3 a Hearts eiiritcratar hy ws Po dt arary Trep et eH Aa abbas ht 3 oo Tet) i) i “| < i : a i #1) i dich hy, oe Kt tie i ‘ a | i i % peng , i eh i My AOR ie an aH ae te ! i uit i b} 4 te 1 r iH ae at cf Cent aE ia eetits f ; ; | nlbatates ; ii tf ; Wah ened ef ite ni th : i va igh ih t's ty ihe eta i ais an Hee tM het saatie idee , : i i ai i malig ——— HRasasua ars ele4 tt op vdedh agate if) Asti) He f Mi seh piadnd hte 4 ii iM x z bal ise an ie a 0 i “¢€ a0 ie i — = _ as ar. vole a ; = Sed ' j ! } ‘ Ne } t { L* i = a ‘ = J = POwIeN Ae OF GEOLOGY THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS THE CAMBRIDGE UNIVERSITY PRESS LONDON THE MARUZEN-KABUSHIKI-KAISHA TOKYO, OSAKA, KYOTO, FUKUOKA, SENDAI THE MISSION BOOK COMPANY SHANGHAI Ne isles, moUIKNAL OF GEOLOGY A Semi-Quarterly Magazine of Geology and Related Sciences EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of STUART WELLER, ALBERT JOHANNSEN, Invertebrate Paleontology Petrology ROLLIN T. CHAMBERLIN, Dynamic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain JOSEPH P. IDDINGS, Washington, D.C. CHARLES BARROIS, France JOHN C. BRANNER, Leland Stanford Junior Uni- ALBRECHT PENCK, Germany versity HANS REUSCH, Norway 7 RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. GERARD DrEGEER, Sweden WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Leland Stanford Junior CHARLES K. LEITH, University of Wisconsin Univerisity WALLACE W. ATWOOD, Harvard University CHARLES D. WALCOTT, Smithsonian WILLIAM H. EMMONS, University of Minnesota Institution ARTHUR L. DAY, Carnegie Institution VOLUME XXVIII JANUARY-DECEMBER, 1920 THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS Paplished: February, March, May, June, August, September November, Plecemiber, 1920 : Gdiuboseit and Printed by The University of Chicago Press Chicago, Illinois, U.S.A. Commurs of YoLtumMe X XV TIT NUMBER I _ DIASTROPHISM AND THE FORMATIVE PROCESSES. X. THE ORDER OF MAGNITUDE OF THE SHRINKAGE OF THE EARTH DEDUCED FROM Mars, VENUS, AND THE Moon. T. C. Chamberlin Tue Laws or Etastico-Viscous Frow. II. A. A. Michelson THE GREAT GLASS-SPONGE COLONIES OF THE DEVONIAN; THEIR ORIGIN, RISE, AND DISAPPEARANCE. John M. Clarke . A QUANTITATIVE MINERALOGICAL CLASSIFICATION OF IGNEOUS Rocks —ReEviseD. Part I. Albert Johannsen. THE PRE-MOENKOPI (PRE-PERMIAN ?) UNCONFORMITY OF THE Gono RADO PLATEAU. C. L. Dake PALEozor1c DIASTROPHICS OF THE NORTHERN Ivieaans Teen ey Charles Keyes . SomE ESTIMATES OF THE aerernnse OF THE Sapam eens oF Onto. T. M. Hills RECENT PUBLICATIONS NUMBER II THE ORIGIN OF GumBOTIL. George F. Kay and J. Newton Pearce DIASTROPHISM AND THE FORMATIVE PROCESSES. XI. SELECTIVE SEGRATION OF MATERIAL IN THE FORMATION OF THE EARTH AND Its NrercHBors. T. C. Chamberlin A QUANTITATIVE MINERALOGICAL CLASSIFICATION OF eeand ees —ReEviseD. Part II. Albert Johannsen REVIEWS ; RECENT Pyaar : NUMBER III COMPILATION AND CoMPOSITION oF BiTumMINOUS COALS. Reinhardt Thiessen - A QUANTITATIVE IA capRAroeR gE fuga OF ieeeee Boer —Revisep. Part III. Albert Johannsen GEOLOGICAL SETTING OF NEW Mexico. Charles Keyes Vv PAGE 126 158 178 182 185 210 233 vi CONTENTS TO VOLUME XXVIII MovEMENTS IN CRYSTALLIZING Macma. Frank F. Grout DEFORMATION OF CRYSTALLIZING Macma. N. L. Bowen REVIEWS RECENT ETON ONG NUMBER IV THe CHESTER SERIES IN ILLINOIS. Stuart Weller A CORRELATION OF THE PRE-CAMBRIAN FORMATIONS OF Nope ONTARIO AND QuEBEC. H. C. Cooke ; ae THE JUAN DE Fuca LOBE OF THE CORDILLERAN ICE San I. Harlen Bretz -. “SLIDES” IN THE Conmuxccn BOnmerond NEAR Mose eran Waer Vireinia. Earl R. Scheffel . A REPLACEMENT OF Woop By DOLOMITE. S. 'F. Adanis A Discussion oF ‘‘NOTES ON PRINCIPLES OF OIL Accomm ont By A. W. McCoy. ~Chester W. Washburne REPLY TO Discuss1on By C. W. WASHBURNE ON Noel ON Peete CIPLES OF Ort AccumuLATION.” A. W. McCoy RECENT PUBLICATIONS NUMBER V RECENT STUDIES OF THE UPPER CRETACEOUS OF TENNESSEE. Bruce Wade. : : 5 : : THE CHESTER SERIES IN Teamron, II. Stuart Weller . CONCERNING THE PROCESS OF THRUST FAULTING. Terence T. Quirke ae : PLEISTOCENE Mortecen FROM TTR AND “Oene. Frank Collins Baker 2 3 ; : FACTORS PRODUCING Conner Srectane IN See AND ITs OCCURRENCE NEAR MELBOURNE, AUSTRALIA. Albert V. G. James REVIEWS RECENT Banaiaomone NUMBER VI DIASTROPHISM AND THE FORMATIVE Processes. XII. THE Puy- SICAL PHASES OF THE PLANETARY NUCLEI DURING THEIR FORMA- TIVE STAGES. T. C. Chamberlin : ; : , Notes on THE MECHANICS oF GEOLOGIC Snuucmnes, Warren J. Mead PAGE 255 265 268 278 281 304 333 340 356 366 371 374 S01 T/ 395 417 439 458 470 471 473 5°5 CONTENTS TO VOLUME XXVIII PRELIMINARY DESCRIPTION OF A NEW SUBORDER OF PHYTOSAURIAN REPTILES WITH A DESCRIPTION OF A NEw SPECIES OF Phytosaurus. E. C. Case ; : és 4 5 Tue Heart MounrtTAIN ONEREST, ORE D. F. Hewett SUMMARIES OF PRE-CAMBRIAN LITERATURE OF NoRTH AMERICA. Edward Steidtmann NUMBER VII THE KATMAI REGION, ALASKA, AND THE GREAT ERUPTION OF 1912. Clarence N. Fenner A ; : : ; : On SoME PHYSICAL PROPER | OF Tee, Motonori Matsuyama A TEST OF THE FELDSPAR METHOD FOR THE DETERMINATION OF THE ORIGIN OF METAMORPHIC Rocks. Charles Gordon Carlson SUMMARIES OF PRE-CAMBRIAN LITERATURE OF NorTH AMERICA. Edward Steidtmann REVIEWS RECENT Deere TTONS NUMBER VIII DIASTROPHISM AND THE FORMATIVE Processes. XIII. THE BEAR- INGS OF THE SIZE AND RATE OF INFALL OF PLANETESIMALS ON THE MOLTEN OR SOLID STATE OF THE EARTH. T.C. Chamberlin . A PLEISTOCENE PENEPLAIN IN THE COASTAL PLAIN. PRS Ee Cleland : ; THE MECHANICAL [haieReecon OF jonas. lip Walter H. Bucher : GEOLOGIC Rinsonneeetnnes OF THE Soe Oe ¢ OF THE Taos Rance, NEw Mexico. John W. Gruner : SUMMARIES OF PRE-CAMBRIAN LITERATURE OF items Soaanae Edward Steidtmann RECENT PUBLICATIONS INDEX TO VOLUME XXVIII Vil PAGE 524 536 558 569 607 632 643 6590 660 abs EDITED BY \a AS is pee cn AND ROLLIN D. SALISBURY © y j ih the Active Collaboration of P ALBERT JOHANNSEN, Petrology "ROLLIN T. CHAMBERLIN, Dynamic Geology 4 '” ASSOCIATE EDITORS ; ts , RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. WILLIAM H. HOBBS, University of Michigan FRANK D. ADAMS, McGill University .__ _ CHL.RLES K. LEITH, University of Wisconsin WALLACE W. ATWOOD, Harvard University WILLIAM H. EMMONS, University of Minnesota * ARTHUR L. DAY, Carnegie Institution \GE OF THE EARTH DEDUCED FROM MARS, VENUS ’ eee Peiopeta i ay! taht ah Smee Ra = a ig Coy ARTES ORE TO eS BUOW: (TD 8 2 haa = ORY Siprerercont OLONIES OF THE DEVONIAN; THEIR ORIGIN, RISE, eye = = < = - - - Joun M. CLARKE Ee ze & ie 2 2 3 = a Cale DARE 5 z z my eG ake - CHARLES KEYis acai apy nS ete the % METEOR heer - - - — - - - - - ,= - - 7 ICAGOPILLANOIS, USA. , JOHN C. BRANNER, Leland Stanford Junior University - UIBRARY NUMBER 1» C xy See 18 38 61 » ' Journat or GEOLOGY, the University of Chicago, Chicago, Ill. . Yi: ' transit, and when the reserve stock will permit. January-February 1920 EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of STUART WELLER i) ALBERT JOHANNSEN ' Invertebrate Paleontology f ag Ay ROLLIN T. CHAMBERLIN Dynamic Geology Press, 57 50 Ellis Avenue, Chicago, Illinois, semi- se on or about the fellate ‘ihtes F chau a March 15, May 1, June 15, August 1, September 15, November 1, December 15. § The subscription price is $4.00 per year; the price of single copies is 65 cents. Orders for service of less than a half: year will be charged at the single-copy rate. {| Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. {[ Postage is charged extra as follo s For Canada, 30 cents on annual subscriptions (total $4. 30), on single copies 4 cents (total 69 cents) for all other countries in the Postal Union, 53 cents on annual subscriptions (total $4. 53), on sing copies 11 cents (total 76 cents). {[ Patrons are requested to make all remittances payable to The ve versity of Chicago Press in postal or express a orders or bank drafts. ' The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, vounge E. i: 4. Yearly subscriptions, including postage, 21s. each; single copies, including postage, 3s. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchowel Tokyo, Japan. Yearly subscriptions, including postage, Yen 9.00 each; single cOPRESH ee postage, Yen 1.50 each. on For China: The Mission Book Company, 13 North Beas Road, Shanghai. coe sub- scriptions $4.00; single copies 65 cents, or their equivalents in Chinese money. Postage extra, itt é mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies 11 cents. oa Claims for missing numbers should be made within the month following the regular month of publi- 1 1 cation. The publishers expect to supply missing numbers free only when losses have been sustained Business correspondence should be addressed to The University of Chicago Bias, Cees pt Communications for the editors and manuscripts should be. addressed to the Editors of ‘Tae Contributors are requested to write scientific and proper names with particular care and i in citations, to follow the form shown in the pages of the Journal. : Separates, if desired, must be ordered in advance of publication. Separates of ieadite articles will e 3 printed as ordered, of which 50 (without covers) will be furnished gratis, the actual cost of the Temainde! ; (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend 1 upon the amount of work in remaking the pages into forms, presswork, paper, binding, etc. Separates con- . taining half-tones may be expected to cost somewhat more, the increase depending upon the numbe of cuts and the amount of work required upon them. . ae y Entered as second-class matter, March 20, 1893, at the Post-office at Chicago, Ill., under the Act 0 of March 3 +1879. Acceptance for mailing at special rate of postage provided for in Section 1103, Act of October F 3) 1927, authoriz July rs, ror. Among articles to appear in early numbers of the Journal of Geology are the following: Notes on Principles of Oil Accumulation. By A. W. McCoy. Discussed by CHESTER W. WASHBURNE. Diastrophism and the Formative Processes. XI. Selective Aggre- gation of Material in the Formation of the Earth and Its Neighbors. XII. The Physical States of the Earth during Its Formative Stages. By T. C. CHAMBERLIN. (Two of a group of three papers to appear in successive numbers.) The fundamental nature of Professor Chamberlin’s recent cosmological studies is indicated by the titles of these articles. Compilation and Composition of Bituminous Coals. By REINHARDT THIESSEN. Dr. Thiessen has studied bituminous coals in great detail, and his article will be of great value to readers interested in his subject. The Origin of Gumbotil. By Grorcr F. Kay and J. Newton PEARCE. The authors present a satisfactory explanation of the origin of the gumbos associated with glacial drift. Geological Setting of New Mexico. By C. R. KEvEs. Paleozoic Diastrophics of the Northern Mexican Tableland. By CR. KEYEs. Slides in the Conemaugh Formation near Morgantown, West Virginia. By Ear R. SCHEFFEL. Factors Producing Columnar Structure in Lavas and Its Occurrence Near Melbourne, Australia. By ALBERT V. G. JAMES. POSITIONS OF ALL KINDS Never was the demand so great for qualified teachers and specialists. For ten years we have given our time and energy to this work. Write for our free literature. State qualifications briefly. Co-operative Instructors’ Association Marion : - Indiana The Origin of the Earth By Thomas C. Chamberlin A new chapter in “‘the Story of the Birth of the Earth.” xiit+272 pages, 8vo, cloth; $1.50 net, postpaid $1.65. One of ‘“‘The University of Chicago Science Series’’ The University of Chicago Press Chicago - - - - - Illinois (A reprint from University of Chicago Sermons) The Function of Death in Human Experience By GEORGE BURMAN FOSTER “One of the most philosophical, as well as one of the most solacing and beautiful, presentations of the great facts of death and life.” 25 cents (27 cents postpaid) THE UNIVERSITY Oe CHICAGO PRESS Chicago . linois A Short History of Belgium By Léon Van der Essen of the University of Louvain A Fascinating Story The author tells in an interesting way of the formation, struggles, and triumphs of the little kingdom. A new and enlarged edition presents additional material on the reign of Leopold II and describes The Heroic Nation’s Part in the Great War Although this is not a war book, it shows why the Belgian nation preferred honor in the place of dishonor and struggle for free- dom in place of ease. No one can understand nor appreciate the Belgium of today unless he knows something of its prolonged Seale for unity and independence. CHICAGO =- - $1.50, postpaid $1.65 THE UNIVERSITY OF CHICAGO’ PRESS Outlines of Geologic History with Especial Reference to North America Symposium Organized by BAILEY WILLIS Compilation Edited by ROLLIN D. SALISBURY EOLOGISTS and all readers of geologic Ce literature will welcome the publication, in book form, of an important series of essays and discussions on the subject of geologic correlation under the title, Outlines of Geologic History with Especial Reference to North America. | They embody the present state of knowledge and opinion concerning many of the funda- mental problems of North American geology, }| and form an admirable supplement to earlier | treatises and manuals. . The value of the book’ is greatly enhanced by the fifteen paleogeographic maps by Bailey Willis which accompany the papers. 316 pages, Svo, cloth; price $1.50, postpaid $1.65 The University of Chicago Press | Chicago - - Illinois | If you do not receive our advertising matter regularly ask to be put on our mailing list. THE UNIVERSITY OF CHICAGO PRESS - ILLINOIS Rate of Wear Unit= 70 total weight per mile S90 6 ERRATA In Mr. Wentworth’s article in the Journal of Geology, Volume XXVII, pages 507-21, a number of figures were incomplete and others not properly arranged. For the corresponding figures please substitute the following: 10 7 5 ° 20 40 60 80 IO0O 120 140 160 Size in Grams = This curve is under the following conditions: Rock—Niagara limestone Barrel Vel.—27 R.P.M. Barrel Diam.—24 inches Flushed with water Mixture—3,000 grams in 63-inch compartment Fic. 2.—Relation of rate of wear to size of cobble Rate of wear tooo total weight per mile 200 160 Size in Grams 120 40 fo) IOO 200 300 400 500 600 700 800 Miles Fic. 3.—Relation of size to distance traveled. Ideal history of cobble starting at 178 grams’ weight. 16 12 8 Unit 4 Non 50 100 150 Number of cobbles in mixture Average weight = 41.7 gms. Fic. 5.—Effect of amount of mixture in compartment of drum a Values of R ° Too 0200 300 Distance in Miles ROUNDING OF CUBES Fic. 26.—Graph of values of R plotted against distance for the series illustrated in Figures 8 to 25. r=radius of curvature of most convex point d=diameter through same point 2s Reg iii 50 40 Diameters in Millimeters 30 20 ° 5,000 10,000 Revolutions of Barrel (880 Rev.=1 Mile) ROUNDING OF CUBES Fic. 27.—Showing convergence of diameters of cubes rounded by cobbles averaging 70 grams each. iv 40 Diameters in Millimeters 30 20 fo) 5,000 10,000 Revolution of Barrel (880 Rev.=1 Mile) ROUNDING OF CUBES Fic. 28.—Showing convergence of diameters of cubes rounded by cobbles averaging 20 grams each. & Sarr eae > eee * VOLUME XXVIII NUMBER 1 THE JOURNAL OF GEOLOGY JANUARY-FEBRUARY 1920 DIASTROPHISM AND THE FORMATIVE PROCESSES X. THE ORDER OF MAGNITUDE OF THE SHRINKAGE OF THE EARTH DEDUCED FROM MARS, VENUS, AND THE MOON T. C. CHAMBERLIN The University of Chicago PRELIMINARY CONSIDERATIONS During the last century it was the prevalent view that the earth was once a white-hot liquid globe. It was a logical inference from this view that the subsequent shrinkage of the earth arose chiefly from a loss of heat and from effects incidental thereto. On critical inquiry, however, it was found that the contraction assignable to lowering of temperature was disappointingly small. On the other hand, it was found as field inquiry was extended that the sum total of surface shortening implied by foldings, crumplings, overthrusts, and similar evidence was distinctly large. As a result of these divergent disclosures, students of diastrophism came to feel not a little hesitation in following out fully and freely to their logical limits the trend of interpretations suggested by field evidence whenever very great shrinkage was foreshadowed. The restraint thus felt was much like that suffered during the same period from supposed limitation of geologic time, a restraint now happily removed. A radically new aspect, however, was given to the whole problem of earth shrinkage when, near the opening of this century, it was I T. C. CHAMBERLIN discovered that certain of the heaviest known atoms were spon- taneously giving out heat as an incident of their own disintegration. It was found that if radioactive substances pervade the whole body of the earth as richly as they do the accessible portions, the heat generated by them would be much greater than the amount the earth is now discharging at its surface. The certainty that the earth had been cooling at once disappeared; it seemed quite as likely that the temperature of the earth was rising as falling and, so far as heat is concerned, the volume of earth quite as likely swelling as shrinking. This cut at the very roots of the former tenet that shrinkage was chiefly due to the lowering of the tempera- ture. The possible potentialities of the new source of heat were not only embarrassing in themselves, but they made the great heat hypothetically inherited from the white-hot earth a superadded burden of embarrassment instead of the facile explanation of shrinkage and deformation it had once been supposed to be. Nor was this all: It was obviously necessary to devise some special hypothesis to obviate the surplus of heat the radioactive substances would give if they had a uniform distribution throughout the interior of the earth. Since no pressure or other physical condition is known to reduce appreciably the thermal output of radioactivity, a restriction of the radioactive substances themselves to a shallow surface shell seemed the only hypothesis available. But here the tenet of a molten globe arose as a new form of embar- rassment. The radioactive substances are exceptionally heavy, and in a liquid mass they should naturally concentrate toward the center, not toward the surface. Convection, of course, if it were sufficiently active, might be supposed to prevent much concentra- tion toward the center, but convection carries down as well as up and tends to give a more or less uniform distribution—just the distribution the hypothesis is seeking to avoid. Nor is this the limit of the embarrassments attending the old view. A molten state implies that the larger part of the potential resources of shrinkage were exhausted before the formation of a crust made a record of shrinkage possible, at least so far as shrinkage depends on arrangements and combinations of material. A molten state offers nearly ideal conditions for the physical adjustment of DIASTROPHISM AND THE FORMATIVE PROCESSES a all constituent elements and for such chemical combinations as are possible, except in so far as the heat itself stands in the way. With such extraordinary facilities for selective adaptation as were hypothetically offered by the passage of the earth substance from the assigned gaseous to the liquid state and thence at length on to the solidifying state, a large part of all possible adaptation to the demands of pressure—save those restrained by heat—should have taken place before the record of diastrophism began. About all shrinkage left to be registered would be the meager amount that might spring from cooling. Thus in several vital ways the inherited theory of a molten earth came to be a source of embarrassment to investigators who were struggling with the specific demands made by the field evi- dences of actual diastrophism. THE WORKING FITNESS OF THE ALTERNATIVE VIEW However, the case was desperate only from the traditional point of view. These embarrassments may be avoided if the gaseo-molten_ hypothesis is replaced by some form of the view that the earth was built up by the accession of solid particles brought to the earth in succession at intervals. In an earth so built the aggregate should have retained very nearly its maximum resources of combination, adjustment, and compression, while at the same time it was contributing only a small measure of heat to embarrass the threatened oversupply from radioactivity. Furthermore, this view affords an easy and natural explanation of the concentration of radioactive substances at the surface. Under this view the radioactive particles came to the earth at random with the rest of its material. ‘Their spontaneous heat was readily radiated away until they became buried to depths that _ prevented its ready escape. They then tended to become centers of local liquefaction. In so far as this was realized they became enveloped in their own mobile products and were thus carried by the extrusive agencies up to the cold zone or the surface. The liquid blebs thus generated, carrying their self-heaters with them, were well equipped for making exchanges with the eutectic sub- stances encountered on their way out and for concentrating these 4 T. C. CHAMBERLIN in the ascending thread of lava, leaving behind the less eutectic substances. ‘They were thus well fitted to fulfil three functions: (x) to flux their way upward, (2) to separate the more fusible material from the less fusible, and (3) to carry the former out to the cold zone or the surface, taking along the chief source of heat and the heat already generated. By thus draining away selectively the more fusible elements in the mixed material and raising the mean resistance of the rest to fusion, they help to maintain the solidity of the main mass. It is obvious that adjacent threads of hot self-heating lava must render one another assistance in mutually uniting and massing their forces for fusing their way outward. After such tracts have been drained of their more fusible substances and the conduits closed, new paths in ground less depleted of its eutectic material would naturally be chosen and thus the selective work should at length cover the whole field and raise its mean fusion-point, while the self-heating radioactive particles were more completely removed. At all stages of this selective process it is held that the differ- ential stresses of the earth body lent effective aid in extruding the liquid matter. The great pervasive stresses of the earth, static and dynamic alike, are intensest in the deep interior and graduate outwardly. The differential components of these are well suited to squeeze toward the surface the liquefying portions of the interior matter about as fast as these accumulate in sufficient quantities to respond readily to such stresses, while the liquids themselves readily yield to the rise by reason of their heated state and the gases they gather to themselves. To this doubly facilitated extrusion of liquid matter carrying its special thermal source with it is assigned the function of clearing the depths of their original radioactive substances and of their heat products, with the in- cidental effect of perpetuating the solid state of the earth as a whole. If the simile may be pardoned, the lquid ‘lineade may be likened to the sweat pores of an organic body, regulating its temper- ature by a natural perspiratory system. The pools of lava that at times accumulate at the surface function as the sweat drops of the earth body. They seem large, to be sure, in terms of ordinary DIASTROPHISM AND THE FORMATIVE PROCESSES 5° measure, but they are really quite minute when compared with the 260,000,000,000 cubic miles of the earth mass. It will be noted that this view is a reversal of the old inter- pretation. Instead of being a residual effect of former excessive heat, extrusive igneous action is initiated automatically within the body and forms a regulative system through which the solidity of the globe is maintained. The extrusive action is at the same time merely a minor feature in a general genetic process that has con- served the potential resources of shrinkage and rendered them available at such successive stages in the history of the planet’s evo- lution as developed the conditions necessary to call them into action. But, notwithstanding the ampler possibilities of shrinkage which this newer view places at the command of students of diastrophism, the pall of restraint has not as yet been wholly lifted. Thus far there has been no well-grounded estimate of the total earth shrinkage that has actually taken place. Even a theoretical estimate of the shrinkage available for interpretative assignment is still lacking. Workers in this field are thus still more or less under the shadow of restraint. Research will certainly proceed with more equipoise if workers can feel wholly untrammeled by supposed limitations in following to their logical conclusions any leadings of evidence they may encounter, even though its demands may be greater than general considerations have thus far seemed to warrant. THE SPECIFIC FIELDS THAT YIELD DIASTROPHIC EVIDENCE A glance at the field evidences of diastrophism will further pre- pare the way for a study of the probabilities of the case. Diastro- phism is displayed in three great fields. These are closely related,, to be sure, but yet sufficiently different to require individual recognition. I. The first embraces the deformations of the distinctly strati- fied terranes, chiefly those of the Paleozoic and later ages. These are relatively accessible, and the constituent formations usually so far retain their individuality as to be susceptible of being satis- factorily traced throughout the whole tract involved in the defor- mation under study. 