island, 440,496; Hcboken,N. J., 441 ; Syracuse, N. Y., 443 ; Cliester Co., Penu., 437 et seq.; Cornwall, Penn., 442 ; North Carolina, 438, 507 ; Lake Supe- rior, 579; Europe, 449 et seq. ; Scotland, 427, 510 ; Cornwall, i^ug., 428, 449, 510 ; Saxony, 479; Mount St. Gothard, 511 > . seq., 513 ; Alps, 427, 4C3, 465, 469, 472 ; Apennines, 455, 456, 487 ; the two re- gions compared, 484 ; Italy, 450 et seq., 482 et seq.; Liguria, 452, 485; Sestrl Levante, 495 et seq. ; Tuscany, 452 ; INIonteferrato, 490 et seq.; Lombardy, Biellese, 463, 496; Corsica, 474; of Elba, 475 ; Emmons on, 428 ; Mather, 441 ; E. Hitchcock, 428 ; Bonney, 449, 450, 462, 495 ; Goikie, 610 ; King and Rowney, 498; Stapff, 501, 511 et seq.; Delesse, 431, 432; Daubr»5e, 600; Qas- taldi,4S3, 486; Gras,432,-io0. Taconic slates, group of, 527, 585, 627; called Upper Taconic, 527, 630 ; subdi- visions of, 644 ; limestones of, 646 et seq. (see Sparry Lime-rock) ; fossils of, 586, 647; equivalent to Calciferous Sand- rock, 587; overlies Transition Argillite, ibid, i overlaid by Second Graywacke, 686, 598, 600, 627. Is First Graywacke, and Quebec group, which see; also, Hudson-River group. Taconic hills, geology of, 553 et seq., 627, 643, 645. Taconic system, 617 et seq. ; Lower (see Taconinn) ; Upper (see Taconic slates). Talc-scaists of Lory, '66. Taramelli, geology of Italy, 456, 478. Tennessee, Taconi,an of, 659. Terranovan series, 480. Tertiary strata, 190; limonites of, 263, lu^te; in Italy, 474, 485, 490 et seq., 496, 616. Tcschenite, 137. Tesori Sotteranei del' Italia, Jervis, 477. Texas, geology of, 621, 624. Theologians on Hutton and Werner, 76. Thermal springs, mineralogy of, 160 et seq. Thermoohaotic hypothesis, 82, 105, 10i». Thermodynamics, 26. K-l^-ir-il INDEX. 709 ', 553 et seq., 627, sis, 82, 105, 10^. Thiogalenoids, tribe of, 379. Thoinsou, Win., on Imerstellar space, 63. Thorasonite, 1-12, 334. Thomson, ISIinn., argilUtes of, 578, 580. Thunder JJay, Lake Superior, 678, 611. Ticino, Italy, geology of, 470 et seq. Tilden and Sheiistone, solution at high temperatures, 221. Tintio Hills, Utah, geology of, 676. Tonto group, 624. Toreil, crystalline rocks, 418, 419. TOrnebohm, crystalline rocks, 93 ; Iherzo- lite, 508. Torrance, J. F., apatite-veins, 224, 232, notes. Tourmalines, 138, 101,425, wo^e; composi- tion of, 350 et seq. ; table of, 353. Tozzetti, gabbros, 451. Trachyte, origin of, 133, 186, 187, 217; Bunsen's normal, 129, note. Transmutation of rocks, doctrine of, 98, 100, 102. .See Metasomatism. Transition Graywaoke series of Eaton (see First Graywacke) ; rocks of Wer- ner, 70, 100, 190, 402. Trenton limestone, 521, 537, 540 et seq. ; its history and distribution, 600, 604, 606, 620. Tribes in mineralogy, 314, 321. Triads, Eaton's, 518, 527. Tri,i8, supposed altered of Alps, 467, 470, 48 1,683, 684. See Glanzschiefer. Triple division of utiata, Eaton' '.IS, 527. Tridymite, 151, 157, 3(C, 687. Trinity College, Cambridge, Eng., 61. Troy, N. \., Cambrian of, G39. Tschermak on intermediate feldspars, 295, 304 ; on scapolites, 340 et seq. Tungstates, complex, 386 et seq. ; unit- weights, 392. Turgite, 635. Tuscany, serpentines of, 462 et seq., 486, 490 et seq., 492. Tyndall, physics, 12 ; life in matter, 20 ; terrestrial atmosphere, 44. Unit-volcme, 303, 391, 394 ; of various species, 291, 392. See Molecular vol- umes. Unit-weight of species, 303, 391 ; how calculated, 325. Universal anim.ation, 16, note, 18. Uplifts or faults in strata, 693, 602,639 eid, 336. r space, 49. sisses, younger. conditions of lecretions from 138, 307 ; bnses i-sea ooze, 154 ; ition of, 150 it formation of, eiiesia of, 143; )n, 135 ; Kmer- f, with related 11, 313. 4; table of, 336. 231; Taconian, I, 306,' 335, 344, 214 ; varying meionite and ;ellar space, 65.