6 T. C. CHAMBERLIN II. The second embraces the complicated distortions and the metamorphosed phases of the Proterozoic and Archean complexes. Usually these are only partially accessible, and the profound changes they have undergone present formidable difficulties in addition. III. The third includes the deeper and more massive deforma- tions of the earth body. These can be regarded as accessible only in a logical sense by means of indirect evidences or remote intima- tions, or by a priori considerations. I. In the more surficial of these three fields estimates of crustal shortening have been made from time to time in the past, but in the main these have been confined to linear shortening; they have not included the depths involved in the shortening. This is necessary for computing their total quantitative values. Nor has there usually been any determination of the under-configuration of the distorted masses. This carries a very important part of the specific significance which the diastrophism embodies. A notable beginning has been made in adding these two neglected factors and increasing at the same time the reliability of the estimate of the linear factor.‘ But the labor involved in these more adequate determinations is so large that much time must pass before a sufficient number of such determinations can be made available for a total estimate of the shrinkage involved in even the limited field to which the method is adapted. II. In the Proterozoic-Archean field there is little ground to hope for any general application of these superior methods, partly because of the large measure of concealment of the terranes and partly because of the excessive intricacy of the structure and the frequent changes in petrologic nature which render sharp identi- fications of the borders of the several members of the terrane throughout the whole folded tract impracticable. The difficulties of this field are formidable in the extreme. No one, so far as I know, has thus far had the temerity to offer an estimate of the amount of shortening implied by the intricate crumpling of these t Rollin T. Chamberlin, “‘The Appalachian Folds of Central Pennsylvania,” Jour. of Geol., XVIII (1910), pp. 228-51; ‘‘The Building of the Colorado Rockies,” ibid., XXVII (1919), pp. 145-64, 225-51. DIASTROPHISM AND THE FORMATIVE PROCESSES 7 old formations on any great circle of the earth. That it was large, however, goes without the saying. But, taken at their best, the deformations in these two fields are merely surficial. Such foldings as are accessible are mere wrinklings of the skin of the earth body, mere lineaments of the face of the earth. They have about the same relation to the effective framework of the earth body as the shriveled integument of an old man has to the bony skeleton that chiefly gives form to his figure. III. The deeper deformations of the earth have been little more than a field for the imagination thus far. And yet they have given rise to indirect and implied evidences. There are the protrusions of the continents, the sags of the sub-oceanic basins, and the general configurations of the globe. There are tidal, seismic, magnetic, and other dynamic lines of approach. Great light has been thrown on the problems of the interior by the brilliant determination of the value and nature of the body tide and the elastic rigidity of the earth by Michelson and Gale on the experimental side,t and Moulton on the mathematical side. The seismic evidences gathered by many observers indicate that the elasticity of the earth increases downward faster than the density for at least a depth that involves much more than half the volume of the earth. These trenchant determinations bear vitally on the interpretation of the internal deformation of the earth. The lines of approach now available for an interpretation of the master-features of.the earth’s surface promise at least some insight of value into the earth’s fundamental diastrophism. I have ventured to interpret these master-features as simply the adult products of a segmentation that sprang from primitive shrinkage - stimulated and shaped by oscillating rotation and tidal strains. Under this view there are cogent reasons for assuming that the original segments were more or less unequal and asymmetric, and tA. A. Michelson and H. G. Gale, ‘“‘The Rigidity of the Earth,” Jour. of Geol., I (1919), pp. 585-601. 2 F. R. Moulton, ‘‘Theory of Tides in Pipes on a Rigid Earth,” Astrophys. Jour., L (1919), pp. 346-55. 3 The Origin of the Earth (1916), pp. 200-224. 8 T. C. CHAMBERLIN that the large inequalities and asymmetries now observed are largely due to later shiftings, distortions, and outgrowths of the primitive elements. These then are the special subjects of study in the third field of diastrophism. THE PARTICULAR OCCASION FOR THIS INQUIRY Now in a recent study of what could plausibly be assigned to original irregularity in segmentation and what then remained to be assigned to subsequent movements and unequal growths, I was led to see, or to think I saw, evidence of a system of shiftings and of unequal growths which marshaled themselves in a singularly rational way as though they were due to systematic causes of a general nature. The particular adjustments appeared to be such as were directly implied by the configurations which the great features now bear. By reasoning back from the present con- figurations to the assigned primitive configurations, rather specific amounts of shiftings and deformations, abetted by unequal out- growths, seemed to be indicated. The amounts of these shiftings were distinctly larger than the movements commonly assigned to diastrophisms in the surficial fields. Because of this largeness the question, How much shrinkage can reasonably be assigned the earth during its whole history? came ‘up in a new and specific form, and with especial piquancy by reason of unexpectedly exacting demands. | : COMPARISON BETWEEN THE EARTH AND ITS NEIGHBORS In casting about for some independent means of estimating such reasonable possibilities or even probabilities of shrinkage as there ‘might be under the later view of the constitution of the earth, a comparison of our planet with its near neighbors, the moon, Venus, and Mars, suggested itself, as also a comparison with an ideal earth built of material of the average meteoritic type. The earth, Venus, Mars, and the moon form a little group of closely related bodies revolving in the inner part of the sphere of control of the sun under very similar dynamic conditions. We naturally think of them as widely deployed, but, taken all together, the little group spans less than 3 per cent of the radial reach of DIASTROPHISM AND THE FORMATIVE PROCESSES 9 the solar system and probably not more than a thousandth part of the radius of the sun’s sphere of control. This last is the more significant standard, for the sun’s sphere of control is the dynamic field within which the planets had their origin and have'ever since had their being. It is therefore a reasonable inference that the members of the little group shared much the same evolutionary conditions, were formed in much the same way, and of much the same material. But even if there was some gradation in the nature _of the material due to position in the system—which we will con- sider later—its effects are measurably equated in the comparisons, because Mars lies outside the earth and Venus inside, while the moon, as a member of the earth system, presumably partook of the common material from which both earth and moon were derived, though perhaps not in precisely the same way. At any rate, though some differences of original material must be presumed to have entered into the formation of these four bodies, such differences could scarcely have been at all radical. Besides, it will be seen later to be possible to deduce the more important differences . which affected the selection of material in the formation of these bodies. This will be considered in an article following this one, as it goes too far afield to be introduced here. All members of this little group of bodies are small compared with the four giant planets outside them, ‘and yet they are large relative to the majority of satellites and planetoids. They form an intermediate group, and deductions respecting them may be checked by the extremes on either hand. Among themselves they form a graded series well suited to our purpose. The moon is distinctly small and has no appreciable atmosphere or hydro- sphere; it may be taken to represent such bodies as are formed of molecules heavy enough and sluggish enough to be controlled by a limited attractive force, a force too feeble to hold the lighter and swifter order of molecules. Mars represents a stage of growth at which sufficient gravitative power has been reached to maintain a limited atmosphere and apparently the beginnings of a hydro- sphere. Venus represents a much more advanced stage at which the gravitative power is sufficient to hold a very notable atmosphere and probably a rather massive hydrosphere. The earth, as we IO T. C. CHAMBERLIN well know, represents a stage at which a notable atmosphere and a distinctly massive hydrosphere have been acquired and held. The four bodies thus represent those stages of evolution which are most significant in such a study as this. They are all notably dense compared with the great planets that lie outside them and with the sun at the center of the system. ‘The moon, Mars, and Venus will . be treated as representing a typical series of stages of evolution connecting the small atmosphereless type with the largest known cold planet enveloped with a deep water-sphere and gas-sphere, the earth. No special study will be given to the large hot bodies of low density that form the great outer group.” STATISTICAL DATA Some of the more essential statistics on which the study will be based are gathered into Table I. TABLE I* Planet Mean ameter Mass Earth=1 | Density Water=1 Su ete Moonesae erences 2160 0.0122 3.34 0.16 (3476 kms.) / Mars? ater ues 4339 0.1065 3.58 0.36 (6983 kms.) WenUSe 2 ran neti: 7701 0.807 (?) 4.85 (?) 0.85 (?) : (12394 kms.) Harthines ns neers I .000 Bese I.00 791 (12743 kms.) * These statistics are taken from Moulton’s Introduction to Astronomy, Revised Edition, 1916. Venus has no satellite and its mass and density can only be determined by indirect means which are not very accurate, and hence the figures for these are marked with an interrogation point, but they are probably close enough for our purpose. The figure 5.53 for the density of the earth is conservative; figures as high as 5.56 and 5.57 have been used. These higher figures would give greater shrinkage. The dimen- stonal data are given in miles and in kilometers, but the computations are carried out in miles, because, being the larger unit, it is the more convenient. An even larger unit is desirable for most earth studies and so the standard degree of a great circle of the earth is added in circumferential measurements. Degrees are convenient units in working with globes. THE METHOD OF THE INQUIRY As a step preparatory to the proposed comparison there were built up from the moon, Mars, and Venus, each in turn, by using material of its own mean density, parity-earths whose masses were equal in each case to that of the actual earth. A similar parity- earth was built up of mean meteoritic material. The radii and volumes of these parity-earths were then computed and taken as *Cf. W. D. MacMillan, “On Stellar Evolution,” Astrophys. Jour., Vol. XLVIII, No. 1 (July, 1918), pp. 40-41. DIASTROPHISM AND THE FORMATIVE PROCESSES II the basis of shrinkage. The parity-earths were supposed to shrink until their mean densities were identical with that of the present earth. The amount of this shrinkage is recorded in Table II in terms (1) of the earth’s radius in miles, (2) of the earth’s circum- ference in miles, and (3) of the earth’s circumference in degrees, each of these being more convenient than the other in certain specific uses. In building up the meteorite earth Farrington’s mean specific gravity of meteorites seen to fall was taken as the basis of compu- tation.t While the inclusion of only those meteorites that have been seen to fall may not be strictly representative, it is Farrington’s view that this limitation gives the best definite figure that is available. If the meteorites found but not seen to fall were included, the specific gravity would quite certainly be too high, because metallic meteorites are more likely to attract attention on account of their unusual heaviness and the whitish color of the metal, and because they are less liable to disintegration than the stony mete- orites. Nevertheless, if all meteorites that have reached the ground in observable masses were averaged, the mean specific gravity would probably be greater than the figure given. On the other hand, the surfaces of iron meteorites are notably pitted, due probably to the exfoliation of the stony parts, as these are less tenacious than the metallic parts. A naked body sweeping about the sun and likely to be in rotation is quite sure to be subjected to those rapid changes of temperature which promote exfoliation. — The gravitative power of a meteorite is very small and hence these exfoliated chips would be likely to be thrown off into separate paths and thereafter play the part of individual meteorites. It is thus probable that the vast multitude of small meteorites that are burned to dust in the upper atmosphere are much more largely stony than metallic. This consideration probably offsets any weight that ought to be given to the preponderance of metal among the meteorites found some time after their fall. At any rate the mean given by Farrington is the best available and is doubtless near enough the true mean to give the right order of magnitude to the results deduced from it. 10. C. Farrington, Jour. of Geol., V (1897), pp. 126-30. 12 T. C. CHAMBERLIN While meteorites in the main seem to belong to the solar system, they appear to be samples from many sources, for they are extremely numerous and come to the earth from various directions and at very different velocities. It is therefore thought that they fairly represent the nature of any kind of scattered interplanetary matter of the solid type that might once have been available for the formation of small planets and satellites. This view does not rest so much upon their present status as upon the dynamics of the case, for the self-aggregation of small masses implies feeble gravitative control, and under the conditions of such feeble control only the heavier, sluggish molecules can be gathered and held. For this reason meteoritic material is taken to represent the densest type of scattered solid particles and small masses, whether planetesimal, satellitesimal, meteoritic, or otherwise, available now or in the past, for the growth of satellites and planets. This of course does not exclude the availability of lighter material, even gaseous material, to planets massive enough to hold such material, nor does it exclude occluded or combined gases from even the smallest bodies. In building up these parity-earths, the series starts with the moon, the lowest in mean density, rises thence through Mars and the representative meteorite, to Venus, next to the earth in mean density, and ends with our planet. This arrangement should suggest at once that as the last two are the most massive bodies and hence have the greatest power of holding light molecules, they probably have the largest proportions of inherently light matter in their composition. THE NUMERICAL RESULTS The leading numerical results of the computations are gathered into Table II. Let us hasten to admonish ourselves that these results are as yet uncriticized. Before the inquiry may properly rest these results must be scrutinized in the light of the dynamical conditions under which the four bodies were formed, for these conditions were such as to determine the inherent heaviness or the inherent lightness of — the matter that formed them. This critical phase of the study will take us rather far afield and must therefore be deferred to a later article. I feel warranted, however, in saying that this further study will indicate, as does the hint given above, that the more massive DIASTROPHISM AND THE FORMATIVE PROCESSES 8. bodies in all probability contain the larger proportion of light atoms and molecules and that the shrinkage figures of Table II will need to be somewhat increased to satisfy the natural intimations of the laws of planetary organization. For the moment, however, let us regard these prospective increments merely as a measure of assurance that we will be forming first impressions on conservative grounds if we tentatively review the results as they stand. If this review shall raise any questions as to the validity of the results deduced these questions will serve to give piquancy to the deferred discussion. TABLE II* 5 5 & | SHORTENING x ; 3 zZ S oF Parity- Basis oF Pariry-| » ll | «% | Present Votume | Pariry-Votume | %u |Z 3. a)! Crrcum. Earta Be ac Cusic Mites Cusic Mites rales = ze n Bo we | As | a” E> | frie Mites) recs Moon........ 3.34) 1080 5,270,678,626) 430,353,000,000) 4684) 725 | 4555| 66 Mars. . arcane 3.58| 2170] 42,802,460,494| 401,502,000,000) 4577] 618 | 3883) 56 Meteorite..... BOC! tas oblla5 eae 380,506,000,000] 4531] 572 | 3504 52 Wentiseeese oi 4.85] 3851] 230,226,992,649| 296,367,000,000] 4136) 177 | 1112) 16 SBianthy 5.002. G53) 3050 250,0235940)377| «0. s5--+ ae. aaa oes ea ae * The parity-earths may be derived either from the relative densities or the relative masses. The results, however, are not strictly identical in all cases, doubtless because the figures adopted are the weighted means of different methods of determining the masses and densities and these thus lose strict consistency with one another. The differences are not enough seriously to affect the order of magnitude of the shrinkage results. PROVISIONAL DISCUSSION OF THE RESULTS On first thought it may seem that the observed densities of the four bodies compared can be easily accounted for by assigning such specific gravities as are requisite to the material that entered into their formation. Thus the computed amounts of shrinkage may seem to be avoided. If it is legitimate to make purely arbitrary assignments in neglect of the laws of cosmic organization under such hypotheses of genesis as are tenable, no doubt this might be done. But in a naturalistic inquiry that tries to be thoroughly loyal to cosmic laws, so far as the inquirer knows them or can find them out, arbitrary assignments have little or no place. We are here dealing with highly composite results, the products of natural processes of organization. Each of the four bodies is believed to have been formed by a multitude of accessions brought together by forces of like types, acting under similar conditions and surrounded 14 T. C. CHAMBERLIN by similar dynamic environment. It does not seem naturalistically probable that arbitrary variations of sufficient moment to affect the average order of results could have entered into these combi- nations so closely analogous in generalnature. It is well recognized that under the law of probabilities a multitude of random contribu- . tions, uniting under common conditions, give closely concurrent averages even though the individual contributions may be highly variant. It will be seen from the discussion in the succeeding article that a very definite law probably presided over the proportion of the inherently heavy to the inherently light material which entered into the formation of the four bodies compared. Taking then their systematic organization for granted for the time being, the following tentative points are to be noted: ; I. The total shrinkage of the earth implied by the comparisons is very large. A circumferential shrinkage of 4,555 miles in a putative growth from a moon stage by the addition of moon-stufi is certainly large. A similar shrinkage of 3,883 miles in a growth from a Mars stage by the addition of Mars-matter is quite as notable; and a shrinkage of 1,112 miles in a growth from a Venus stage—a stage in which 80 per cent of growth has already been attained, while the material added has the high density of Venus— is even more remarkable. These large shrinkages are ample to meet all the demands that gave rise to the inquiry and leave a good ' working margin beside. ) II. Since the four bodies were treated as spheres, the computed shrinkages apply to all great circles, meridional, oblique, or equa- torial, equally. The special deformations that may be assignable to changes in the rate of rotation are not here included. There was probably always some equatorial bulging and polar flattening, but the geological evidence does not seem to imply that defor- mations of this class were essentially greater during the early ages than they have been during the later ages.‘ The large shortening in meridional circles given by the computations satisfies the require- ments of the Archean crumplings and related phenomena of the high latitudes, which seem to be essentially as great as those of low latitudes.? x“The Tidal and Other Problems,” Publication No. 107, Carnegie Institution of Washington (1909), P- 51. 2 Loc. cil. DIASTROPHISM AND THE FORMATIVE PROCESSES 15 THE STAGE OF MAXIMUM SHRINKAGE If the four bodies under comparison are to be regarded as representing stages of growth, it is a matter of much added interest to deduce from the comparison the stage of growth at which the greatest shrinkage took place. If the bodies were entirely compact at the outset, as they would be if fluid, or pliantly viscous, shrinkage from gravitative pressure might be expected to decline with every stage of compression reached, because resistance to compression usually increases rapidly as compression proceeds; but if the material of growth were minutely fragmental at the outset and the particles rigid and elastic, other factors of importance would come in. At first the porosity would be great. Until the porosity was exhausted the shrinkage would depend largely on the rigidity and the elastic qualities of the constituent particles. Later, the possibilities of chemical, crystalline, and physical readjustments in the interest of density would come into service. Another factor is the presence or absence of effective wash, solution, and redepo- sition. ‘These are dependent on the presence or absence of an effective hydrosphere. The moon has neither appreciable atmos- phere nor hydrosphere and if originally built up of minute rigid particles it would retain a deep porous zone of relatively low specific gravity. This would notably affect its mean density. Besides, there is evidence of much explosive eruption and the pyroclastic products arising from this would, in the absence of wash, solution, and redeposition, remain highly porous. :The Mare once regarded as seas and later as lava plains may perhaps really be tracts of volcanic ash. Lines of projected débris crisscrossing Mare Imbrium are well shown in a recent photograph by the too- inch reflector of Mount Wilson. Mars is on the ragged edge of doubt; it may perhaps have enough water on its surface to wash fine material from the exterior into the interior and to dissolve more or less of the surface material and deposit it in the pores below, or, on the other hand, the water may be so scant as to have little effect in cementing and solidifying the outer zone of the planet. But in the case of Venus, inwash and cementation are probably efficient, while they are known to be on the earth. All these factors seem to have played important parts in the results. 16 T. C. CHAMBERLIN An inspection of the mean densities themselves gives some hint of the general nature of the compression: 3.34 for the moon, 3.58 for Mars, 3.69 for meteorites, 4.85 for Venus, and 5.53 for the earth. It is notable that the mean density of meteorites falls between the two bodies suspected of deep porosity and the two bodies in which wash, solution, and cementation are effective. The results given in Table II bear an analogous import: 725 miles radial shrinkage for the lunar parity-earth; 618 miles for © the Martian parity-earth; 572 miles for the meteorite parity-earth; - and 177 miles for the Venus parity-earth. However, these shrink- ages represent quite different ranges of growth; to be strictly comparable they must be reduced to a common basis. A conven- — ient unit is an increase equal to 1 per cent of the mass of the earth. This is equal to the weight of about 14 billion cubic miles of water. Reducing the several shrinkages to this unit of mass increase, they become: for the mean rate of shrinkage between the moon stage and the mature earth, 7.44 radial miles per unit increase of mass; between the Mars stage and the mature earth, 6.90 radial miles per unit; and between the Venus stage and-the mature earth, 9.17 miles per unit. This brings to attention the very suggestive fact that the rate of shrinkage per unit of mass increase 7s greatest in the last stage. Next to this, it is greatest in the growth from the stage represented by the moon, the body suspected of being the most porous, and the least affected by wash, solution, and cementation. The first seems to imply that massiveness is the dominant influence. Next to this porosity seems to be influential. These inferences will appear to be still more strongly suggested if we reduce all the four natural bodies to parity-bodies, using mean meteoritic material as the basis. The results appear in Table ITI. The third column shows that if the moon had been built up to its present mass with material of the mean density of meteorites, its radius would fall short of what it actually is by 35 miles; if Mars had been built up in a similar way its radius would be 22 miles short, while the radius of Venus built in the same way would be 367 miles greater than it actually is, and the radius of the earth under like conditions 572 miles greater than it is. If all these bodies were actually built up of mean meteoritic material the figures DIASTROPHISM AND THE FORMATIVE PROCESSES 17 would seem to mean that the porosity of the moon is represented by 35 miles in terms of radius, over and above such compression as its center has suffered. This may be taken tentatively as representing the deep porosity of the moon. Similarly, the porosity of Mars would be represented by 22 miles in excess of its central com- pression. On the other hand, the actual compression of Venus seems to be represented by 367 miles, while the corresponding figure for the earth is 572 miles. TABLE III COMPARISON OF METEORITE-PARITY WITH ACTUAL BODIES Shrinkage “ieee THES Bersecn Banity- co . Pen ss ass Units ody an Body We anced ae ae eats t Per Cent Earth |Actual Body per Mass Unit =1 Per Cent Earth a Mass Miles Miles Miles Miles Moony: >: 1080 1045 — 35 1/22 —28.7 Wars yet... 2170 2148 — 22 10.6 — 2.0 Venus 3851 4218 +367 80.70(?) + 4.5 Riaaplieee o.5. 3059 4531 +572 100.00 + 5.7 In the fifth column the degrees of compression are reduced to a common unit-mass. ‘This brings out the essence of the matter in a striking way. It appears that the moon, built up as it actually was, failed to compress itself to the meteorite standard by 28.7 miles per unit of mass-growth, and Mars by 2.0 miles per unit, while Venus compressed itself beyond the meteorite standard to the extent of 4.5 miles per unit of mass-growth, and the earth by 5-7 miles per unit. This seems to put the first two bodies in one category and the last two in quite another category, while it greatly emphasizes the progressive nature of the compression from the least to the greatest, even per unit-mass of increase. Let us, however, hold all these tentative results in abeyance until we have more critically considered the probabilities in respect to the inherent nature of the material that entered into the con- stitution of these four bodies. Meanwhile this preliminary in- spection may serve to give point to the study of the genetic con- ditions that affected these results. This study will be the theme of the succeeding article. THE LAWS OF ELASTICO-VISCOUS FLOW. II* A. A. MICHELSON University of Chicago In a paper of Harold Jeffreys entitled ‘“‘The Viscosity of the Earth’? the author makes use of a formula which combines the laws of Larmor and of Maxwell. See See I Fat dt Tr The integral implies a permanent set which, as the author indicates, would be inconsistent with the ‘‘accepted theories of tidal friction and variation of latitude. Hence 7: must be practically infinite.” The formula is thus reduced to the expression F=n, S tre Experiments made on a great variety of materials show, how- ever, that this expression must be seriously modified to represent the facts. Thus it has been shown} that the displacement produced by a stress P is given by the expression S=C,Pe"? +C,Pe”" (1 —e -*V#) + C,Pe"t?.4 The last term produces permanent set, so that for the : : f I I if present it may be omitted. Putting C,= and C, er = Vi. I 6 ae [acre(s -<)| whence nN nN ; ds Nt m( Str 4 =P(4.+%4,) 0 “The Laws of Elastico-Viscous Flow. I’ appeared in Jour. Geol., XXV (1917), Pp. 405-10. 2 Monthly Notices of the Royal Astronomical Society, LXXVII, No. 5. 3 “The Laws of Elastico-Viscous Flow,” Jour. Geol., XXV (1917). 4 The strains in these experiments were torsional, thus involving only the rigidity constant x. In the formula as given in the paper referred to the coefficients C and the exponents / are functions of the temperature. The stress P is constant and p is approximately one-half. this becomes 18 THE LAWS OF ELASTICO-VISCOUS FLOW me) For small stresses A,=A,=1, and if 7, equal , this expression takes a form resembling that given by Jeffreys. 3 It is important to note, however, that this formula is based on the assumption that the viscosity is ‘‘external,” that is, it acts as : ; eens though the viscous resistance were due to an absolute velocity dt But this is by no means evident; and indeed the probability is that a considerable part if not the major part of the viscous resist- ance may be “‘internal,” that is, due to the relative motion of parts. Thus if an element consists of two parts y and z, y being coupled to the next adjacent element by an elastic coupling 2, and with z by an elastic coupling 7., together with a viscous coupling e,, while ef and e: represent the “‘external”’ viscosities, the equations of motion will be pr3 =e.(3— y)+n.(2—y)+el3 6 3 ; i d*y p2) =e2(3 —) IrmE—y)TeaIt mss If p.e and e. be considered negligible, the solution, for not too rapid extinction, is z=ae—** cos p(t—vx), in which put ae) TPE p N2(N2.—pp’)+ pre? B= p/spre 27/ Mi|n2(m2—pp?)+ pe?” lf pe is large compared with 1, v= "(1400 p é p= er 2ev’ so that in this case the higher the viscosity the less rapid the decay of the oscillations—quite the reverse of the conclusions on the former 20 A. A. MICHELSON assumption. But the appearance of = 1/7 is a more serious matter, making the use of the formula much more difficult. The operator which should replace 7 is therefore _ds I+ ary te rene apeeae But the application of this formula to such a problem as the earth’s viscosity is still further complicated by the fact that all the constants are functions of the pressure and of the temperature in the earth’s interior. Even though more or less probable assump- tions may be made regarding the value of temperature and pressure as functions of the distance from the center, we know but little regarding the effect of these factors on either rigidity or viscosity. . It was found that the temperature effect may be represented with considerable accuracy by the expression : A = pS) 0 2 in which P is the applied stress, 9 the temperature, and E, K, and b constants. For room temperature the values of 60=h are given in Table IV. ; If we take h,=0.2 as fairly representative S,= Pee Pe Vt). The unit P=100 gm., so that G the couple=Pr gm. cm. Thus we get for the displacement, after a sufficiently long time, WY ea coy 2 TERS S=Pee and z : 2 ty THE LAWS OF ELASTICO-VISCOUS FLOW 21 Table I shows the very great increase in importance of the elastico-viscous term for large stresses. The same is also true for the purely viscous term. TABLE I JB = S/P GER cies wait iarenstan ele c's I ES BiG CHORE C ea Cone Tee OPP edt araie tus Sav olaeaiqret As: ¥ a6 7 ROM Mic rie eioais cast aes s 20,000 2 - Ne ° : Table II gives the ratio = for twenty-two materials, showing that there are certainly two elasticities, one of which is not accom- panied by viscosity and the second is thus affected. In every case excepting that of sealing wax, where the ratio is unity, the second elasticity is much greater than the first, and in some cases enormously greater. TABLE II Na Ne x Mx TTT 5 piglet eee GOES late Tak een hea mien ete 150 LC s.2¢5 Se Ae a eee AO SAL iey cu pene vate teotesia ree aneveueneyeate 50 JEON ES ae en eae Aa SOAPSTONE A siete orto © a5 WAMIESTONE Ho). 2 aed os: FOU) MCAS Lee hrs be ee yceyare epee 40 WU QUIECN Ce e.sialee inicio < Seaa Cachiiiiuine koe Gh mip eer 65 LEROTIL. 4 Sic ae een DIS) NCO GLA ARB AeA OC Chal O 150 SES. 35 ee ene TP OOOM MiapaeSIUIN eye mie eee tials 250 CUUDSE: ee ae a aa rey BCD SNIKN Nd oe cee MeO ale come 7 PUMA, ss hs ke AVAGO! EVOLY — ee ete lets teria 50 LOS. RSG See eee ae Hose USM ay oda dodou dod. dee 80 Class. Gatien ea TOO! SCALING Wak aie aie ae ca erte vic. I The introduction of 1/¢ instead of ¢ itself is a step so radical that it may be well to give an illustration in its justification. For this purpose it is desirable to choose a material in which the elastico- viscous effect is well marked. This is notably the case for vul- canite, which has the added advantage of the relatively small importance of the third or purely viscous term. This illustration is perhaps the most striking in showing the appropriateness of /é instead of t; but all the materials investigated give similar results. 22 . A, A. MICHELSON Table III is a table of results for R., the return at the time ¢ after releasing the stress.‘ R,/, gives the result of calculation from R=890(1—e *7V#) - R, gives values calculated from R=840(1—e" °”’) . The differences between calculated and observed values under A, and A, show that the former expression is very near the truth, while the latter is entirely inadequate- TABLE III Zt Ro Ry; Ar R; : Aa TA atetctsheic teks 380 387 7 277 —103 Dik Sect eee 490 492 2 462 — 28 7 aay SPSS 600 605 5 672 + 72 Ohi thse 730 720 =i 820 + go TO eee ee ei 800 802 D 838 + 38 Desc nusyetlene - 840 841 I 840 (ojo) BO se erase hela 853 851 =2 840 — 13 OOiicatielacus ee 890 890 fo) 840 — 50 While the term involving a permanent set may not have any application to the problem of the earth tides, yet it may not be amiss to draw attention to the fact that in some cases and espe- cially at temperatures approaching the melting-point, this term becomes the most important of all. The temperature coefficient in this case enters in the form 6/T —86, giving as it should perfect fluidity at T, the melting-point. In the former article the expression given for this viscous term is S;=(Ft)?, in which F =C,Pe’ and p is stated to be approxi- mately one-half. From more recent data the average value of p is .41; and if from the nineteen substances examined four be excluded the average is .35, which makes it much nearer one-third than one-half. ™It was found by experiment that for stresses not too great the “direct” curve (on applying the stress) and the ‘“‘return” curve (on releasing) were the same; or rather if the former is S and the latter R, then S+R=Ci. a THE LAWS OF ELASTICO-VISCOUS FLOW ¢ fo) o'L ‘Auo} I'o go foie) v fo) Zz 9 ogh ORE OO Oa ON OOS OOO. Otho rood JOATIS oexe) ore) Silelie) 6) 60a) phe be 2(o) OL lolol fefe) mute ‘oc? ,O1 OS CRORONT CaiC eC On ah uO COR OO: Ot tnd Cat 0 AIOAT gO eae. [fou a ayos05 z‘O ZI 00° oo ‘PI "096 oot alae aielerle fal miwleue)isviesiay eiielieiaianaitsline sii sitens aye oo amar 9°0 v'o Cz lolol oS °Q v6 oofz COMmoOO OOO WOOO oon G sess UMISOUse IAT fS-o oO’! Sure) F Ano} to 00 O'L ZO ¢ oS O*O Git WOOO Cay) ONO OO Oni scr Oi 20) (089) co oO'L g°0 ore) z ro) oo Vz vs ol oor ale. «(se 6-6) 0.6: e)\6|\0)\0/ ele) @)/6) «)el-wis) eile) w **uunuaIpe) to I'l g°€ 6°1 ZO fore) Y a or ooSt Mahe feruevieelvijelekelichiedyselaiyiiclivis lefeiissjisieas hice aarems peo'T to o°T go Se: fon zo Bi o'r v8 OOU Se Spee Ness ON asvAraio jarjered oyeys eee ee ee VI (ie lolol lolol fete) I'o 00° 9's Scr Bhis Dot I'o fofe) Sr lofo) go Corr WiKenceliviisiiehisis}vwiispisileliaitelei(s) le isis peyeouue Lacon at qos L‘o (So) ean | eb ae zo IO 00 00" oSl “Clo Kom Siig Poo ste oo Poa one UWiIa}9p Puodes 9z1UO qT S-o DG _|[Pecvoaws 0:0 I'o fofe} Ke) 2.009 -O0QI pOI XE Pred 0-0 Ot GOI O40 O40 WLI9}9p SIU a} IUO gn fo ot. go C-o +e) OL IL 5 (6) “OG 009 COO) CeO On OO 080. On 0eOh0 0 & ded iGep trad 9U0]SOW'T r Ane) at o-oo g°0 Q 1 VI “be ov QI org OG Oe os PO, I Ch OOS OME res: te Gachitrers 0 IIe So ol gr oe) vo oo 00° Ors v8 COLE ies ser rah aire: p2fooo ATyomb ourz S-o oo roe) £0 vo 00° O-% vee ae CLG aire es eh ae eee aN p2foos ATMo|S dUrzZ 9:0 go gt g'¢ gir 00° S-o0 9:0 "9 obg COO FOOD Goto o-0sd-U 0-5 p2toos Apyomb ULL, (oe) 3:0 SY Sic ol OL ol o€ oa 0 ae | an “Teese e ss *=palooo ATMOS UL, d 2 7] ty ty IY 76) i9) 79) 1) AI 2IAVL 24 A. A. MICHELSON The expression for the viscous term should be S;=(F?)? if the stress (P) is constant. If P is a function of time =([Fdt)> . Thus if P be given a constant value for a time / and then changed to P, the corresponding value of the viscous term would be S3= (Fo lo—F xt) e If the first stress be considerable and act for a long time the effect of the second stress is negligible. Table IV is a provisional table of the constants which spose in the formula for the torsional strain at room temperature. She AG en 2) AAs in which 4 =CPe"”. P is the weight acting on a pulley of radius 5 cm., the unit of weight being 100 gms. and the unit of time one minute. The specimen is a cylindrical rod 7.5 cm. long and 4 mm. in diameter. The term A,, which may be termed the ‘‘lost motion,” should probably be considered as a part of the viscous term, but with a very small exponent 7, so that the whole viscous term may be represented by S;=C,Pé? (?+Br) - THE GREAT GLASS-SPONGE COLONIES OF THE DEVONIAN; THEIR ORIGIN, RISE, AND DISAPPEARANCE JOHN M. CLARKE New York State Museum, Albany A very striking feature of the biota of the Devonian as repre- sented in the state of New York is the extraordinary development in its late stage, the Chemung period, of its silicious hexactinellid sponges. At various levels in the sandy deposits of this time they are found, sometimes as scattered individuals and sometimes in plantations of uncounted numbers, so that it is safe to say that from the bottom of the formation in the “southern tier” of counties to nearits summit, the hexactinellids of this order, the Dictyospongida, are many times more abundantly represented than in all the rest of the world together. In their extensive monograph of these sponges Hall and Clarke ascribe seventy-seven species in sixteen genera to this formation within the borders of New York and the same rocks in northern Pennsylvania. Clarke has described a number of additional Chemung species, so that there are now about ninety outstanding specific designations for this Devonian assemblage. Sometime it will be a subject for discussion among morphologists whether this so-called order, Dictyospongida, is homogeneously constituted; probably it is not, but so seldom is the spicular struc- ture retained in the sandy matrix that on the basis of general form and habit, and the arrangement of the spicular bands which are usually sharply preserved in impressions, all the sponge occurrences in this formation and those of like composition in the Mississippian faunas of Ohio and Indiana are now for convenience put in this single group. That they are for the most part accurately referred to the hexactinellids is abundantly shown by the spicular structure of the Mississippian species which has been demonstrated. The described species and genera have been established with the best 25 26 JOHN M. CLARKE knowledge available; more exact determinations must await better preserved materials. AREA OF OCCURRENCE The area of deposition in which these Devonian hexactinellids are most prolific is quite distinctly limited along the line of outcrop in the region running from Cattaraugus County on the west to Otsego County on the east, about 150 miles; they accompany the area of most typical sandy sediment. As soon as the formation begins to lose in sand in its extension westward and to the east the sponges disappear. NUMBER AND NATURE OF COLONIES While scattered individuals and groups of sponges occur at — random through these rocks and add much to the variety of the fauna, it is the great plantations or colonies that are here the subject of special reference. It is probable that we know as yet but few of the colonies that once existed. Some have been irreparably lost and doubtless others await discovery. But we may here take note of the following: ; 1. The Hamlin Farm Colony, Naples, Ontario County.—This lies nearest the base of the formation and is the oldest of all the colonies known. It appears to have been entirely composed of the species Hydnoceras tuberosum Conrad, of the tuberous or “alli- gator tail’? type. Some hundreds of specimens have been found here. 2. The Brown Hill Colony, near Avoca, Steuben County.—Here again Hydnoceras tuberosum, the type of the genus and the first of all the dictyosponges to be described, is the prevailing if not the exclusive form. Wagonloads of these sponges have been taken from this place. 3. The Jenks Quarry Colony, near Bath, Steuben County.— The sponges here are also of the tuberous type but belong to the species H. bathense, H. & C., with an occasional representative of _ H. botroedema H.& C. This is the largest of all the assemblages. Workmen in the quarry, 30 years ago, found a layer of sandstone with a ‘‘curly grain” running through it that made it unfit for their GLASS-SPONGE COLONIES OF THE DEVONIAN 27 market and it was thrown out on the spoil bank. The discarded blocks came to the attention of the state’s geologists and a carload of the layer was specially quarried for them. One slab of this layer, 8’ 4’, now exposed in the State Museum, carries about 250 sponges lying as they were left, knocked over on their sides by some heavy tide. The carload of sponges contained probably not less than 5,000 individuals. The layer carrying them extended over the full face of the quarry, 120 feet, and indefinitely inward. The census of the colony cannot be estimated except in very large figures of tens of thousands. 4. The Irish Hill Colony, near Bath—This is known only by the multitude of specimens of H. botroedema found loose in the soil at this place. 5. Lhe Halli Colony at Wellsville, Allegany County.—Here the horizon is high in the formation and the species is Thysanodictya Edwin-Halli H., of which several hundred specimens were found by the late E. B. Hall, of Wellsville. PREVIOUS HISTORY OF THE DICTYOSPONGIDA Limiting the term to the characteristic expressions of the Devonian and Mississippian, they have little record of previous history; “there is a single doubtful specimen from the mud beds of the Hamilton shales (Clathrospongia ? hamiltonensis Hall) and some hexactin patches in the black Marcellus shale, D. ? marcellia Clarke. Fragments of like type, but heretofore unrecorded, have been found in the Rochester shale of New York. In this statement we are eliminating from the group the Cyathospongia forms of the Utica shale, the Levis beds of Little Metis (Ordovician), and the extensive assemblage of similar hexactinellids in the Cambrian, especially those found by Walcott but not yet described. It is proper to exclude these even though they may have full ordinal relation with the Devonian species, because of the vast vacant interval of time between the earlier and later records. There were species of these sponges in the upper beds of the Portage group (three of Hydnoceras, one of Prismodictya, one of Dictyospon- gia, and one of Clepsydrospongia), but they must be regarded as 28 JOHN M. CLARKE belonging to the advance guard of the Chemung army, as in various respects the faunal relations of the two are close. HABITAT OF THE DEVONIAN SPONGES These glass sponges obviously grew on sandy bottom at a depth which could not well have been more than one hundred fathoms, and probably not more than fifty. The waters of the epicontinental seas were always shallow; even the clay and lime muds betoken no depth comparable to the deep-sea oozes and blue muds of the present oceans. There is no single exception to this interpretation that could carry any weight in the presence of such overwhelming proof of the adaptation of this great array to conditions of life wholly unlike those under which their successors are living. When the Devonian time was over the simpler and typical expressions of the obconical, prismatic, and nodose sponges disappeared and were replaced by accelerated species of like stock (twenty species are recorded by Hall and Clarke) in the Mississippian stage, the Waverly and Keokuk divisions, in which there is a notable increase of lime sedimentation and consequent evidence of a deepening sea. HABITAT OF LIVING HEXACTINELLIDS Depth habitat.—Generally and specifically, these are deep-sea animals. Agassiz dredged them from 2,410 fathoms, but the “Challenger” expedition, as shown in Schulze’s report, determined a greater range and a greater depth. The “Challenger” garnered about one hundred species, none of which grew at less than ninety- five fathoms. The summary of the record is as follows: ENOL OS Wo) “AC OBNONOINS -4 gp kobseacncenor 24 species AJ) ON 9 SOS MEHU MONS. A npg odo b oe Hcdoe 6 none KOMI) WGkolo) MERION. Lh dn bub godess o46 35 species 7OE oO) “OOOMa THOM Sse eee none Or La Toco tathoms ane em er emer 2 species FOOT £02900 mathoms) 4... eee 47 species Thus all the species are of the deeps and many of the very great depths. GLASS-SPONGE COLONIES OF THE DEVONIAN 29 Ground habitat.—The nature of the ground determined for tor of these species is given by Schulze as follows: Material Number of Species SHUANG!. 5 o wits bisinh ¢ OR OG oh CO CE Eee ee nn eae. 5 GiaveMa i GESLONES Hera felt ee asso lec. BS Sad Mala bone dtoee 2 RATAN SOU memantine itas Seles el lols on tbe eyeisas. heve 6 (COMPEN TUMIUG| oe oh eae ec ee a ie a a Fy) Wiles TOS SRaNG |S hc AA ee ae ae ee oA 14 (CHPBEID TENG! WW 4-As Gos aR oa eS I TRG! PMG! S545 o SIGS Bre ee EE ea 2 Manca nchicme lie mm ud)).5 2... es os os eye ce wiles « 32 TRG! GEST 6 8s 6 Me e.Sie oa eee II GiCIDN ETE, COVADS hn he an an ee Em 13 RECOM OCKOGZE mM tslosae oles as dc ates. s CAs Seiwa ds 7 TRG CIETY CYS YAS 9 40 ea ee 2 HO ECMO AG SONI TNC aa Pes, alc see 5 Seem sic hee aged a Riabele re ele 9 Schulze observes that forms equipped with root tufts were principally found in soft muddy ground, and in the Devonian seas the species generally were provided with a more or less conspicuous tuft of this kind. Temperature habitat—The temperature of the Chemung period was probably pretty cool. Glacial ice had formed over the ele- vated land of the middle Devonian on the Atlantic border of this continent and now with partial resubmergence the refrigerated waters were discharging themselves, with abundant landwash, into the shallow seas. In the Portage division of the late Devonian an immigrating warm water fauna (the Manticoceras intumescens fauna) coming in from the west was blocked and stopped on its way, driven out or destroyed by the presence of the cool waters carrying the Chemung fauna. Today the hexactinellids show an apparently different tempera- ture control, as witness the ‘‘Challenger’s” record: Nonuiehemperate ZONne.. ...:5-.. 4.62 ote 3 y- 20 species TETSOWOS 65, 3.3 shen ee ee ee 45 species Southihemmenrate ZONE. <..20<.0:dea eee vee bee 35 Species This apparent difference is compensated by the cold of the deep waters. 30 JOHN M. CLARKE RELATIVE ABUNDANCE OF LIVING AND DEVONIAN SPECIES Continuing to use the “Challenger” reports we find a contrast in the abundance of individuals growing in any one place, but in making this comparison we must remember that the zodlogist is dipping into the sea bottom with a rake on the end of a string, while the paleontologist is on the sea bottom itself, with dynamite, crowbar, and hammer. The ‘““Challenger’s” dredgings rarely found any considerable number of individuals at any one place; “gen- erally only one or two specimens of each species were obtained at the same locality. Sometimes, however, a considerable number of specimens were found at once.” None of this evidence seems to point toward colonies or plantations in such vast numbers as in the Devonian. Today the strongholds of the hexactinellids are about the Philippines, Little Ki and Kermadoc islands; in the deeps of the southern Indian Ocean between Prince Edward and Crozet islands; and in Atlantic waters, off the Bermudas and St. Thomas. ONTOGENY OF THE DEVONIAN SPONGES AS AN INDEX OF THEIR ADVANCED AGE AND SPECIALIZATION There are four simple types of morphology, contemporaneous and combined, among the Devonian dictyosponges: (1) the smooth obcone, regularly expanding like a cornucopia and gently con- tracting about the open aperture; (2) a six-sided prismatic or banana shape; (3) a subprismatic obcone with successive trans- verse rows of tufted nodes, typically eight toa row; (4) long obcones with con-entric rings, like the horn of an Oryx. These simple expressions have a successive value in ontogeny. The first group constitutes the genus Dictyospongia. The second, in its typical expression, is the genus Prismodictya; but in several species not included in that genus the prismatic phase is superinduced upon and subsequent to the smooth phase. In other genera or species the later growth of the prism may show a tendency to develop nodes at the prism angles. Hydnoceras is the name applied to the typical tuberous or nodose forms and Ceratodictya expresses the annulated phase. These different expressions are, as just observed, essentially successive in the chronologic order of GLASS-SPONGE COLONIES OF THE DEVONIAN 31 ontogeny. In a progressed species the preliminary phases may be reduced by acceleration and even suppression but they are usually determinable; that is, a species of Hydnoceras (cf. H. Walcotti 7eeege Seu Fic. 1 Fic. 4 Fic. 5 Fic. 6 Fic. 7 Fics. 1-3.—The typical or radicle form as expressed by (1) Cyathodictya and (2, 3) Dicyospongia. Fic. 4.—The development of the Prismodictya faces. Fics. 5-7.—The Prismodictya type with inception of Hydnoceras tufts. Clarke) will show over its initial surface, first the smooth and then the prismatic and tufted development, and even in its mature and final expression, decided development of the concentric rings. 32 JOHN M. CLARKE These features of ontogeny are brought out for a number of species in the accompanying figures. Fic. 9 Fic. to Fic. 11 Fic. 12 Fic. 13 Fics. 8-11.—Various expressions of the Hydnoceras tufted type in which the Prismodictya and the earlier Dictyospongia stages are indicated. Fic. 12.—Hydnoceras Walcotti; an extreme expression of this combination, show- ing notable retention of the Dictyospongia stage, union of the prismatic and tufted stages, and the development of the annulated condition. Fic. 13.—Botryodictya, a condition in which the Dictyospongia stage is pro- tracted and abruptly develops into a cup-shaped, tufted, and pouched condition. In many species ontogenetic development is carried into extremes of nodosity and annulation, resulting, especially in later species, in great variability of form even within the confines of the Devonian. GLASS-SPONGE COLONIES OF THE DEVONIAN 33 Enough however has been given to indicate the degree of mor- phologic specialization displayed by these ,hexactinellids in the Fic. 14 Fic. 15 Fic. 16 Fic. 18 Fic. 19 Fic. 14.—The incipient annulated or Ceratodictya stage. Fic. 15.—The annulated stage combined with an extreme expression of the prismatic form (Rhabdosispongia). Fics. 16, 17.—Phases of the Ceratodictya stage, Hydnocerina (16), showing departure from the type. Fics. 18, 19.—Expressions of two of these phases (Rhabdosispongia and Hydno- ceras) from the Frasnian formation of France. Devonian, a condition which must have required ages of time to work out. The Devonian sponge colonies must therefore have had a vast ancestry. 34 JOHN M. CLARKE WHERE DO THESE DEVONIAN SPONGES COME FROM AND WHAT WAS THEIR ANCESTRY ? The answer to the first query may take this form: Between the species in the dark shale of the early Ordovician (Levis shales) and this invasion in the late Devonian, there is but a single recorded - species which would seem safely placed among the Dictyosponges, viz., Dictyospongia danbyi McCoy from the Upper Ludlow (Silu- rian) of Westmoreland. We have referred to a similar occurrence in the Silurian of New York. It is quite possible that these were but derelicts tossed shoreward. The striking hexactinellids described by Dawson from the Levis shales seem to have for the most part the simple obconical foundation with special developments of spicular tufts which indicate that up to this time there had been no wide departure from the simple type of Cyathospongia which is repeated in the genus Dictyospongia. ‘The other differentials of the Dictyo- spongida do not appear." The dark shales in which those early species (Ordovician and Cambrian) are preserved indicate a greater depth of water than do the Devonian colonies. We may therefore think of them as having invaded the deeper epicontinental seas from the much deeper waters of the continental edge at a time when the way was freely open to the margins of the platform. If they were traveling in toward ever-shallowing water there would or should be remains of them in the black shales and the sands of the interval deposits. There are none, and the fact constrains us to think that, instead of traveling into shallow waters, they were moving back to the deeper waters, where, concealed from the accessible rock records, they were working out their evolution. Then some impulse which may not be defined? drove them into the shallow epicontinental waters, «Jt is presumed, but not proven, that these early hexactinellids were Lyssacine, ~ that is, had the parenchymous tissue filled with detached spicules as contrasted to the fused parenchymalia of the Dictyonina. 2 Austin H. Clark, writing of causes of marine migrations, says: “Internal specific pressure due to enormous increase in the number of individuals within a species operates not only to cause a species to colonize bathymetrically undesirable locations or unnatu- rally cold and uncongenial regions such as the polar seas but also to force species into small localized areas.”” (Quoted by Ruedemann, Paleontology of Arrested Evolution, p. 128, 1918.) GLASS-SPONGE COLONIES OF THE DEVONIAN 35 clothed in their new differentials. No shallowing of the sea or positive diastrophy is required for this explanation. By the time the Chemung outburst of species was effective all egress to or access from deep water was shut off. Examination of Schuchert’s paleogeographic maps of this time will bring out this condition clearly. ‘There was no deep-water Devonian in the vicinity at that period; to the east and south lay the Appalachian lands; to the north Laurentia, and on the west a long and, we must say, putative channel reaching in from the Pacific border. Through this channel the sponges may have gone out. We conclude that the long evolution of these sponges from their appearance in the dark Cambrian and Ordovician muds to their immigration in the late Devonian was passed in the deeper waters of the continental edge and is recorded in sediments beyond the present reach of our observation. Barrois discovered the Dictyosponges in the Psammites du Condroz of Jeumont in Brittany in sandy sediment at a horizon equivalent to the Chemung of New York, and four of these species in three genera were described and illustrated by Hall and Clarke (op. cit.). This is interesting collateral evidence of the widespread influence which in the Northern Hemisphere impelled these sponges on to the platform seas. WHY AND WHERE DID THE CHEMUNG DICTYOSPONGES GO? The course of their ontogenetic development shows that their later expression assumed gerontic and adaptive characters in great variety. The stratigraphic record indicates that the sandy bottom on which the New York colonies and their contemporaneous species grew was overwhelmed by incursions of coarse gravel washed in from the rivers of the eastern Appalachian land. ‘These terminated their local existence and the Devonian period as well. ‘Their emi- gration from southern New York was westward and into deeper waters of the Waverly group of western Pennsylvania and Ohio and the Keokuk lime muds of Indiana. In these Mississippian sediments they make their last appearance. But they were on their way down to deeper waters and we find no reason against the assumption that it was the westward course they followed on to 26 JOHN M. CLARKE the epicontinental margin. As Dictyosponges they have never reappeared, nor as Lyssacine hexactinellids. That réle was played. When their successors came back in the Jurassic and Cretaceous times their independent spicules had been fused into continuous networks and, as Dictyonine sponges, they carried on their impor- tant work as reef and rock builders. Seldom, however, did they reproduce the form of their Devonian predecessors; indeed the ancient form is far better revived in the glass sponges of today. That, too, is an interesting illustration of once more passing the same point on the cycle of their development history. SUMMARY 1. In the Cambrian and Ordovician times the Lyssacine hexactinellids grew freely in the black muds of moderately deep -epicontinental waters. 2. They were on their way off the American continental plat- form and down to the marginal seas. ; 3. Here they carried out their evolution during the long Silurian (when a stray species came ashore) and all the early stages of the Devonian. | 4. In the later Devonian they return in great force and with their evolution fully under way, but not to such an extent as to conceal their ontogenetic stages and the radicle expression on which they were based. To this reappearance on the epicontinent they were evidently impelled by some vis a tergo, some compelling external force, probably the invasion of their province by a dominat- ing life-element of another kind. They were caught in a general migration of the time into the shallow and cool waters of the Chemung, and in these waters they perfected their evolution. 5. Out of these shallow waters they were driven by an incursion of fresh waters which flooded the Devonian province with gravel from the eastern lands. : 6. They migrated thence westward into the deeper waters of the Mississippian and from there once more to the circumcontinental seas. 7. In their return to the epicontinents of the Jurassic and Cretaceous their development had advanced to a change of skeletal structure and a wide variation of form. GLASS-SPONGE COLONIES OF THE DEVONIAN oii 8. Their departure from the Mesozoic epicontinents came with the opening of the Tertiary. They have never returned to epi- continental waters. 9. Today their descendants present a wide vertical range in the ocean waters, indeed an extraordinary adaptation through 2,800 fathoms. The deepest water forms are hopelessly isolated—they cannot climb the hill back to the zones of evolution and they will be as they are for future generations of observers. It looks as though all of them were traveling down to the depths; in which case the race will become stabilized and “‘immortal.”’ 10. The history then is one of cycles of migration and develop- ment, of compelling impulses governing the former and probably inducing the latter. A QUANTITATIVE MINERALOGICAL ‘CLASSIFICATION OF IGNEOUS ROCKS—REVISED ALBERT JOHANNSEN University of Chicago PART I In response to a request published in a former paper on a quanti- tative mineralogical classification of igneous rocks, numerous letters have been received from petrographers in this country and abroad. Practically no objections were raised except against the separation of the feldspar molecules. With this separation the writer himself was not fully satisfied, and he reverted, shortly after the paper was published, to the subdivisions used by him when the classification was first presented to his students some nine years ago, namely, that of dividing the feldspars simply into plagioclase, on the one _ hand, and the remaining feldspars, on the other. In the following article this change is shown. A further modification is introduced in Class 4, although no objection was made to it as previously given. The new subdivisions are somewhat simpler than before. An extensive change, embracing the omission of the 72 families of the monzonite series (Fig. 1), was contemplated, and personal letters were sent to a considerable number of petrographers asking opinions. Unfortunately the answers were so far apart that it has seemed best to allow the divisions to remain as they were. The need of more uniformity in classification was brought out clearly by the replies to this one question, as a comparison of the granite quartz-diorite series of several petrographers will show (Fig. 2). Essentially the system now is as follows. For a detailed descrip- tion the reader is referred to the former paper. Classes.—On the basis of the amount of dark minerals (mafites) present, the igneous rocks are divided into four classes, the division points being o—5—50—-95—100 per cent matfites. t Albert Johannsen, ‘“‘Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,” Jour. Geol., XXV (1917), 63-97. In that paper delete eudialyte, p. 90, l. 15, and change Jess to more under Class 1, p. 91, l. 26. 38 CLASSIFICATION OF IGNEOUS ROCKS 39 Orders.—Each of the first three classes is divided into orders (Fig. 3) according to the Ab-An ratio in the plagioclase. The division points are Ab,ooAbo, Aby;An;, Abs.Anso, Ab; Ang;, AboAnyoo. There are thus formed, for each class, four double triangles (Fig. 4), in each of which three angles represent (1) quartz (Qu), (2) all feldspars except plagioclase (Kf), and (3) the feldspathoids (Foids). Adamellite Granite Granodiorite artzZ nZO bog g artz-monzo Le K-Gr. Normai|Gran. Soda-lime-Gr JQu-Monzonite Geanesponiee QuzDior. Fic. 2.—Variations in the usage of names in the granite quartz-diorite series: A, divisions used in the present classification; B, divisions suggested (cf. Fig. 1); C, Lindgren’s original divisions; D and E£, divisions as used by certain other petrographers. The right ordinate is orthoclase, the left, plagioclase. -mo nite G way a ow The remaining angle (Plag) represents albite (Naf), oligoclase to andesine (CaWaf), labradorite to bytownite (NaCaf), or anorthite (Caf), and these constitute the basis for the separation into Orders 2 ANG, A), In Class 4, owing to the absence of light constituents, it is neces- sary to make the subdivisions on a different basis. In this class, 40 . ALBERT JOHANNSEN therefore, each order represents an increased amount of the “‘ores” (Fig. 5). The division points, as in the other case, are O—5—so- As 95-100. Vey 2 ae ee ie EG 6 5 18\_17 1 [19 7g Plag 31 Foids Fic. 3.—Subdivisions of the tetra- Fic. 4.—Family divisions in Classes hedron into orders. I to 3. Families.—Finally, each order is subdivided into families. In Classes 1, 2, and 3, the divisions, as shown in F igure 4, are at o-5—35— 65-95-100 on the feld- spar base line, and at O-5-—50-95—-100 from this line up or down toward quartz or the feldspath- oids. Families o, 1, 6, HI WO Ch AO. Banal air occur only once in each __aictite i i Amphibole 8 class (cf. Fig. 3), since the amount of _plagio- clase in each, whether it be albite, acid plagio- clase, basic plagioclase, or anorthite, is too small to make an essential difference in the rock. These “hinge families” are classed, for convenience, in Order r. Pyroxenes Fic. 5.—Orders and families in Class 4 CLASSIFICATION OF IGNEOUS ROCKS 4I In Class 4 the orders are subdivided asin Figure 5. The families are numbered from 1 to 12, as shown, and are divided at o-5—so— -adamell Quartz Adamellite EES: Monzonite | Monzodiorite \ \Diorite CaNaf Neph.bear'g M/_/nephelite- PulaskitoWeph.bear'g Sy \Ne =e poerine Diorite ephelite-monzoni Fergusite Foids Fic. 6.—Families in Class 2, Order 2 95-100. The three corners, in Orders 1, 2, and 3, represent respec- tively olivine, biotite and (or) amphibole, and the pyroxenes. In Order 4, if thought desirable, the corners may be taken to 42 ALBERT JOHANNSEN represent the various ores; the writer, however, groups the ores in one family, for, considered as rocks, they are unimportant and hardly worth separating. THE MINERAL GROUPS The constituents of the rock are divided into three primary groups: (Qu) Quartz (Kf) | Orthoclase, microcline, microperthite, anorthoclase, etc. (Plag) The whole isomorphous Ab-An series of plagioclases (Foids) The feldspathoids (nephelite, leucite, sodalite, hauynite, noselite, melilite, primary analcite, etc.) QUARFELOIDS MAFITES Dark micas (biotite, phlogopite, etc.) Amphiboles Pyroxenes (including uralitized pyroxene) Olivine Tron ‘“‘ores” (magnetite, ilmenite, chromite, pyrite, hematite, etc.) Cassiterite Garnet Minor } Primary epidote Mafites | Allanite, zircon, rutile, primary titanite, spinel, and other dark minor constituents AUXILIARY CONSTITUENTS The auxiliary constituents are seldom of importance. Topaz Primary scapolite Tourmaline Primary calcite Cordierite Muscovite Corundum Lepidolite Fluorite Zinnwaldite Andalusite Apatite, etc. Most of the auxiliary constituents are light in color; they are, consequently, computed among the leucocrates. SECONDARY CONSTITUENTS Secondary constituents are to be calculated as the originals from which they came. Thus ore replacements of the mafites are com- puted as mafites, kaolin as feldspar, etc., chlorite as a biopyribole, ° analcite as feldspathoid, pseudoleucite as leucite, etc. CLASSIFICATION OF IGNEOUS ROCKS 43 GLASS Glass must be computed from an analysis. One can usually surmise its composition from the character of the phenocrysts and the appearance of the rock as a whole. When undetermined, the rock must be given a tentative name, such as hyaline-rhyolite, etc. RULES FOR COMPUTING ROCKS FROM THEIR MODES 1. The sum of the minerals in the mode should be 1too+o.5. Tf less it should be recalculated to 100. The sum of the leucocrates (quarfeloids plus auxiliary minerals) determines the class. Class 1. Leucocrates form more than gs per cent of the total component. Class 2. Leucocrates between 95 (inclusive) and 50 per cent. Class 3. Leucocrates between 50 (inclusive) and 5 per cent. Class 4. Leucocrates between 5 (inclusive) and o per cent. 2. Determine the orders in Classes 1, 2, and 3 directly from the Ab-An ratio in the plagioclase. Order 1. AbrooAn, to Aby;An;. Order 2. Ab,;An; (inclusive) to Abs:Anso. Order 3. AbscAnso (inclusive) to Ab;Angs. Order 4. Ab;An,; (inclusive) to AbopADioo. In Class 4 the orders are determined by the percentage of “‘ores.”’ Reduce the sum of biotite, olivine, pyribole, and ‘‘ores’’ (including cassiterite, chromite, etc.) to 100, dropping the minor mafites, apatite, garnet, perofskite, any small amount of quarfeloids, etc. The percentage of ‘‘ores’’ determines the order. Order 1. 0 to 5 per cent “ores.” Order 2. 5 (inclusive) to 50 per cent “‘ores.” Order 3. 50 (inclusive) to 95 per cent “ores.” Order 4. 95 (inclusive) to roo per cent “ores.” 3. Determine the family. In Classes 1, 2, and 3, first recalculate the quarfeloids to 100. The amount of quartz (or feldspathoid) thus determined immediately locates a row of five horizontal pigeonholes, in one of which the rock belongs. Recalculate’ Kf plus plagioclase to 100 and determine the proper point on the tIt is immaterial whether the orthoclase-plagioclase ratio is taken from the original values or from those reduced as quarfeloids to 100. The results are naturally the same. 44 ALBERT JOHANNSEN Kf-Plag base line. This determines the vertical series of pigeon- holes, and its intersection with the horizontal series gives the proper position for the family. (If plotted graphically, the family is directly determined by the position of the intersection of three lines, as shown in Example 1, below.) Only when the point falls very close to a division line is it necessary to compute the position accurately. The separation points for Kf-Plag are at o-5—35-65— 95-100. In Class 4, Orders 1, 2, and 3, recalculate the olivine, pyroxene, and biotite plus amphibole to 100 (Fig. 5) and find the proper position graphically, or find the position analytically by taking the ratio of the mineral of one corner to each of the others; thus amphibole to olivine, amphibole to pyroxene, and olivine to pyroxene. The division points are o-5—50—-95—I00. In Class 4, Order 4, the writer groups all the ores in a single family, but classifies the various hematite, ilmenite, magnetite, etc., ores as subfamilies. If desired, they may be further separated. If accessory dark minerals, not used in the computation, are abun-. dant, they determine subfamilies and may be mentioned in the rock name. 4. Subfamilies. In all classes subfamilies are based on o—5—50-— 95—I00 division points, after the manner shown in Figure 1. A FEW POINTS TO BE OBSERVED Any percentage value falling exactly on a line should be moved toward the opposite apex of the triangle except as indicated below. Thus a syenite with 5 per cent quartz is classified with granite, a rock with 95 per cent mafites belongs to Class 4, and one with 95 per cent quarfeloids to Class 2; Ab ;An,; belongs to Order 2, and” Ab;An,; to Order 4. If the divisions fall on the 50-50 line of quartz they are moved upward, or on the 50-50 line of foids downward, toward the apices; that is, they are placed in Families 1 to 5 or 26 to 30. Rocks falling on the 50-50 light-dark line are classed with the dark, while along the plagioclase line Ab;.Ango is classed with basic plagioclase. Rocks falling on the line separating the two triangles, namely, on the feldspar base line, usually should be classed on the quartz side, that is, on the normal side, but if CLASSIFICATION OF IGNEOUS ROCKS 45 the rock has affinities with alkalic rocks it should be placed on the foid side. In Class 4 the medial lines pass through the center of gravity of the triangle, consequently rocks falling on these lines are arbi- trarily moved into the families shown by the small arrows in Figure 5. For classificatory purposes, it is seldom necessary to make exact determinations of the mineral percentages. Unless the proportions are such that the rock falls near a division line, a simple inspection will answer. ‘Thus it is usually possible to determine very quickly whether the dark constituents are less than 5 per cent, between 5 and so per cent, etc.; whether the quartz forms more or ‘less than 5 per cent of the light constituents, and whether the ortho- clase forms more than 95 per cent, between 95 and 65, between 65 and 35, between 35 and 5, or less than 5 per cent of the total feldspar. Of course, if the rocks are to be plotted as points in the triangle, more careful determinations are necessary. EXAMPLES Example 1.—A granodiorite having the composition (ORURIAEA aie le, ek hae ee een ea 18.0 = 23.1 Wrihoclase says ines esac s ss 18.0 = 23.1= 30.0 Andesine (AbjAnj).........- 42.0 = §5.5= FO Total quarfeloids......... 78.0 100.0 100.0 LEV OPER its 6c oe eA 12.8 EVormblendes egret. oiled slew < 9.0 IMGT EIEC ie syc)cte aie wc crs ale os od ANE AN ES) ALS ae out Motalkmantessc t5/.' 60 ess 220 100.0 Percentage quarfeloids=78. ‘The rock belongs to Class 2. Ab,oAn3 falls between 95 and 50. ‘The order, therefore, is 2. The family may be rapidly determined graphically. Plot 23.1 Qu, 23.1 Or, and 53.8 CaNaf by drawing a horizontal line 23.1 above the base of the triangle, and an inclined line, parallel to the Qu-Plag line, 23.1 from that line. The intersection of the 46 ALBERT JOHANNSEN two in Family 9 determines the position of the rock. Asa check, the point must also lie 53.8 from the Kf-Qu line. | To compute the family analytically: From the presence of 23.1 per cent quartz, the family must lie between numbers 6 and io, since there is more than 5 per cent and less than 50 per cent quartz. Further, the ratio of Or to CaNaf is 30 to 70, and since the orthoclase is between 5 and 35 per cent the family number is 9. The rock number, therefore, is 229, that is, Class 2, Order 2, Family 9. (The number is to be read two two nine.) Example 2.—A nephelite-monzonite with FE eS Gis aaa eh cemelg BTS = 39.0 Oligoclase (AbyAng) ......... Bans = 87.0 Motalvfeldspare 2 vo cee. 55-0 100.0 DNF) 0) 0 Page Anes elena hs asa 27.5 Sodal eee ae eaten sem 8.5 Total feldspathoids....... 36.0 Total quarfeloids......... QI.O INGO RAU crocs ciers alielelseaner ets 5.0 BiO te asset ove ieie cone ea erat seats Dal HNC CESS tere? aly Guateia anne aietacaerehe 15 9.0 Quarfeloid ratio 91, Class 2. Ab, Ans, Order 2. Foids to feldspars = 36:55=39.5:60.5. Between peeelice 21 and 25. Kf to CaNaf=39.0:61.0. Family 23. Rock number is 2223. (To be read two two twenty-three.) Example 3.—A lherzolite with VNU a) C cium MURMUR em 20 ea Es A5cOll\ ua Hypersthene 5.0 ln aes "2010) am Beets Olivineg ei. pote eiaee cireener 30.0 = 30.6 Hornblende tee eee 3.0 = 3.0 IMO MetLes enieracraelererucreiete 2). 100-0 CLASSIFICATION OF IGNEOUS ROCKS 47 Since there are neither feldspars, feldspathoids, nor quartz, the rock must belong to Class 4. The ratio of ferromagnesian minerals to ores is 98:2; therefore the order is 1. The ratio of pyroxene to olivine is 65:30=68:32; therefore the rock lies along the line of Families 6, 10, and 11. The ratio of pyroxene to hornblende is 65:3=096:4; therefore it belongs to Family 1, 4, 11, or 12. Family 11 is the only one common to both computations, consequently the rock number is 4111. (To be read four one eleven.) The rock may be plotted by drawing one line parallel to the base and 30.6 above it, and another parallel to the amphibole-olivine line and 66.4 from it. The intersection of the two lines locates the rock. As a check, the third line, parallel to the remaining side of the triangle and at a distance of 3.0 from it, should cross the other two at their intersection. : NAMES PROPOSED FOR VARIOUS FAMILIES On the basis of the foregoing subdivisions, nearly a thousand modal analyses have been plotted and names have been given to many of the families, most of them derived from plutonic rocks falling at the center points. In some cases, as in the quartz-rich types, family names were taken from differentiation rocks. In the tabulation (pp. 12-15) there are many blank pigeonholes, owing to lack of good modal descriptions. There are undoubtedly many rocks in most of the families here left blank, especially in Classes 2 and 3, but the majority of published rock descriptions lack mineral percentages, making them unavailable for classification. The writer is at present engaged in measuring the components of a great number of thin sections, most of them of classic rocks or of rocks which have been chemically analyzed. Blank spaces in the tables below do not necessarily mean that rocks are wanting in these pigeonholes but may indicate that none falls near the center point, although, on the other hand, a solitary rare rock may, in some cases, give its name to the family, even though it is not at the center. 48 ALBERT JOHANNSEN A certain system is used for the prefixes. The terms “granite,” “syenite,” “monzonite,”’ “diorite,” ete., are defined, and the addition of a prefix to any one indicates a definite modification. Where no specific name is available, “leuco-” is used to indicate rocks of Class 1, “‘meso-”’ those of Class 2, and “mela-”’ those of Class 3. In most cases the prefix “meso-” is unnecessary, since normal rocks belong to Class 2, and these are written without the prefix, the class being understood. Thus there are leuco-granites, granites, and mela-granites, respectively, in Classes 1, 2, and 3. Furthermore, syenites, monzonites, granodiorites, diorites, and even gabbros normally contain more than 5 and less than 50 per cent of dark constituents, whereby the prefix “‘meso-”’ is unneces- sary. Analogous rocks in the four orders of each class similarly have distinctive prefixes where no other names are available. The rocks of Order 1 have albite as their plagioclase; therefore an albite- monzonite is a monzonite whose plagioclase is albite, and in Order 4 an anorthite-monzonite is one containing orthoclase and anorthite. An albite-diorite means a rock all of whose plagioclase is albite; an albite-granite, on the other hand, means a granite containing some albite in addition to orthoclase, since granite itself is defined as a rock consisting of quartz, a biopyribole, orthoclase, and less plagio- clase. That is to say, the term “granite”’ in itself conveys the idea of an orthoclase rock with some plagioclase, the latter indicated, except in normal rocks, by the prefix. ‘The plagioclase in Order 2 is oligoclase to andesine, and that of Order 3 labradorite to bytown- ites. Acid and basic cannot be used as prefixes for these orders, since albite and anorthite, the end members of the acid and basic plagioclases, are set apart as Orders 1 and 2. Lime-soda and soda- lime are so much alike that one must always stop to think which is which. The prefixes ‘‘sodi-” and “‘calci-’’ are here suggested. As in the names of normal classes, here also normal rocks drop the prefix; ‘‘sodi-,” therefore, is seldom necessary. To the rocks of the hinge families, namely those which contain no plagioclase, “‘ortho-”’ is prefixed; the feldspar present is orthoclase, microcline, microperthite, or anorthoclase. CLASSIFICATION OF IGNEOUS ROCKS 49 TABLE I CLASS 1. Quarelods between ~°° and 2° Mafites fo) _ Order r Order 2 Order 3 Order 4 AbzooAne to Aby;Ans Aby;Ans to AbsoAnso AbsoAns0 to AbsAngs Ab;Ang; to AboAnyoo OMSTIERIEC Hs et. cisco ates (=110) (=110) (=1TI0) xr Orthotarantulite (=111) (Scere) (=111) 2 Tarantulite Granite-greisen 3 Adamellite-greisen 4 Granodiorite-greisen 5 Tonalite-greisen 6 Orthoalaskite (=116) (=116) | (e116) 7 Alaskite Leucogranite 8 Leuco-albite-adamellite | Leucoadamellite 9 Leuco-albite-granodiorite| Leucogranodiorite Leucogranogabbro to Leuco-albite-tonalite Leucotonalite Quartz-anorthosite rz Orthosite (=1111) (=111T1) (=1111) 12 Leuco-albite-syenite Leucosyenite . 13 Leuco-albite-monzonite Leucomonozonite 14 Leuco-albite- Leucomonzodiorite Leucomonzogabbro Leuco-anorthite- monzodiorite monzogabbro t5 Albitite Leucodiorite Anorthosite Anorthitite ra 16 (=1116) (=1116) (=1116) 17 18 19 20 Dungannonite 2I (=1121) (=1121) (=1121) 22 23 ‘24 Leucolitchfieldite 25 Leucomariupolite 26 =1126) * =1126) (=1126) 27 28 20 30 Craigmontite 31 (=1131) (]rr37) (=1131) 50 ALBERT JOHANNSEN { TABLE II CLASS 2 Quarfeleids between 2° and 52 i Mafites 5 50 Order 1 Order 2 Order 3 Order 4 AbsooAne to Ab,s;Ans Ab,sAn; to AbsoAnso AbsoAnso to AbsAnos Ab;Ang; to AboAnzoo o Meso-silexite (=210) (=210) =(2T0) t Moyite (=211) (=211) (=211) 2 Quartz-granite 3 Quartz-adamellite 4 Quartz-granodiorite ’ a) 5 Rockallite Quartz-tonalite ai 6 Orthogranite (=216) (=216) (=216) 7 Albite-granite pala Granite Calcigranite Anorthite-granite 8 Albite-adamellite Adamellite Calciadamellite Anorthite-adamellite 9 Albite-granodiorite Granodiorite Granogabbro north aonoe been to Albite-tonalite Tonalite Quartz-gabbro Quartz-anorthite- gabbro rr Orthosyenite (=2111) (=2111) (=2111) 12 Albite-syenite Syenite Calcisyenite Anorthite-syenite 13 Albite-monzonite Monzonite Calcimonzonite Anorthite-monzonite 14 Albite-monzodiorite Monzodiorite [ Monzogabbro Anorthite-monzogabbro 15 Albite-diorite Diorite Gabbro, Norite Anorthite-gabbro 16 Pulaskite (=2116) (=2116) (=2116) 17 Nephelite-bearing syenite 18 Nephelite-bearing monzonite 19 Nephelite-bearing monzodiorite 20 Nephelite-bearing diorite 2r Ortho-nephelite-syenite | (=2121) (= 2121) =2T21) 22 Albite-nephelite-syenite ‘| Nephelite-syenite 23 Albite-nephelite- Nephelite-monzonite Kulaite monzonite 24 Litchfieldite Nephelite-monzodiorite | Nephelite-monzogabbro 25 Mariupolite Nephelite-diorite Nephelite-gabbro 26 Naujaite (=2126) (=2126) (=2126) 27, Beloeilite Heronite 28 20 30 Toryhillite Lugarite 31 Urtite, Fergusite, (=2131) (=2131) = 2131) Uncompahgrite CLASSIFICATION OF IGNEOUS ROCKS Si TABLE III CLASS 3. Quaricleids between 2° and > ; Mafites 50 Order r Order 2 Order 3 Order 4 AbrooAne to AbysAns AbysAns to AbsoAnso AbsoAnso to AbsAngs AbsAngs to AboAnsoo ° (=310) (+310) (=310) I (=311) (=311) (=311) 2 3 4 5 6 Mela-orthogranite (=316) (=316) (=316) 7 Mela-albite-granite Melagranite Mela-calcigranite 8 Mela-albite-adamellite | Mela-adamellite 9 Mela-albite-granodiorite | Melagranodiorite Melagranogabbro 10 Mela-albite-tonalite Melatonalite Mela-quartz-gabbro ir Mela-orthosyenite (Senn) (che) (=3111) 12 Mela-albite-syenite Melasyenite 13 Mela-albite-monzonite | Melamonzonite 14 Mela-albite-monzodiorite] Melamonzodiorite Melamonzogabbro Ricolettaite as 15 Mela-albite-diorite’ Meladiorite Melagabbro Yamaskite 16 Orthoshonkinite | (=3116) (=3116) (=3116) 17 Shonkinite Oligoclase-(andesine-) | Labradorite-(bytown- shonkinite ite-)shonkinite 18 19 20, 21 Nephelite-shonkinite (=3121) (=3121) (+3121) 22 23 24 Melalitchfieldite Mela-nephelite- monzogabbro 25 Melamariupolite Theralite 26 (=3126) (=3126) (=3126) 27 28 20 30 31 Bekinkinite, Missourite, | (=3131) (=3131) (=3131) Farrisite 52 ALBERT JOHANNSEN TABLE IV uarfeloids fo) CLASS 4. winaielods between os and — Mafites 05 100 Order 1 Order 2 Order 3 Order 4 **Ores”’ less than 5 Per Cent “Ores”? between “Ores”’ between “Ores”? more than 5 and 50 Per Cent 50 and 95 Per Cent 95 Per Cent Chromite-dunite Olivine-chromitite Chromitite xr Dunite Magnetite-dunite Olivine-magnetite Magnetite 2 Mica-peridotite, amphibole-peridotite, hornblende-picrite, cortlandtite. 3 Valbellite, hornblende-diallage-peridotite, etc. 4 Lherzolite, diallage-peridotite, wehrlite, harzburgite, saxonite. 5 Included with Family 2 at present. 6 Included with Family 3 at present. 47 Cromaltite, hornblende-hypersthenite, etc. 8 Amphibolites, hornblendites. 9 Included with Family 7 at present. to Included with Family 3 at present. rr Included with Family 4 at present. 12 Diallagite, hypersthenite, websterite, ilmenite-enstatitite, etc. CLASS I, ORDER I (110) Silexite MILLER. The term silexite was proposed by Miller’ for pure igneous quartz rocks. Such rocks, frequently described under the names igneous quartz, quartz dikes, quartz veins, etc., represent the end members of pegmatitic intrusions. Greisen, an old Saxon miner’s term, cannot be used for the family name, since it should properly be restricted to tin-bearing leuco- granites. Furthermore, according to some authors, it is an altered rock, feldspar having been changed to quartz. Beresite ROSE? likewise cannot be used; it was shown by Helmhacker’ to be an altered quartz-porphyry. Besides pure quartz there are a number of other rocks falling in this family, namely such as are free from feldspars and mafites, and t William J. Miller, ‘‘Silexite, a New Rock Name,” Science, XLIX (1919), 149; also “‘Pegmatite, Silexite, and Aplite of Northern New York,” Jour. Geol., XXVII ' (1919), 28-54, in particular p. 30. 2 Gustav Rose, Mineralogisch-geognostische Reise nach dem Ural, dem Altai und dem Kaspischen Meere (Berlin, 1837), I, 186. 3 R. Helmhacker, “Der Goldbergbau der Umgebung von Berézovsk am éstlichen Abhange des Urals,” Berg- und Hiittenmannische Zeitung, LI (1892), 83-84. CLASSIFICATION OF IGNEOUS ROCKS 53 consist only of quartz and an auxiliary mineral, such as mica, tourmaline, topaz, etc. Among these are: Esmeraldite Spurr. The terms greisen and beresite have been used more or less loosely for quartz-muscovite rocks, although, as mentioned above, they have other meanings. Spurr' definitely applied the term esmeraldite to rocks from the southern Klondike district, Esmeralda County, Nevada, which consist only of quartz and muscovite. Syn.: Greisen (in part), Glimmer-greisen JOKELy.? Tourmalite.s This term is here suggested for rocks consisting only of tourmaline and quartz, to which the name hydrotourmalite was given by Daubrée.* Other synonyms are Schérlquarzit, Schérlfels, Schorlschiefer, Turmalinschiefer, Turma- linfels, schorl-rock, Carvoeira,' etc. Topazite. Topazite is here suggested for rocks con- taining only quartz and topaz. Syn.: Topasfels WERNER,° Topazogéne CHARPENTIER,’ topaz-rock. (111) Orthotarantulite. The prefix ‘“‘ortho-” is used here and in all of the hinge families to indicate feldspathic rocks which con- tain less than 5 per cent plagioclase. They include, therefore, orthoclase, microcline, microperthite, and anorthoclase rocks. See note under (112). Thus defined, orthotarantulite is a tarantulite with less than 5 per cent of its feldspar plagioclase. tJ. E. Spurr, ‘‘The Southern Klondike District, Esmeralda County, Nevada,” Econ. Geol., I (1906), 382. 2 Johann Jokély, “‘Das Erzgebirge im Leitmeritzer Kreise in Bohmen,” Jahrb. d. k. k. geol. Reichsanst., UX (1858), 566. 3 In the following pages proposed new names are in bold-face type. Where the prefix indicates a newly formed group it is shown thus, orthogranite. 4M. Daubrée, ‘‘Sur le gisement, la constitution et l’origine des amas de minerai d’étain,” Ann. d. Mines, XX (1841), 84. 5 W. L. v. Eschwege, Beitrage z. Gebirgskunde Brasiliens (1832), p. 178. 6A, G. Werner, Kurze Klassification und Beschreibung der verschiedener Gebirgs- arten (Dresden, 1787), p. 15. 7J. de Charpentier, Vom Schneckenstein, oder dem sdchsischen Topasfelsen (Prag, 1776). 54 ALBERT JOHANNSEN Syn.: Alaskite-quartz Spurr (in part),' Feldspar-greisen JoKELy (in part). (112) Tarantulite. Spurr? used the term alaskite-quartz for a transition rock between alaskite and igneous quartz which occurs in the Missouri mine, near Tarantula Spring, Nevada. Asasimpler term the name tarantulite is here proposed for Family 2, which contains orthoclase and less albite. Syn.: Alaskite-quartz Spurr (in part), Feldspar-greisen JoxKELy (in part). (116) Orthoalaskite. The term alaskite was given by Spurr? to leucocratic rocks which contain alkali feldspars (orthoclase, microcline, microperthite), with or without albite. Here they are divided into two groups. ‘Those without albite are included under the name orthoalaskite (see note under (111) ), while those with albite are normal alaskites of Family 7 (117). Runite PINKERTON. Syn.: Pegmatite Hativ,5 Schrift- granit, Hebraischerstein, graphic-granite, etc. Orthotordrillite. The extrusive equivalent of ortho- alaskite is orthotordrillite. The name tordrillite was given by Spurr® to the extrusive equivalent of alaskite. In the present classification only the tordrillites with no plagioclase are included in this family. (117) Alaskite SpurR. See under orthoalaskite (116). Tordrillite Spurr. See under orthotordrillite (116) above. Syn.: Rhyalaskite.? tj. E. Spurr, “‘Ore Deposits of the Silver Peak Quadrangle, Nevada,” U.S. Geol. Surv., Prof. Paper 55 (1906), p. 61. 2 Ibid. 3J. E. Spurr, “Classification of Igneous Rocks According to Composition,” Amer. Geol., XXV (1900), 229-30. 4J. Pinkerton, Petrology (London, 1811), II, 85. 5 Ascribed to Haiiy by Brongniart, ‘‘Essai d’une classification minéralogique des roches mélangées,”’ Jour. d. Mines, XXXIV (1813), 32. 6J. E. Spurr, ‘Classification of Igneous Rocks According to Composition, Amer. Geol., XXV (1900), 230; also “Reconnaissance in Southwestern Alaska in 1898,” U.S. Geol. Surv., Ann. Rept., XX, Part VII (1900), p. 189. 7 J. H. Farrell, Field Geology (New York, 1912), p. 160, Table II. ”? _ CLASSIFICATION OF IGNEOUS ROCKS 55 (118) Leuco-albite-adamellite. No confusion can result here from the use of the word albite as a prefix (see p. 48, above). It cannot be thought to mean that the albite replaces (proxies) the potash feldspar, since the term adamellite conveys the quartz- monzonitic idea, but it clearly shows that the plagioclase present is albite. For the use of the term adamellite for quartz-monzonite see note under (228). Leuco-albite-dellenite. The extrusive equivalent of the preceding. (119) Leuco-albite-granodiorite. See note under (118). (1110) Leuco-albite-tonalite. See note under (118). For the use of tonalite for quartz-diorite see (2210). : (1111) Orthosite TuRNER. These rocks are leuco-potash- syenites or leuco-orthosyenites. For those composed entirely of orthoclase, Turner? proposed the term orthosite. Pure anortho- clase or sanidine rocks belong here also. The former might properly be called anorthosites. This term, however, is now so firmly attached to plagioclase rocks without mafites that no change is possible. See note under anorthosite (1315). Sanidinite TsCHERMAK.? Microclinite LoEwINson-LESSING.? Anorthoclasite LoEwINson-LEssING.4 (1112) Leuco-albite-syenite. A temporary term for leucocratic syenites, consisting essentially of orthoclase with some albite. See note under (217). (1115) Albitite TURNER. Turner’s® term for rocks which con- sist essentially of albite. (1124) Leucolitchfieldite. See note under (2124). This term is preferable to the longer term leuco-albite-nephelite-monzodiorite. tH. W. Turner, ‘‘The Nomenclature of Feldspathic Granolites,” Jour. Geol., VIII (1900), 106-10. 2 Gustav Tschermak, ‘“‘ Verhandlungen der k. k. geol. Reichs. Sitzung am 6 Marz, 1866,” Jahrb. d. k. k. geol. Reichsanst., XVI (1866), 34. 3 F. Loewinson-Lessing, ‘‘ Kritische Beitrage zur Systematik der Eruptivgesteine,” Tscherm. Min. Petr. Mitth., XX (1901), 114. 4 Loc. cit. SH. W. Turner, “The Nomenclature of Feldspathic Granolites,” Jour. Geol., VIII (1900), 111. 56 ALBERT JOHANNSEN (1125) | Leucomariupolite. See note under (2125). It is a leuco-albite-nephelite-diorite. (1131) Here belong rocks consisting of one or more feldspath- oids, without feldspars or dark constituents. ‘The type rock is pure nephelite. For the corresponding extrusives Loewinson- Lessing proposed noseanite, nephelinolith, and amphigenite. Noseanite, however, was previously used by Boricky? for an amphibole-(plus 50 per cent) nephelite-(20o-40 per cent) noselite rock with small amounts of magnetite and olivine, while amphi- genite was used by Cordier for rocks now called leucite-tephrites. Sodalitsten STEENSTRUP‘ probably belongs to this family. CLASS I, ORDER 2 (120) These rocks are included under Order 1, since the varia- tions produced by small amounts of different plagioclases are un- essential. : (121) Included under Order 1. (122) Granite-greisen JOKELY. Syn.: Feldspathgreisen JOKELY. Jokély applied the term Granit- or Feldspathgreisen to rocks con- sisting essentially of quartz and feldspar with some muscovite. The feldspar was determined megascopically only and was spoken of as ‘‘allem Anschein nach durchgénglich Orthoklas.” If ordinary granites are divided into orthogranites and normal granites, so also should the granite-greisens be divided, and there would be orthogranite-greisens (121) and granite-greisens (122). But (121) is the hinge family and is the same as (111); consequently the former name need not be considered, and the term granite-greisen can be applied to the quartz-feldspar rock of Family 2. The presence or absence of muscovite will not change the classification, 1 F, Loewinson-Lessing, ‘‘Kritische Beitrige zur Systematik der Eruptivgesteine, IV,” Tscherm. Min. Petr. Mitth., XX (1901), 114. 2 Emanuel Boricky, ‘‘ Petrographische Studien an den Basaltgesteinen Bohmens,”’ Arch. f. d. naturw. Landesdurchf. v. Bihmen, Band II, Abt. ii, Th. ii, pp. 41, 78-79. 3 Cordier and d’Orbigny, Description des roches (Paris, 1868), pp. 114-15. 4N. V. Ussing, ‘‘Mineralogisk-petrografiske Underségelser af Grénlandske Nefelinsyeniter og beslaegtede Bjaergarten, 1894,’ Meddel. om Gronl., XIV (1898), 128. 5 Johann Jokély, ‘‘Das Erzgebirge im Leitmeritzer Kreise in Bohmen,” Jahrb. d. k. k. geol. Reichsanst., IX (1858), 567. CLASSIFICATION OF IGNEOUS ROCKS 57 though when it is present the rock may be classed as a sub- family, comparable to one in the normal granite family, namely as Muscovite-granite-greisen. (123) Adamellite-greisen. The rock for which the term adamellite-greisen is here proposed is related to the quartz- monzonites (adamellites) as granite-greisen is to normal granite. For the use of the term adamellite for quartz-monzonite see note under (228). (124) Granodiorite-greisen. See note under (123). (125) Tonalite-greisen. See note under (123). (127) Leucogranite differs from normal granite (227) in the practical absence of mafic minerals. It therefore consists of quartz, orthoclase, and a less amount of oligoclase or andesine. Leucorhyolite. The extrusive equivalent of the preceding. (128) Leucoadamellite. See note under (127). Leucodellenite. The extrusive equivalent of the pre- ceding. (120) Leucogranodiorite. See note under (127). Leucorhyodacite. The extrusive equivalent of preced- ing. See note under rhyodacite (2209). (1210) Leucotonalite. See note under (127) for the relation- ship between this rock and normal tonalite. For the use of tonalite for quartz-diorite see (2210). According to the kind of feldspar present, the leucotonalites are divided into quartz-oligosites TURNER (quartz-oligoclasites), and quartz-andesinites TURNER. See note under (1215). Quartz-oligosite TURNER. Quartz-andesinite TURNER. Leucodacite. The extrusive equivalent of leucotonalite. (1212) Leucosyenite. See note under (127). Leucotrachyte. The extrusive equivalent of preceding. (1213) Leucomonzonite. See note under (127). Leucolatite. The extrusive equivalent of preceding. (1214) Leucomonzodiorite. In the former paper the writer" suggested the term syenodiorite, from analogy with granodiorite, t Albert Johannsen, ‘Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,” Jour. Geol., XXV (1917), 89. 58 . ALBERT JOHANNSEN for quartz-free plagioclase rocks with some orthoclase. Under (229) he shows the objection to the term granodiorite, an objection which also applies to the word syenodiorite. The latter term, conse- quently, is here withdrawn, and monzodiorite is substituted as better indicative of a rock intermediate between monzonite and _. diorite. Leucomonzodiorite is the name of the corresponding rock in Class 1. Leucoandelatite. See note under (2214). The extru- sive equivalent of preceding. Leucotrachyandesite would be the extrusive name by analogy with granodiorite, but trachyandesite has been used in the sense of latite as well as for an intermediate rock of the foyaite series* _and comparable to trachydolerite of the gabbro series. (1215) Leucodiorite. According to the kind of plagioclase present, the leucodiorites are divided into oligosites TURNER? (oligoclasites) and andesinites TURNER.3 Oligosite TURNER. Andesinite TURNER. Leucoandesite. The extrusive equivalent of leuco- diorite. (1220) Dungannonite ADAMS and Bartow. A _ leucocratic nephelite-bearing diorite with considerable corundum (13.24 per cent) from Dungannon, Ontario, was described and named dun- gannonite by Adams and Barlow.‘ It is not typical of the family on account of the presence of corundum, but the name is here used since this rock is the only apa of the family yet located in the literature. (1230) Craigmontite ADAMS and Bartow. The name craig- montite was given by Adams and Barlow’ to a nephelite-oligoclase- muscovite rock from Craigmont, Ontario. While the mode given tH. Rosenbusch, Mzkroskopische Physiographie der massigen Gesteine (4th ed.; Stuttgart, 1908), p. 1036. 2H. W. Turner, ““The Nomenclature of Feldspathic Granolites,” Jour. Geol., VIII (1900), 111. 3 Ibid. 4 Frank D. Adams and Alfred E. Barlow, ‘‘Geology of the Haliburton and Ban- croft Areas, Province of Ontario,” Geol. Surv. Canada, Mem. 6 (Ottawa, 1910), p. 322. 5 Ibid., p. 313. CLASSIFICATION OF IGNEOUS ROCKS 59 in the report of the Canadian Geological Survey is calculated from the analysis, the actual mineral composition is probably also repre- sented by it. The craigmontite type contains muscovite, but the name may be applied, with a proper prefix, to any oligoclase- nephelite rock of Family 30, for example aegirite-craigmontite, etc. CLASS I,*ORDER: 3 (139) Leucogranogabbro. While adam-gabbro would be more correct than granogabbro for the rocks of this family, the term is objectionable in sound, and, furthermore, since granodiorite is so firmly established that it must be retained, granogabbro as an analogous term should also be HORNE, See note under grano- diorite (220). Leucorhyobasalt. For the same reason that rhyodacite is retained rhyobasalt is used. See note under (229). Strictly speaking the term should be leuco-rhyo-quartz-basalt, but the prefix “‘rhyo-” may be considered as indicative of the presence of quartz. (1310) Quartz-anorthosite. See note under (1315). Leuco-quartz-basalt. ‘The extrusive of the above. (1314) Leucomonzogabbro. The objection to syenodiorite, mentioned under (1214), applies also to syenogabbro, which was previously proposed by the writer." That term is now withdrawn and monzogabbro is substituted. See note under (2314). (1315) | Anorthosite Hunt. The term anorthosite was pro- posed by T. Sterry Hunt? for rocks composed chiefly of plagioclase (labradorite in most Canadian occurrences). The name is derived? from anorthose, originally used by Delesse for triclinic feldspars, although now used for anorthoclase. Anorthosite, consequently, properly should not be applied to plagioclase rocks. It is in such general use, however, that it must either be dropped entirely _ or else used in the sense of Hunt. Turner would apply the term t Albert Johannsen, “‘Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,”’ Jour. Geol., XXV (1917), 80. 2T. Sterry Hunt, Geology of Canada (Montreal, 1863), p. 22. 3Frank D. Adams, “Uber das Norian oder Ober-Laurentian von Canada,” Neues Jahrb., B.B., VIIL (1893), 423- 4H. W. Turner, ““The Nomenclature of Feldspathic Granolites,” Jour. Geol., VIII (1900), 106-11. 60 ALBERT JOHANNSEN to anorthoclase rocks without mafic minerals, and would divide the basic plagioclase rocks, now called anorthosites, into labradites and anorthitites, according to the kind of plagioclase present, pre- sumably dividing the bytownite rocks between them. Since most anorthosites are labradorite rocks, the term may well be confined to the labradorite-bytownite rocks of Class 1, Order 3, leaving the anorthite rock, anorthitite, in a class by itself (1415). Labradite TURNER. Bytownitite. Leucobasalt. The extrusive equivalent of anorthosite. CLASS I, ORDER 4 (1414) Leuco-anorthite-monzogabbro. A simpler name should be used here when a rock near the center point is described. (1415) Anorthitite TURNER.’ See note under (1315). 1 Ibid. ‘ [To be continued] THE PRE-MOENKOPI (PRE-PERMIAN ?) UNCON- FORMITY OF THE COLORADO PLATEAU C. L. DAKE Missouri School of Mines \ There is a growing appreciation of the importance of the uncon- formity below the “‘Red Bed” series of the Colorado Plateau Province, as witnessed by the recent paper of Leet on the subject. During a reconnaissance in southeastern Utah, northeastern Arizona, and northwestern New Mexico, the writer had an oppor- tunity to gather some additional data on the extent and possible magnitude of the stratigraphic break, and these facts are herewith presented. The first part of the paper deals with the local evidences of the unconformity at several points, particularly at Mule Twist Canyon, Utah; at Fruita, Utah; in Quartzite Canyon, near Fort Defiance, Arizona; at Ramah, New Mexico, on the Zuni Uplift; and near Tolchico, Arizona, on the Little Colorado River. The location of all these points is indicated on the accompanying sketch map (Fig. 1). The second part of the paper begins with a summary of the known areal extent of the unconformity, and this is followed by a discussion of the magnitude of the stratigraphic break. Since the formation names are largely local and probably unfamiliar to most readers, the following partial columnar section is given. Jurassic, La Plata sandstone Dolores (Chinle) formation Triassic ‘ ‘ Shinarump conglomerate DeChelly sandstone (local) Permian (?), Moenkopi red beds = W. T. Lee, ‘‘General Stratigraphic Break between Pennsylvanian and Permian in Western America,” Bull. Geol. Soc. Am., Vol. XXVIII (1917), pp. 160-70. 61 62 C. L. DAKE Kaibab limestone Coconino sandstone ; Upper Aubrey group, Pennsylvanian Supai formation Redwall limestone, Pennslvanian and Mississippian Carboniferous, DETAILED OCCURRENCES OF THE UNCONFORMITY Mule Twist Canyon.—The first locality at which the uncon- -formity was noted is about four miles northwest of Mule Twist oo S%, $ JS, ; On Tolchico nx, ‘a Fic. 1.—The Colorado Plateau Canyon, a well-known pass through the La Plata “‘reef”’ or “‘ledge,”’ a prominent escarpment on, the east flank of the Water Pocket Flexure, west of Henry Mountains, Utah (Fig. 1). At this point the unconformity is distinctly indicated by basal conglomerates, by the uneven surface on which the Moenkopi rests, and by the PRE-MOENKOPI UNCONFORMITY OF COLORADO PLATEAU 63 variable thickness of the Moenkopi itself. The following detailed section was measured in this vicinity: SECTION NEAR MULE TWIST CANYON (4) 100 feet red and yellowish-gray sandy shale and some sandstone (3) 60 feet well-bedded gray fossiliferous limestone (2) 40 feet regularly bedded calcareous sandstone (1) 75 feet massive cross-bedded gray sandstone, base not exposed The gray standstone and limestone are undoubtedly the Aubrey group of the Carboniferous mentioned by Gilbertt in the Henry Fic. 2.—Cross-bedded Coconino(?) sandstone near Mule Twist Canyon, Water Pocket Flexure, near Henry Mountains, Utah. Mountains. The limestones and upper bedded sandstones (mem- bers 2 and 3) were sparingly fossiliferous, and several species were collected, but the collection was unfortunately lost by the burning of an office building in which the writer was temporarily quartered. The fauna was totally unfamiliar to the writer, and consisted of very small crinoid stems, several pelecypods, and one or two small brachiopods. They did not resemble at all the faunas so abundant in the Goodridge formation of the San Juan Oil Field, a formation ™G. K. Gilbert, The Geology of the Henry Mountains, 1880. 64 CTD AGE provisionally correlated with the Redwall limestone. The cross- bedded sandstone (member 1) resembles very closely the Coconino. (Cf. Figs. 2 and 3 of this paper with Plates XXIX A and B, U.S. Geol. Surv., Bull. 613, which illustrate the Coconino in Wal- nut Canyon, Arizona.) The red and yellowish-gray sandy shales of number four rest on the slightly eroded surface of the limestone (member 3) with a very sparing development of conglomerate at the contact. The shales are gray at the base, but grade upward Fic. 3.—Cross-bedding in Coconino(?) sandstone, same locality as Fig. 2. Photo by Zoller. irregularly into the typical red sandy shales of the Moenkopi, which are in turn overlain, again with unconformity (Fig. 4), by the coarse sandstone and conglomerate of the Shinarump (restricted), and this is followed by typical Dolores and La Plata. ) At this point the unconformity is not very pronounced. About two miles northwest, however, near the south end of Wagon Box Mesa, a large mesa capped with Shinarump, the deeper valleys show red shale grading down through gray sandy shale into coarse sandy conglomerate. The conglomerate consists of pebbles of chert and limestone in a sand matrix, and rests directly on gray, cross-bedded sandstone. At this point the limestone seems to have been removed by the pre-Moenkopierosion. Here, at the south end of the Wagon PRE-MOENKOPI UNCONFORMITY OF COLORADO PLATEAU 65 ~ Box Mesa, the Moenkopi is about four hundred to five hundred feet thick. Two miles north of the north end of the mesa, perhaps four miles from the last point described, occur some smaller mesas, also capped with Shinarump. Here the Moenkopi is not over two hundred feet thick. The change may in part be due to the post-Moenkopi, pre-Shinarump erosion, but is more probably owing to the uneven surface of the Aubrey beds, on which the Moenkopi was laid down. This seems the more probable, since Fic. 4.—Unconformity of Shinarump conglomerate on Moenkopi beds, west of Wagon Box Mesa, Water Pocket Flexure, Utah. at this point the Moenkopi is again noted to rest on the limestone member. In this vicinity the upper ledge of the limestone is itself distinctly conglomeratic. The foregoing facts indicate a decided erosion of the Aubrey before the Moenkopi was deposited, an erosion amounting in places to probably two hundred feet or more, with the complete removal, at certain points, of the limestone beds, allowing the shales to rest directly on the cross-bedded sandstone member. Fruita.—At Fruita, about forty miles northwest of the Mule Twist locality, the canyon of the Dirty Devil (Fremont) River cuts deeply into gray limestones and sandstones below the “ Red 66 CSL DAKE Beds.”’ A rather detailed section was measured along this canyon, and the section is given below. SECTION IN DIRTY DEVIL CANYON, NEAR FRUITA 26 50-+feet red shale, Moenkopi, slightly irregular contact, with a few pebbles 25 50 feet limestone in massive ledge 24 55 feet. red shale 23 1o feet covered slope 22 60 feet limestone, chalky at base, sandy in middle, cherty at top 21 Teer covered slope 20 40 feet white chalky limestone, with chert concretions 19 20 feet covered 18 20 feet gray limestone, weathers with a powdery surface 17 20 feet covered 16 DZ Nea chert 15 170 feet non-bedded sandstone, very coarse at base, finer above 14 30 feet — sandy limestone, with quartzite ledges 13 13 feet yellowish, very cross-bedded sandstone 12 4 feet sandy gray limestone II 12 feet cross-bedded coarse white sandstone 10 15 feet sandy gray limestone 9 10 feet coarse gray sandstone 8 18 feet thin-bedded argillaceous limestone 7 2 2eeeet cross-bedded coarse white sandstone 6 125 feet cross-bedded white sandstone with some thin shale lenses = 2 feet thin-bedded shaly sandstone 4 AGuateer cross-bedded white sandstone 3 2-6 inches greenish-gray shale 2 40 feet massive cross-bedded coarse white sandstone I 12 feet covered slope above river ——_—_ 957 feet, all of which, except number 26, are Aubrey. There is nothing in the section above that can be correlated with any great certainty with the individual beds in the vicinity of Mule Twist Canyon. Member number 15 may possibly repre- sent the sandstones of the Mule Twist region, though this is uncertain. If it does, there are many more beds between it and the Moenkopi here than in the former section. The unconformity PRE-MOENKOPI UNCONFORMITY OF COLORADO PLATEAU 67 itself is not particularly well marked in the vicinity of Fruita, but the entirely different character of the beds on which the Moenkopi rests here and near Mule Twist constitutes a difference which perhaps might result from lateral variation, but which more probably indicates a different horizon as the base on which the Moenkopi beds were deposited. No fossils were seen in the Fruita section. About ten miles northwest of Fruita, and perhaps three miles southeast of Torrey, the following section was measured across the contact: SECTION NEAR TORREY, UTAH 16 Bepteee red sandy shale 15 15 feet gray shale 14 75 feet red sandy shale 13 20 feet gray to brown, thin-bedded sandstone 12 35 feet red sandy shale, ripple-marked II Spee covered slope ite) ireet thin-bedded sandy gray limestone, upper contact con- cealed 2 feet very crystalline, pitted gray limestone 16 feet thin-bedded argillaceous limestone Fceta covered 15 feet crystalline gray pitted limestone feet argillaceous white sandstone 3 feet gray crystalline limestone 3 feet fine-grained argillaceous white sandstone G2reieet massive gray limestone 42+feet red shale with gypsum veins, base not exposed Hnbwtu an Cw Os The boundary of the Pennsylvanian and the Moenkopi is here believed to occur between members to and 11. The red shale (number 1) of this section, below the fifty-two-foot ledge of gray limestone, is believed to correspond to the fifty-five feet of red shale (number 24) in the Fruita section, also below fifty feet of massive limestone. Beds 3 to 10 of the Torrey section seem to have been eroded off at Fruita. Ramah.—On the west flank of the Zuni Uplift, about six miles east of Ramah, New Mexico, along a sharp canyon, followed by the main wagon road, the situation is as follows: about fifty feet of red sandy shales and sandstones rest with very uneven contact 68 CoE =DAKE: on gray massive limestones, of which perhaps fifty to one hundred feet are exposed. No fossils were noted, but it is confidently believed that the gray limestone is Aubrey and the red sandy shales and sandstones Moenkopi. The marked uneven character of the contact leaves no room for doubt as to the unconformable relation between them. . Above the Moenkopi is a thin conglomeratic sandstone taken to be the Shinarump, since above it rest characteristic ashy gray and purple shales highly suggestive of the typical Chinle (Dolores), E : | Fic. 5.—Nearly flat-lying Moenkopi unconformable on folded pre-Cambrian, Quartzite Canyon, near Fort Defiance, Arizona. so widely exposed in the De Chelly (Fort Defiance) Uplift: The exceptional thinness of the Moenkopi here (so to roo feet) may be due in part to post-Aubrey and pre-Moenkopi erosion, in part to post-Moenkopi and pre-Shinarump erosion, and possibly in part to lack of deposition. No information was secured which would enable one to decide which of these might be the most important factor. Fort Defiance—In Quartzite Canyon, near Fort Defiance, Arizona, the Moenkopi rests directly on steeply dipping, much jointed, vitreous quartzite (Figs. 5 and 6). ‘This relation has tH. E. Gregory, “‘Geology of the Navajo Country,” U.S. Geol. Surv., Prof. Paper 93, 1917. 2h PRE-MOENKOPI UNCONFORMITY OF COLORADO PLATEAU 69 already been described by Gregory,’ who says of it: “If strata of Pennsylvanian age once covered the quartzite the pre-Moenkopi erosion interval, elsewhere poorly marked, becomes here a strati- graphic feature of great significance. In my opinion the quartzite is a monadnock of a pre-Cambrian erosion surface—an elevated mass which outlived its contemporaries through Cambrian, Silurian, - Fic. 6.—Same unconformity as in Fig. 5 Devonian, and early Carboniferous time, only to be itself buried by the streams of Permian time.’’ It would appear from this state- ment that Gregory does not incline to believe that the quartzite- Moenkopi unconformity is related to the Pennsylvanian-Permian erosion interval, but represents earlier unconformity with overlap. Here also it is not possible to determine the facts. tH. E. Gregory, loc. cit. 70 C. L. DAKE Tolchico.—Gregory* sums up briefly evidence presented by several writers, pointing to unconformity at the base of the Moen- kopi, which seems to be especially marked along the Little Colorado. The writer has seen the area described by Gregory near Tolchico, and while there is clear evidence both of pre-Moenkopi erosion channels and basal Moenkopi conglomerate, neither the eroded contact nor the conglomerate are as well developed here as near Mule Twist Canyon, west of Henry Mountains, unless the condi- tions described below are related to pre-Moenkopi erosion. At a sharp bend in the Little Colorado River, perhaps two or three miles northwest of Tolchico, a deep sharp canyon is cut in the Kaibab limestone, and a sharply cut tributary canyon enters at the bend, from the southwest. Perhaps a quarter of a mile up the tributary canyon from its mouth occur conditions of peculiar interest. Within the canyon, which here cuts between seventy- five and one hundred feet into the Kaibab, and resting against the Kaibab walls with distinct and coarse basal conglomerate, are very friable, intricately cross-bedded, rather coarse sandstones. In color they are dark red, deep brown, and in places almost black, in which respect they contrast strongly with the gray limestone walls of the canyon between which they lie, and also with the cream- colored, white, or gray drifts of dune sand which surround them, in turn, and partly bury them. Their age was not determined, since they appear to be absolutely unfossiliferous. At first the writer inclined to the idea that they were a phase of the basal Moenkopi, resting in an old pre-Permian channel, and while he still admits the possibility of this interpretation, it is also considered possible and perhaps probable that they are a wind-blown deposit of Tertiary or early Quaternary age, consisting largely of materials derived from the Moenkopi. That they rest in a channel cut at least one hundred feet into the Kaibab limestone, that they contain a coarse basal conglomerate of limestone bowlders, and that they are distinctly older than the present dune sands, admit of no question whatever. The general character and relations of these deposits can be studied more clearly from the illustrations given (Figs. 7 and 8). *H. E. Gregory, op. cit., p. 21. PRE-MOENKOPI UNCONFORMITY OF COLORADO PLATEAU 71 AREAL EXTENT OF THE UNCONFORMITY Many other workers who have studied the stratigraphy of this region have recognized the prevailing unconformity at the base of the Moenkopi, but in general it has heretofore been considered a minor break. The evidence goes to show that it is recognized over an area from the Little Colorado River in Arizona, east to the Zuni Uplift in New Mexico, and northwest to the Dirty Devil (Fremont) River in Utah. Fic. 7.—Cross-bedded sandy shale (the two dark masses to the left of the center) resting in erosion channel in Kaibab limestone; dune sand in the foreground. On Little Colorado River, near Tolchico, Arizona. In the San Juan Oil Field, which lies about midway between the Ramah and Fruita localities, the Moenkopi, according to Woodruff, rests on the Goodridge (Redwall ?) limestone with ‘‘a sharp litho- logic break,” though he does not mention an actual erosion interval. Of the same place Gregory’ says: “‘ No undisputed evidence of uncon- formable relations between the Pennsylvanian and Permian (?) was obtained at this locality.”” He indicates, however, evidence of a probable break of importance in sedimentation. In view of tE. G. Woodruff, “‘Geology of the San Juan Oil Field, Utah,” U.S. Geol. Surv., Bull. 471, p. 87. 2H. E. Gregory, op. cit., p. 21. 72 €. L! DAKE f the unconformable relations both northwest and southeast, it is highly probable that the unconformity also exists along the San Juan. > STRATIGRAPHIC MAGNITUDE OF THE UNCONFORMITY Stratigraphically, near Mule Twist Canyon, this unconformity represents at least two hundred feet of erosion. If the beds in the Mule Twist vicinity are actually different geological horizons. Fic. 8.—Detail of contact of sandy shale shown in Fig. 7, on Kaibab limestone. Note pebbles of limestone in the shale. from those near Fruita, rather than different facies of the same horizon, a conclusion that seems highly probable, then the pre- Moenkopi erosion was even greater. Let us examine now the significance of certain tentative cor- relations made by Girty and presented by Woodruff.‘ Regarding the age of the Goodridge limestone in the San Juan Field, Girty says: T have already examined and reported upon a collection from the Honaker trail, where a good portion of Mr. Woodrufi’s material was obtained. This collection was made by Robert Forrester, of Salt Lake City, Utah. Mr. For- rester, who has done much work of a very accurate kind involving the Mesozoic tE. G. Woodruff, op. cit., pp. 75-104. PRE-MOENKOPI UNCONFORMITY OF COLORADO PLATEAU 73 and late Paleozoic rocks of Utah, reports that his fossils came from what was called Lower Aubrey group in the reports of the Wheeler Survey, their Upper Aubrey being our Kaibab limestone. The lists of fossils given by Meek as representing the fauna of the upper Redwall limestone show the same general- facies as Mr. Woodrufi’s collection. The typical Redwall we know to be of Pennsylvanian age in the upper part and Mississippian age in the lower part, so that the facts at hand seem to indicate that the strata involved in Mr. Wood- ruff’s collection represent the upper part of the typical Redwall limestone. I do not regard it as certain, however, that the marked dissimilarity of the Kaibab fauna to anything which Mr. Woodruff found in his section may not be regional and that by gradual modification some of his faunas may not pass into the Kaibab fauna at the same geologic level. If Girty is right in his tentative suggestion that the Goodridge is the equivalent of the Redwall, the pre-Moenkopi erosion interval at once assumes greatly added significance as a stratigraphic break of notable magnitude, for this would indicate the removal, or non- deposition, in the San Juan Field of the Supai sandstones and shales, the Coconino sandstone, and the Kaibab limestone, all of which occur above the Redwall in the Grand Canyon section, the com- bined thicknesses of which are not far from two thousand feet. This supposition is perhaps somewhat strengthened by the facts observed near Mule Twist, where the formations resemble closely the Kaibab and Coconino. The fossils collected by the writer, had they not been lost, might have settled this point. It is to be hoped that other collections may be secured soon from that locality. From this it would seem that the Moenkopi may be resting on Kaibab limestone near Tolchico and in the Grand Canyon, on Redwall limestone in the San Juan Oil Field, and again on Kaibab limestone and Coconino sandstone near Mule Twist Canyon. The foregoing is a possibility which, the writer finds, has already been considered by Cross‘ in explaining the fact that the Pennsyl- vanian directly beneath the ‘‘Red Beds” at Moab carries a different and possibly older fauna than was found by Powell and Newberry below the Red Beds farther west on Colorado River. Cross says: There may be a stratigraphic break, due to uplift and erosion, through which the Aubrey strata found by Powell and Newberry have been removed at Moab, in the Sinbad Valley, and to the mountain region to the east. This t Whitman Cross, ‘Stratigraphic Results of a Reconnaissance in Western Colorado and Eastern Utah,” Jour. Geol., XV (1907), pp. 634-79- 74 : C. L. DAKE implies that the Hermosa beds of Moab are present beneath the section examined by Powell and Newberry. Such a break must occur at the base of the Paleozoic “‘Red Beds,”’ and no suggestion of such a hiatus has come from observations in Colorado; but it is to be remembered in this connection that in southern Utah and northern Arizona, Powell, Gilbert, Dutton, Walcott, and others have noted a persistent unconformity by erosion between the Aubrey and the succeeding strata now commonly referred to the Permian through Walcott’s discovery of fossils in the Kanab Valley. All of the above-named geologists have observed a conglomerate more or less widely distributed at * the base of the Permian series, composed in large part of pebbles derived from the Aubrey rocks, as shown by fossils contained in them. It is, of course, possible that the denudation at this horizon may have been much more exten- sive than the observations thus far reported would suggest. The direction even of this change corresponds to the facts observed by the writer. According to the statement by Cross, the beds on which the Moenkopi rests are younger to the west, older to the east. Similarly the writer finds that in the San Juan Oil Field, about in the longitude of Moab, the Moenkopi rests on beds which are possibly as old as Redwall, while farther west, at Tolchico and at Mule Twist Canyon, the Moenkopi seems to be resting on younger beds, probably the Kaibab. That the beds below the Moenkopi near Mule Twist ‘Cangan were equivalent to the Kaibab and Coconino, and were not equiva- lent to the Goodridge, was the independent conclusion reached by the writer, even before he was aware of the foregoing statements by Girty and by Cross. In view of the uncertain condition of this correlation, it was felt that these few notes might add to the general knowledge regarding the extent and magnitude of this break. PALEOZOIC DIASTROPHICS OF THE NORTHERN MEXICAN TABLELAND CHARLES KEYES Des Moines, Iowa A little below the southern boundary of Colorado the Rocky Mountains, in a triple cluster of canoe-shaped folds, plunge steeply beneath the general plains surface of the northern prolongation of the Mexican tableland, never again to reappear. In marked con- trast to the prevailing relief expression of the Cordillera, with its stream-cut profiles, the physiognomy beyond is that of typically enisled landscape of the desert, fashioned mainly by the winds. The chain aspect of the Rockies gives way to solitary ridges. All mountains assume the character of short, lofty ranges which, with startling abruptness, rear themselves like volcanic isles jutting from a summer sea. So thickly do these isolated piles stud the vast smooth plateau plains that Dutton aptly likens them on the map to an army of caterpillars crawling northward out of Old Mexico. Inasmuch as the northern extension of the Mexican tableland is included mainly within the present limits of the state of New Mexico its topographical features, so far as the United States go, are in many respects quite unique. Throughout this region’ the areal distribution of the geological formations is probably the least understood of any considerable tract in our country. In only a few circumscribed districts is the geological structure brought out properly and correctly in mapping. Elsewhere, according to published information, the region seems to be a veritable terra incognita. Even the larger relationships of the formations are so little known that they have yet to be exactly determined. Over this northern segment of the lofty tableland the general plains surface lies evenly about a mile above the sea. Towering still another mile in the air are the innumerable mountain masses. The intermontane plains being chiefly desert or semi-arid, rock outcrops are few in number; and drifting sands and mobile earths 75 76 CHARLES KEYES prevail. Over such a country geological boundaries are not easily traced; and determination of the original areal distribution of the various terranes is beset with exceptional difficulties. In most other parts of the world the local stratigraphic succession and general mapping of rock tracts are based chiefly on the rock exposures bordering the valleys of incised streams. ‘These outcrops as a rule lie below the general upland level of the country. In the New Mexican field, the rock sections lie principally above the general plains surface. Correlative determinations of outcrops are thus exactly the reverse of what they usually are. Disposition of the sections, a mile or more high in many instances, is that of a myriad of drill-cores set upon a board.. Spaces between sections are voids in nature as in model. Under ordinary circumstances these intersectional intervals are filled up by the rock masses of the interstream areas. In order properly to visualize the geological formations of the tableland the various sections have to be connected and projected on a common plane. Such a ground plan is very different from a normal geological map. Yet it is the only kind of a diagram that satisfactorily depicts the larger relationships of the geological formations. For the region under consideration such a projection is represented in the accom- panying cut (Fig. 1), the base plane selected. being the ancient peneplain lying at the base of the great Pennsylvanian limestone plate. The fact that at the southern extremity of the Rockies the Penn- sylvanian limestones everywhere rest directly on pre-Cambrian schists long led to the inference that a region of continental pro- portions had been a land area during the greater part of Paleozoic times. Such apparently was not the case. When, a decade ago," the geological formations of New Mexico were briefly described in a systematic way, a circumstance that was expressly pointed out was that while over all the northern half of the state there were no Paleozoics below the Pennsylvanian limestones, it did not preclude the existence of some, or even all, of the early periodic sections else- where. In fact, in southern New Mexico isolated sections of these t Report of Governor of New Mexico to Secretary of Interior for 1903, pp. 337-41, 1904. _ DIASTROPHICS OF NORTHERN MEXICO gyi older rocks were incidentally noted very early. These outcrops were so widely disconnected as to give rise in some quarters to not a little doubt concerning the actual presence of some of the terranes. — oo enc cme cm [es 6 + came oe mee mene 6 2 CE COREE + ¢ GHEE 6 ¢ GumEED 00 GES © eae wa l | | - | NEW MEXICO Fic. 1.—Areal distribution of Paleozoic periodic formations Pioneer observations now seem to be fully verified. As early as 1874 Dr. W. P. Jenney‘ called attention to the presence of Cam- brian rocks in the Franklin Range north of El Paso. In the same locality Dr. A. Wislizenus,? thirty years before, collected charac- teristic Ordovician fossils. Numerous organic remains of Silurian age t Proc. New York Lyc. Nat. Hist., Vol. I (1874), p. 60. 2 Memoir of Tour through Northern Mexico in 1846-47 (1848), 141 pages. 78 CHARLES KEYES were obtained at Santa Rita, in 1873, by Professor G. K. Gilbert. Devonian strata were first recorded in New Mexico by Thomas Antisell;? and in the following year the fossils were described from this region by Professor James Hall. Mississippian forms collected by Professor E. D. Cope were described in 1881 by Mr. S. A. Miller.4 The most northerly points at which they were recognized recently were in the Magdalena Mountains, west of Socorro,5 and in the Sierra Ladrones 25 miles north. So in New Mexico prior to the year 1880 all of the periodic terranes of Paleozoic age had been already fully identified. In this tableland region the outstanding feature of the stratig- raphy, and a characteristic which is perhaps nowhere else met with, - is a notable segregation, instead of the usual alternation, of the hard and soft strata. Resistant beds are confined chiefly to the bottom half of the vertical section; and nearly all of the weak rocks occur only in the upper part. For a succession more than 10,000 feet in thickness this circumstance is certainly a quite remarkable one. Almost the entire Paleozoic sequence is thus composed of lime- stones of such uniform lithologic texture and aspect that it is not usually possible by casual glance to detect the parts of different geologic age. Only by careful discrimination of the successive faunas at the various stratigraphic levels are even the larger, or periodic, subdivisions rendered determinable. Yet, on the whole, the sequence is one of the most complete on the American con- tinent. Out of twenty-five major terranes holding serial rank only five seem to be missing. Both by reason of its completeness and because of its peculiar continental relationships this general geological section of the New Mexican Paleozoics is for purposes of reference one of the important successions of the country. As determined by various parties from the United States Geological Survey, and the State Geological Sur- * U.S. Geog. and Geol. Surv. W. rooth Merid., Vol. III (1875), p. 117. 2 Explo. and Surv. Pacific Railroad, Vol. III (1856), Pt. II, p. 181. 3 United States and Mexican Bound. Surv., Vol. I (1857), Pt. I, p. 104. 4 Jour. Cincinnati Soc. Nat. Hist., Vol. IV (1881), p. 314. 5 Proc. Iowa Acad. Sci., Vol. XII (1905), p. 169. DIASTROPHICS OF NORTHERN MEXICO 79 vey, and by others who have worked more or less extensively in the region, the essential features of the section are well epitomized in recently published tables." Since it is with marked unconformity that the Paleozoics rest upon the pre-Cambrian crystallines it is evident that long before — Paleozoic deposition in the region set in, the old continental com- plex was beveled off to a smooth plain. This ancient erosion surface cuts evenly the folded, faulted, and altered pre-Cambrian strata, the more or less highly metamorphosed clastics, and the strictly igneous masses and intrusions. It doubtless represents as true a peneplain as ever existed, and one that remained longer and nearer base level than any other one known in geological history. The presence of this once low-lying plain and the near-shore deposi- tion of the vast piles of homogeneous limestones appear strongly to support the idea of the existence of a close genetic relationship between the two phenomena. Notwithstanding the fact that such exceptional homogeneity prevailed throughout the Paleozoic succession of the northern Mex1- can tableland, no less than a dozen major unconformities attest the frequency and extent of notable diastrophic movement. Of these by far the most conspicuous hiatus is that at the base of the Penn- sylvanian limestones. In every way it is the most pretentious. Its character and position associate it with the similar phenomenon displayed in the Mississippi Valley. From that it differs in the apparent absence of the Coal Measures. However, this dissimilar- ity fades since remnants of the latter are now known to be actually present. How extensive they originally were is yet a matter of conjecture. In the Ladronesian series, exposed in a circumscribed basin near Socorro,? are represented the all but vanquished coal shales which may be the southwestern extension of the great Arkansan series of the Ozark region. The unconformity plane is comparable in every way with that found at the base of the Coal Measures of Iowa, Missouri, and Illinois. It extends far to the north in Colo- rado; and far to the south in Old Mexico. Although in the Rocky t Proc. Iowa Acad. Sci., Vol. XXII (1916), pp. 249-71. 2 Journal of Geology, Vol. XII (1904), pp. 250-51. 80 CHARLES KEYES Mountains region all strata down to the pre-Cambrian complex are removed and the Pennsylvanian limestones rest directly upon the ancient crystallines, in southern New Mexico the same limestone plate reposes on the beveled edges of all of the older and somewhat deformed Paleozoics (Fig. 2). Certain peculiarities in the areal distribution of the Paleozoic formations in southern and central New Mexico at once raise far- reaching questions in diastrophics. Among them not the least significant is whether the northern limits of the several major ter- ranes are approximately the original boundaries of deposition, or Ceanett : XOo ge PROTEROZOIC §) 9° Fic. 2.—Relations of periodic formations south of Rocky Mountains whether these periodic terranes once extended indefinitely northward over the tract which later was upraised into the Rocky Mountain arch. Casual consideration of present conditions points sometimes to one conclusion and sometimes to the other. Critical evidence centers on the character of the marked unconformity at the base of the Pennsylvanian section. The present northern boundaries of the other periodic terranes are very close together (Fig. 1). The strata are all virtually limestones. There are practically no characteristic shore deposits represented. Positions of none of the formations indicate that there are any well-defined overlaps. All features considered, the conclusion appears inevitable that the strand oscillations at different times were of great magnitude, DIASTROPHICS OF NORTHERN MEXICO 81 amounting to hundreds of miles, instead of very short distances as at first glance seems probable. Paleogeographical maps commonly show the southern Rocky Mountain region as a huge island persisting throughout Paleo- zoic times. Orographic arching of the tract appears to have taken place only late in the era. For the first time since pre-Cambrian days general peneplanation does not appear until the Mississippian or Pennsylvanian period. That the peneplanation at the beginning of Pennsylvanian times, when Coal Measures were being deposited elsewhere around the growing American continent, was extensive is strongly supported by many facts. Since farther north in the Rocky Mountains area the older Paleozoics are present in a few limited and isolated belts, where they are preserved through infolding with more ancient — rocks, it is presumed that Cambrian, Ordovician, Silurian, Devonian, and Mississippian strata as they are represented in the South doubt- less once extended entirely over the province before its epeirogenic uprising. Over the Mexican tableland district last lingering traces of the old formations remained until the grand erosional period represented elsewhere to the eastward by the Arkansan (Pennsylvanian) deposition. It may be that the Pennsylvanian peneplanation epoch of the southwestern region is to be exactly paralleled with that of Iowa where it is designated as the Arkan- san hiatus. Another reason why in the Cordilleran region north of the Mexi- can tableland the Paleozoics do not appear more frequently than they do is that Triassic peneplanation was also profound. This surface is largely covered by mid-Cretaceous sediments before the eastern front of the Rockies is reached. Along this border the more ancient rocks are thus not open to inspection. The rather abrupt termination of the several periodic terranes of the Paleozoic toward the north in central New Mexico does not appear to be altogether a direct result of successive advances of the ancient sea in that direction over a low-lying even coast. If any part of the abrupt thinning is thus to be ascribed it is entirely lost through repeated and profound planation effects. After the lay- ing down of the great Pennsylvanian limestones no less than two of 82 CHARLES KEYES these planation periods are clearly indicated by marked unconform- ities displayed in the general stratigraphic succession of the region. When the great peneplanation of Jurassic or early Cretaceous times took place, as marked by the basal surface of the extraordinarily widespread Dakotan sandstones, it was accompanied by extensive and diverse deformation with some display of volcanic action. So extensive was this main evening that so far south as central New Mexico the Dakotan sandstone is seen to repose upon the upturned Seale ‘ 200 feee Cretacious _ Sandslones OX IWS tron Ore Z70R Ore Fic. 3.—Unconformable relations of Cretaceous and Pennsylvanian beds on Chupadera Mesa, New Mexico. and locally vertical edges of the Pennsylvanian limestones. A most notable section demonstrating these relationships is well displayed on the Chupadera Mesa, at Dios Springs, and in the Arroyo Chupadera, about thirty miles northeast of Socorro (Fig. 3). The fancied complexity of the stratigraphy of the Mexican tableland is therefore more apparent than real. ‘Two features in particular tend to obscure the actual mass relationships of the formations. Of these the wide separation of exposed rock sections assumes an importance out of all proportion to its difficulties or its merits. The phenomena attending diastrophic movements in the region are thus liable to serious misinterpretation. Neither the DIASTROPHICS OF NORTHERN MEXICO 83 moderate flexing nor the profound faulting are found to be so recent as to retain their impress in full force upon the present relief. These major crustal movements are mainly quite remote. Begin- ning at the close of Paleozoic time they continue without interrup- tion until the present day. The effects which their supposed dominancy produces in the existing desert ranges are now known to be due entirely to other causes. On the whole the desert ranges seem to owe their physiognomy to vigorous wind erosion under the stimulus of aridity rather than to recent deformation. Under these abnormal conditions old structures are brought out into strong relief by simple differential erosion of weak and resistant rock belts. This desert degradation so vigorous at the present day may have been of long duration, having gone on since the beginning of Tertiary times. The post-Cretaceous wrinkling of the Rocky Mountains is reflected in the areal distribution of the rock formations far beyond the southern terminus of that cordillera, reaching many miles into the Mexican tableland. SOME ESTIMATES OF THE THICKNESS OF THE SEDIMENTARY ROCKS OF OHIO T. M. HILLS Ohio State University Since the discovery of crystalline rock at a depth of 3,320 feet at Waverly* in 1911, and at Findlay," at a depth of 2,770 feet, in 1912, deep wells have been drilled in many parts of Ohio. From their records, the following estimates are made of the thickness of* the sedimentary rocks along the eastern and southern borders of the state. : The data used? have been taken from wells located along three lines, two of them running east-west, the other north-south. The first extends from Findlay through Cleveland eastward to the central part of Ashtabula County, the second from the city of Columbus east to northern Muskingum County, the third from Norwalk to Jefferson Township, Jackson County. The two formations used as datum planes in the calculations are the Trenton and the Clinton. The tops of these formations are recognized with a reasonable degree of certainty by drillers. Unfortunately, the two wells mentioned above are the only ones that have passed through the Trenton rocks of the state. There- fore the distance from the top of the Trenton to the crystalline rocks must be taken from these wells and considered as constant for the area. ‘The wells in the eastern and central parts of the state do not extend below the Clinton; therefore the depth from this formation to the Trenton, and to the crystalline rocks below, must be supplied from the known wells of other parts of the state. This assumes that the Trenton-to-crystalline-rock interval is constant over a wide area, and that the Clinton-to-Trenton interval varies at auniform rate. These are both broad assumptions, but with the data available cannot be avoided. * Condit, Amer. Jour. Sci., Fourth Series, Vol. XXXVI, p. 123, August, 1913. 2 Acknowledgments are due The Ohio Geological Survey for the use of its files of well data. 84 SEDIMENTARY ROCKS OF OHIO 85 The northern line of wells —At Findlay the interval between the crystalline rocks and the top of the Trenton formation is 1,605 feet. This will be considered as constant for northern Ohio. At Lorain, the interval between the Clinton and the Trenton is 1,075 feet, at Cleveland 1,658 feet, an increase of 21 feet per mile to the eastward. If this continues to the state line sixty miles eastward, the interval would be 2,918 feet (1,658+1,260). In Wayne Township, Ashtabula County, the Clinton formation is found 2,940 feet below sea level. Data from wells at Lorain, Avon, East Cleveland, Chester Township, Geauga County, Harts- grove and Wayne townships, Ashtabula County, show an average, although not constant, decline of the surface of this formation of 22.7 feet per mile. If the eastward decline continues at the same rate to the state line, the surface of the formation should be about 3,172 feet below sea level at this point. Add to this figure the estimated distance between the top of the Clinton and the top of the Trenton, 2,918 feet, and the distance from the top of the Trenton to the crystalline rocks, 1,605 feet, and we have a total of 7,695 feet, the depth below sea level at which crystalline rocks should be found near the northeastern corner of Ohio. An addition of at least 1,000 feet should be made for the thickness of strata above sea level, giving a total estimate of some 8,700 feet of sedi- mentary rocks in this part of the state. Central Ohio.—The wells extending from Columbus to north central Muskingum County are not on a straight line, but the departures to the north side are practically balanced by those to the south. It will be seen later that the variation in thickness of the sedimentary rocks along a north and south line is comparatively slight in short distances. The Clinton occurs 186 feet below sea level in a well along the Mifflin Township line in the eastern part of the city of Columbus. From this well to one at Basil, the top of the Clinton declines at an average rate of 41 feet per mile. Eight other wells, some of them as far east as north central Muskingum County, show a decline ranging from 56 feet to 34.9 feet per mile, with an average of 46.6 feet. Using the last figure, the top of the Clinton should be 5,778 feet (=186+5,592) below sea level at Wheeling, 120 miles eastward. 86 T. M. HILLS At Waverly the top of the Clinton is 310 feet below sea level, and the top of the crystalline rocks, 2,730 feet. The interval between is 2,420 feet. If at Wheeling the interval is the same, the crystalline rocks would be found at 8,198 feet (=5,778+2,420). Add 1,000 feet for the strata above sea level, and the estimate is . brought to 9,198 feet. This figure does not include the increase in the interval between the top of the Clinton and the top of the Trenton found along the northern part of the state, which was. 21 feet per mile. Addition for such thickening would add 2,520 feet, bringing the total to 12,028 feet. The north and south line of wells —This row of wells is practically at right angles to the other two. At Norwalk the Trenton was reached 1,945 feet below sea level. In Jefferson Township, Jackson County, it was found at 2,885 feet, a difference of 940 feet in 162 miles, a decline of 5.8 feet per mile. If this decline continues southward to the Ohio River, twenty-five miles farther, the Trenton would be found there 3,030 feet below sea level. The Trenton (top)-to-crystalline interval of 1,220 feet, found at Waverly, would place the bottom of the sedimentary rocks 4,250 feet below sea level. -Add a thousand feet for strata above sea level, as in the previous cases, and we have 5,250 feet for the thickness of strata above the crystalline rocks at the cee River, near Ironton. Summary.—From the northern line of wells the sedimentary strata of northeastern Ohio are estimated to be nearly 9,000 feet, from those of central Ohio to be over 12,000 feet in the eastern part of the state, and in southern Ohio to be more than 5,000 feet thick. Wells along other lines give results of the same order of magni- tude, the same assumptions concerning the interval between the top of the Trenton and the crystalline rocks being made. Since the post-Trenton strata thicken toward the Appalachian trough, either by the increase in the thickness of the formations themselves, or the introduction of new formations, it is reasonable to suppose that the pre-Trenton sediments do the same, so that the results obtained are probably underestimates rather than overestimates. The presence of the Cincinnati Arch in the southwestern part of the state adds so many complicating features that it does not now seem advisable to attempt an estimate for this part of the state. RECENT. PUBLICATIONS —SEIBERT, F. M., and Harpster, W. C. Use of the Interferometer in Gas Analysis. [U.S. Bureau of Mines, Technical Paper 185. Washington, 1918.] —Sewarp, A. C. Fossil Plants. Vol. III. Pteridspermeae, Cycadofilices, Cordaitales, Cycadophyta. Cambridge University Press. [London: Fetter Lane, E.C. New York: G.P. Putnam’s Sons. 1917. Price 18s. net.| —SMELLIE, W. R. The Igneous Rocks of Bute. Transactions of the Geo- logical Society of Glasgow, Vol. XV, Part III, pp. 334-73. Plates XXXVI-XXXIX. [Glasgow, 1916.] —SmitH, C.G. Cost Accounting for Oil Producers. [U.S. Bureau of Mines, Bulletin 158. Washington, 1917.] —Smitu, P.S. The Lake Clark—Central Kuskokwim Region, Alaska. [U.S. Geological Survey, Bulletin 655. Washington, 1917.] ———. The Mining Industry in Alaska in the Calendar Year 10916. [U.S. Bureau of Mines, Bulletin 153. Washington, 1917.] —SToNE, R. W. Gypsum Products, Their Manufacture and Uses. [U.S. Bureau of Mines, Technical Paper 155. Washington, 1917.| —Tay1or, C. H., Editor. Bulletin of the American Association of Petroleum Geologists. Vol. II. 1918. Bulletin of the Southwestern Association of Petroleum Geologists. Volels> 10917: —Taytor, G. B., and Cops, W. C. Initial Priming Substances for High Explosives. [U.S. Bureau of Mines, Technical Paper 162. Washing- ton, 1917.] —Tennessee, The Resources of. Vol. VIII. 1918. [Tennessee Geological Survey, Nashville.] —TuHompson, J. W. Abstracts of Current Decisions on Mines and Mining, January to April, 1917. [U.S. Bureau of Mines, Bulletin 152. Wash- ington, 1917.| Abstracts of Current Decisions on Mines and Mining. [U.S. Bureau of Mines, Bulletin 159. Washington, 1917.] —THORNBERRY, M. H., and Mann, H. T. The Effect of Addition Agents in Flotation. Part I: Sulphates, Hydroxides and Nitrates. University of Missouri School of Mines and Metallurgy, Bulletin, Vol. IV, No. 2. [Rella, 1917.] 87 88 RECENT PUBLICATIONS —Tovucu, F.B. Methods of Shutting Off Water in Gas and Oil Wells. [U-S. Bureau of Mines, Bulletin 163. Washington, 1918.] - —TyRRELL, G. W. The Whangie and Its Origin. Scottish Geographical Magazine, XXXII, 14-24. 10916. The Petrography of Arran. Geological Magazine, New Series, Decade VI, Vol. III, pp. 193-96. [London: Dulan & Co., 1916.] Further Notes on the Petrography of South Georgia. Geological Magazine, New Series, Decade VI, Vol. III, pp. 435-41. [London: Dulan & Co., 1916.] —Warinc, G. A. Ground Water in Reese River Basin and Adjacent Parts of Humboldt River Basin, Nevada. [U.S. Geological Survey, Water- Supply Paper 425-D. Washington, 1918.] —Washington Academy of Sciences, Journal of. Vol. VIII. —Wells, R. C. Sodium Salts in 1917. [U.S. Geological Survey, Mineral Resources of the United States, 1917, Part II: 23. Washington, 1919.] —Wiuson, M. E. Timiskaming County, Quebec. [Canada Department of Mines, No. 1695, Memoir 103, Geological Survey, No. 86. Geological Series. Ottawa, 1918.| —Woop, H. O. A Further Note on Seismometric Bookkeeping. Seismo- logical Society of America, Bulletin, Vol. VII, No. 3. [Hawaiian Volcano Observatory, 1917.] %) aN Short History of Belgium. By LEon VAN DER EssEN, Professor of History in the University of Louvain. $1.50, postpaid $1.65. In this new and enlarged edition the author tells in an interesting way of the formation, struggles, and triumphs of the little kingdom; presents additional material on the reign of Leopold IT; and dicscries the heroic nation’s part in the Great War. _A Field and [abaratory Guide in Physical Nature-Study. By Exzior R. Downine, Associate Professor of Natural Science in the School of Education, the University of Chicago. “Loose-leaf, 110 pages of text, interleaved, paper covers; $1.00, postpaid $1.10. Tells how to make toys that work by air, heat, electricity, light, "and other physical agencies. The child learns the principles without effort oe while he is absorbed in making the toys. Also tells about the common minerals and ¥ Gd rocks and the stars that are the wonder of every child. a The Geography of the Ozark Highland of Missouri. By Cart Ortwin SavER, the University of Michigan. $3.00, postpaid $3.20. The purpose of sucha study is twofold: to furnish an adequate explanation of the conditions of life in a given area,.and to contribute proved statements which will aid in working out funda- mental principles. A valuable feature _ the volume is the iorty-four figures in the ea and twenty-six plates. sid A Maviual of Style (Revised Edition). By the STAFF OF THE UNIVERSITY OF CHI- J caco Press. $1.50, postpaid $1.65. An invaluable handbook for writers, editors, _ teachers, librarians, ministers, lawyers, printers, publishers, proofreaders—for all who need guidance on matters of propriety in style. ii erent Economic Problems (Revised Edition). Edited by Watron H. Hamitton, Professor of Economic Institutions, Amherst College. $3.50, postpaid 1°$3-75.. A selection of readings from the widest range of sources. The revision pre- brought about by the war in many lines of thinking on économic subjects. -_ postpaid $1.90. As the problem of fertilization deals with the union of male and | female sex elements, which forms the beginning of each individual’s life-history, it has important bearings upon the problems of heredity: It also has close relations to several other fundamental biological problems. "The Revelation of Jehu. Is the Book of Revelation a Mystery to Your By SuirLEy J. Case, Professor of Early Church History and New Testament, the Uni- versity of Chicago. $2.00, postpaid $2.15. This is a popular presentation of the subject and not a technical commentary. The author tells why and when the book “was written. Moore, Harvard University. $2.00, postpaid $2.15. The book gives a survey of the expansion of Christianity since the beginning of the modern era, about 1757, and depicts the movement against the background of general history. THE ‘UNIVERSITY OF CHICAGO PRESS ~ a ~ - - - - ILLINOIS” S(O AIEAIB ANNAN AIAS/OA/03/09)0108 + BAB BAB BANE RISFGS ASCH HIS Ball sents new material made available as a result of the war, and the new point of view Pee blamis of Fertilization. By Frank R. Lrtuir, University of Chicago. $1.75, The ‘Spread of Christianity in the Modern World. By Epwarp Capwett ~ TLE oe By CARL ORTWIN SAUER, the University of Michigan (Publications of the Geographic Society of Chicago) The Ozark Highland of Missouri was selected for investigation because of unusual wealth of geographic responses and because little is known concerning its Co ditions and possibilities. The purpose of such a study is twofold: to furnish an adequate explinaeve 0 the conditions of life ina given area, and to contribute proved statements iste | il aid in working out fundamental principles. The book is divided into three parts. The first is an outline of the environm nt Only those things which are pertinent to an understanding of the conditions under whic the people live are introduccd. - Ne The second part considers the influence of environment on the settlement aude ment of the different parts of the highland. The third part is a study of the economic conditions as they exist today. In co clus a forecast is offered of the lines along which the future of the region will be worked pura ga A valuable feature of the volume is the 44 figures in the text and 26 plates. xviti+ 246 pages, Svo, cloth; $3.00, postpaid, $3.20 The University of Chicago Press Chicago, Illinois Paragon Typewriter Ribbons For all makes of machines. Of every desired color and in many variations in two colors. Made in our own factory because we are unwilling to trust any outside concern with the manu- facture of an item so essential to the good appearance of your letters. Send for the folder, “‘What the Magnifying Glass Shows.’ “ned le | Emancipate yourself from the use of corrosiv. e am ill-smelling inks and adhesives and adop tells why Paragon Ribbons last . ‘Higgins Inks and Adhesives. They willl revelation to you, they are so sweet, c and give sharp, clear im- put up, and withal so efficient. : : At Dealers Generally pressions. REMINGTON TYPEWRITER CO. CHAS. M. HIGGINS & CO., i Branches: Chicago, London — 4 Broad N York City | ial nehevadee 3 eee 271 Ninth Street Branches Everywhere aS NUMBER 2~ EOLOGY a SEM QUARTERLY 6 EDITED BY “THOMAS c. CHAMBERLIN AND ROLLIN D. SALISBURY ee y With the Active Collaboration of Berets Paleontology ALBERT JOHANNSEN, Petrology ROLLIN T. CHAMBERLIN, Dynamic Geology i ing Pa _ ASSOCIATE EDITORS JOHN C. BRANNER, Leland Stanford Junior University RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. WILLIAM H. HOBBS, University of Michigan FRANK D. ADAMS, McGill University CHARLES K. LEITH, University of Wisconsin - WALLACE W. ATWOOD, Harvard University WILLIAM H. EMMONS, University of Minnesota ARTHUR L. DAY, Carnegie Institution — * FEBRUARY-MARCH 1920 , ; Serre =" #2... =. Grorcr F.. Kay ann J. NewTon Pearce. ~ 89 iE UNIVERSITY, OF CHICAGO: PRESS CHICAGO, ILLINOIS, USS2/Ag ail _ THE CAMBRIDGE UNIVERSITY PRESS, Lonpon : “MARUZEN-KABUSHIKL-KAISHA, Toxyo, Osaka, Kyoto, Fukuoka SeNDAt eh) sa ee MISSION BOOK COMPANY, SHANGHAI | Pie ae Shae 1920 IME JOURNAL OF EDITED’ BY.2>* iy - se a an THOMAS C. CHAMBERLIN AND ROLLIN D. ‘SALISBURY _ Noe With the Active Collaboration of STUART WELLER’ .- Pa ke et ALBERT JOHANNSEN a Invertebrate Palcaneelsey i Re Peoley ROLLIN: Hib CHAMBERLIN Dynamic Geology Press, 5750 Ellis Avenue, Cac Tllincis, semi- ey on or about the “followin March 15, May 1, June 15, August 1, September 15, November 1, December 15. price is $4.00 per year; the price sf single copies is 65 cents. sores for service of les: eae will be charged at the single-copy Tate. ‘] Postage is prepaid by the Pa on all Tine Philippine Islands, Guam, Se ae Shanghai. § Postage i is ‘charged oe For Canada, 30 cents on annual subscriptions (total $4.30), on single copies 4 cents. (total 6 for all other countries in*the Postal Union, 53 cents on annual subscriptions (total $4. 53), 0 copies 11 cents (total 76 cents). {| Patrons are requested to make all remittances payable t to - versity of Chicago Press in postal or express money orders or bank drafts. ; The following are authorized to quote the prices indicated: hes. For the British Empire: The Cambridge University Press, Fetter Lane, Lanta E. ey e subscriptions, including postage, 21s. each; single copies, including postage, 3s. 6d. each. | ‘For Japan and Korea: The Meee a eneeK eta, tr to 16 Nihonbashi Tree Tokyo, Japan. Yearly subscriptions, including posteee, Yen 9.00 ce single copies inch postage, Yen 1.50 each. a. For China: The Missioh Book Company, 13 North Szechuen Road, Shanghai, "Yearly su - scriptions $4.00; single copies 65 cents, or their equivalents in Chinese money. Postage extra, mailed direct gutside of Shanghai, on yearly subscriptions 53 cents, on single copies Ir cents. a Claims for missing numbers should be made within the month following the. regular month f . cation, The publishers expect to supply missing numbers free only yon losses have been sustain do transit, and when the reserve stock will permit. Business correspondence should. be addressed to The University of Chicago ree Chicago, ll Communications for the editors and manuscripts should be addressed to the. Editors, of SGUENAL. 01 or GEOLocy, the Vater os ue Chicago, ill. ee aa. to follow the form shown in the pages of the Journal: : _ Separates, if desired, must be ordered i in advance of publication. Separates. of leading articles (and covers, if desived) to be paid for by the author. The actual cost may vary, and will deena the amount of work in remaking the pages into forms, presswork, paper, binding, etc. Separate: taining half-tones may be expected to cost somewhat more, the increase copes ope the n n of cuts and the amount of. work required upon on Acceptance for mailing at special rate of postage provided for in Section 1103, we of October Br July 15, 1918. : : Among articles to appear in early numbers of the Journal of Geology are the following: Notes on Principles of Oil Accumulation. By A. W. McCoy. Discussed by CHESTER W. WASHBURNE. Reply to Discussion by C. W. Washburne on Notes on Principles of Oil Accumulation. By A. W. McCoy. Diastrophism and the Formative Processes. XII. The Physical States of the Earth and Its Neighbors during Their Formative Stages. By T. C. CHAMBERLIN. Compilation and Composition of Bituminous Coals. By REINHARDT THIESSEN. Geological Setting of New Mexico. By C. R. KEYES. Slides in the Conemaugh Formation near Morgantown, W est V irginia. By Eart R. SCHEFFEL. Factors Producing Columnar Structure in Lavas and Its Occurrence near Melbourne, Australia. By ALBERT V. G. JAMES. Movements in Crystallizing Magma. By FRANK F. GRovutT. Deformation of Crystalline Magma. By N. L. BOWEN. The Chester Series in Illinois. By STUART WELLER. The Juan de Fuca Lobe of the Cordilleran Ice Sheet. By J H. BRETZ. POSITIONS OF ALL KINDS Never was the demand so great for qualified teachers and specialists. For ten years we have given our time and energy to this work. Write for our free literature. State qualifications briefly. Co-operative Instructors’ Association Marion . - Indiana ) A Manual for Writers By JOHN M. MANLY Head of the Department of English of the University of Chicago and JOHN A. POWELL A book designed to aid authors and all others who are concerned with the writing of English. It treats ina clear and convenient way the matters of grammar, spelling,, and general form about which writers need to be informed, and gives full directions on the preparation of “copy” for the printer and the correcting of proof. ‘The chapter on letter-writing is unique and gives just the help that is constantly wanted and that is not found in other manuals. viit+226 pages, t2mo, cloth; $1.50 net, postpaid $1.65 The University of Chicago Press Chicago - - - - Illinois The Tale of “The Cockpit of Europe” That is what Napoleon called the famous little country that defended itself so brave- ly in the Great War. Read about its wonderful struggles for unity and independence in the fascinating “Short History of Belgium” Professor a ve der Essen of the University of Louvain BELGIUM BY A BELGIAN A New and enlarged edition which includes an account of Belgium’s part in the Great War. It is a story every American will appreciate. $1.50 net, postpaid $1.65 The Pnizers\(3) of Chicago Press | Chicago - - - Illinois ‘ a ose, Leo 3 y = - ee ee ee ee ee ee ee a The Geography of the Ozark Highland of Missouri By CARL ORTWIN SAUER, the University of Michigan (Publications of the Geographic Society of Chicago) The Ozark Highland of Missouri was selected for investigation because of its unusual wealth of geographic responses and because little is known concerning its con- ditions and possibilities. The purpose of such a study is twofold: to ane an adequate explanation of the conditions of life in a given area, and to contribute proved statements which will aid in working out fundamental principles. The book is divided into three parts. The first is an outline of the environment. Only those things which are pertinent to an understanding of the conditions under which the people live are introduced. The second part considers the influence of euinommnen: on the settlement and develop- ment of the different parts of the highland. The third part is a study of the economic conditions as they exist today. In conclusion a forecast’ is offered of the lines along which the future of the region will be worked out. A valuable feature of the volume is’the 44 figures in the text and 26 plates. xvitit246 pages, 8vo, cloth; $3.00, postpaid, $3.20 The University of Chicago Press Chicago, Illinois a — i eyes iO VOLUME XXVIII NUMBER 2 THE FOURNAL OF GEOLOGY FEBRUARY-MARCH 1920 THE ORIGIN OF GUMBOTIL GEORGE F. KAY AND J. NEWTON PEARCE University of Iowa The name gumbotil' was proposed recently for clays of dis- tinctive characters which lie on glacial till and which are related closely to till. As originally defined, gumbotil is “‘a gray to dark colored, thoroughly leached, non-laminated, deoxidized clay, very sticky and breaking with starchlike fracture when wet, very hard and tenacious when dry, and which is chiefly the result of weathering of drift. The name is intended to suggest the nature of the material and its origin.” In Iowa there is gumbotil on the Nebraskan, Kansan, and Illinoian drifts. It has not been developed on the Iowan drift nor on the Wisconsin drift. SOME OF THE FORMER VIEWS REGARDING THE ORIGIN OF SUPER- DRIFT CLAYS Until recently these superdrift clays which are now called gumbotils had been found only on the Kansan and Illinoian drifts, in connection with which drifts the clays had been described under the name gumbo by several geologists. Regarding the origin of this gumbo there have been various interpretations, some of which will be outlined briefly. t George F. Kay, ‘‘Gumbotil, a New Term in Pleistocene Geology,” Science, New Series, Vol: XLIV, November 3, 1916. 89 go GEORGE F. KAY AND J. NEWTON PEARCE The gumbo of McGee.——Dr. W J McGee applied the name gumbo to the peculiar, tenacious clay which he found on his Lower Till,* and in referring to the habit of weathering of this till he states :? Where the clay is plastic and sand free and of the usual blue color, as in the superior peripheral portion generally, it commonly weathers whitish or ashen to a limited depth and forms a tenacious, intractable soil, drowning when wet and baking when dry. This phase is colloquially known as ‘‘gumbo,” sometimes as “hardpan,”’ and locally as “white clay,” or (from its behavior below the plow) “push land.” ; The gumbo of Leverett.—Mr. Frank Leverett in his monograph “The Illinois Glacial Lobe” described the gumbo which he found associated with the Illinoian and Kansan drifts in southeastern Iowa. He believed that the gumbo on the Illinoian drift was of the same age as that on the Kansan drift and favored the inter- pretation, although he was not satisfied fully with the view, that the gumbo is the result of aqueous deposition following submergence of the region. The gumbo and loess-silt of Bain.—Dr. H. F. Bain, in his report on the geology of Decatur County, Iowa, states that blue to drab- colored gumbo which in places overlies the Kansan drift and is distinct from the drift belongs stratigraphically with the loess, and presents the view that the gumbo suggests a quiet-water deposit which has been compacted or puddled by water.* In earlier geological reports of counties in Iowa the same author refers to similar material. In his report on Keokuk County, Iowa, he refers to a stiff, yellow to blue-gray, plastic, non-calcareous clay which is on the Kansan drift and beneath the loess on the uplands, and states:5 “‘It seems to be a deposit closely akin to the loess and 1 Kansan drift of present classification. 2W J McGee, ‘‘The Pleistocene History of Northeastern Iowa,” U.S. Geol. Surv., Eleventh Ann. Rept., Part I (1891), p. 500. 3 Frank Leverett, ‘‘The Illinois Glacial Lobe,” U.S. Geol. Surv., Monograph XXXVIII (1899), pp. 28-33. 4H. F. Bain, ‘‘ Geology of Decatur County, Iowa,” Iowa Geol. Surv., Vol. VIII (1897), p. 292. 5 H. F. Bain, “‘ Geology of Keokuk County, Iowa,” Iowa Geol. Surv., Vol. IV (1894), p. 302. gt THE ORIGIN OF GUMBOTIL A f) a ~ aa Rw ee: ES eee tek oe a 5 ay 2 ae PL i Z | a " al F A a We, a , ¥ 1 : , | ‘ r ; ALY 10R0 : : | ems }09) — Dyas [Asal a myrarsoay lad b, ANAM e (fed oval Ail NYSNY¥ . Merit meant - 4S NVdd¥ | Bovkio e A\untvo a| Is MN we a A [artes] ANIVYOW { Po vovsuing4on o- 4 on Ree _ A Ly a Etsy) nvionin) J» .-- a a i I. . SAGER ad 2 fsanion saa (| ss301f -~- - = FY AuNan (zzz) wo Ua aa EERE] JNIVHOW | Raa | NISNOOSIM QNa537 oe ihe Oe WY 190 4SN 4| 5 STB, afitouy NA WYYM DIYNK |- @ F0nrsps a A UOLOIUSTM @ > TAX .,' NOLONIRSYA | Hufate? 3 AKY ¥ Ay ¢ C JHU © Clee, ee¥ : aaa f% i q A ts IN AOR Se Seat NE Ns f SRC ore -/, /; Oelol g (4 i Sus ‘a NO wh ria Mi By Peagyin ae’ ag Sa1099 aualed q H S):tp: NO al ae WH J LITE, NA | Ah NAS QZ 4: . NOUNK GAY HK 301g ZZ ine. ¢ Kipp uF SAS win Z 7 Z Fac fre 7 Webi GNC Ve, VZV a ER a wow ‘OOM BLLAAZ aa 1 se Y, I as kiffeffracm Hl now kad : AS a dag Sabra ap Sp Na a. (2% Sb amo .f °° AL AIRE ENI EDEN ll g/cm eral eel ar 329 Iie Ul © Baa NOIS1A3¥ 01 493P ans WAAOLT 40 Si39HS 14140 JHL 40 avy i Neb sage Zs BS 4 TINO TI Pee Zy re y se UP «HNN, Ge Y CIC Y SGD Ei et te MEE Tamara) Pace Ce 92 GEORGE F. KAY AND J. NEWTON PEARCE probably genetically related to it. Very likely it is but a phase of that deposit though differing from it in its plasticity, color, and density.” In his report on Appanoose County, Iowa, Dr. Bain describes a loess-silt' on Kansan drift which he considers to have been deposited later than the drift, and which in character and origin seems much like the white clays of Ohio Valley described by Leverett.’ : In a geological report on Madison County, Iowa, Dr. H. F. Bain and Dr. J. L. Tilton refer to a dark-colored, impervious, unfossiliferous clay on Kansan drift as the lower of two ees of loess in the county.3 The gumbo of Udden.—Mr. J. A. Udden in a report on the geology of Pottawattamie County, Iowa, describes in considerable detail the gumbo or red clay associated with drift and makes the following statement regarding its origin: It would be premature at the present time to express any opinion as to the origin of this deposit. Probably it is mostly an old loess, which has been clogged up by interstitial deposition of fine ferruginous material through the agency of the ground water. Perhaps it is in part a fluviatile deposit, made at a stage of semi-stagnant drainage, or possibly it is of varied origin, being in some places a surface wash, or a disintegration product derived from an underlying bowlder clay, and at other places a modified upland loess, or a river silt. The Dallas formation of Tilton.—Dr. J. L. Tilton gave the name Dallas deposits to gumbo and related materials overlying Kansan drift in southern Iowa. He considered the deposits to have been formed during the closing stages of the Kansan glacial epoch.* The gumbo of Arey.—In a report on the geology of Davis County, Iowa, Professor M. F. Arey states that the gumbo which lies on the Kansan drift is perhaps a water deposit.® 1H. F. Bain, ‘Geology of Appanoose County, Iowa,” Iowa Geol.,Surv., Vol. V (1895), pp. 407-8. 2 Frank Leverett, ‘On the Significance of the White Clays of the Ohio Region,” American Geologist, Vol. X (1892), pp. 18-24. 3H. F. Bain and J. L. Tilton, “Geology of Madison County, Iowa,” Iowa Geol. Surv., Vol. VII (1896), p. 523. 4J. A. Udden, “Geology of Pottawattamie County, Iowa,” Iowa Geol. Surv., Vol. XI (1900), p. 258. 5 J. L. Tilton, ‘A Pleistocene Section from Des Moines South to Allerton,” Jowa Acad. of Science, Vol. XX (1913), pp. 218-20. 6M. F. Arey, “Geology of Davis County,” Iowa Geol. Surv., Vol. XX (1909), presi THE ORIGIN OF GUMBOTIL 93 The Loveland of Shimek.—Professor B. Shimek proposed the name Loveland formation’ for gumbo-like deposits related to Kansan drift in Harrison, Monona, and adjacent counties in southwestern Iowa. Detailed descriptions of the formation are given in his report on the geology of Harrison and Monona counties.’ Here he refers to the fact that Mr. Udden described similar material in Pottawattamie County as gumbo or red clay. It should be pointed out that the Loveland of Shimek differs in some important respects from the Kansan gumbotil of southern Iowa. ‘The gumbo- til is found only on glacial till and has a definite topographic position. According to Professor Shimek the Loveland in places lies not on till but on gravels. Moreover, it is not confined to a particular stratigraphic plain; the term has been applied to material in the lower part of the bluffs along the Missouri River in some places as well as to material 180 feet higher than the bases of the bluffs. Shimek considers the Loveland formation to be a water deposit which was formed during the stage of melting of the Kansan ice, and which has the same relation in general to the Kansan drift as have the Buchanan gravels to Kansan drift. This interpretation may, however, be somewhat open to question, as recent studies indicate. The super-Kansan gumbo of Alden and Leighton.—In a recent publication by Dr. W. C. Alden and Dr. M. M. Leighton there is a discussion of super-Kansan gumbo in Iowa. The authors do not commit themselves definitely with regard to its origin, but they present evidence which they consider favorable to the view advanced by Kay that the gumbo is the residuum of thorough weathering and long leaching of the upper part of the Kansan till. Sufficient evidence has been submitted to show clearly that the students of superdrift clays—the gumbotils and related materials— have been far from agreement regarding their origin. Some geologists have considered these clays to be mainly of fluvioglacial origin, others believe that they are aqueous, and still others have t B. Shimek, ‘‘Aftonian Sands and Gravels in Western Iowa,” Bull., Geol. Soc. Amer., Vol. XX (1910), p. 405. 2B. Shimek, ‘Geology of Harrison and Monona Counties,” Jowa Geol. Surv., Vol. XX (1909), pp. 371-75. 3W. C. Alden and M. M. Leighton, ‘‘The Iowan Drift,” Iowa Geol. Surv., Vol. XXVI (1915), p. or. O47 x. GEORGE F. KAY AND J. NEWTON PEARCE thought them to be related to loess. Until recently McGee was the only geologist who had stated definitely that the material which is now called gumbotil is the product of weathering of drift. FIELD STUDIES OF GUMBOTILS AND RELATED MATERIALS The conclusion which was presented by Kay in his paper in Science is: ; .... that the gumbotil is the result chiefly of the chemical weathering of drift was reached only after the field relations of gumbotil had been studied carefully, and detailed chemical analyses of Nebraskan gumbotil, Kansan gumbotil, Illinoian gumbotil, and the glacial tills underlying these gumbotils had been made. The field relations of Kansan gumbotil to the underlying Kansan till have been already briefly described.* The Kansan gumbotil, there called super-Kansan gumbo, reaches a maximum thickness of more than twenty feet, and is limited to tabular divides and other remnants of a gumbotil plain which, before it was affected by erosion, was as extensive, apparently, as the original Kansan drift plain. This gumbotil occupies a definite topographic position, and where it is exposed in railroad cuts it is seen to lie horizontally in the cut and not to conform to the surface slopes which have been developed by erosion. The gumbotil is dense, sticky, and very slippery when wet, but is hard and very tenacious » when dry. It is usually dull gray to drab in color; in places the gray color is mottled with brown and reddish tints. It is leached, but in many places it contains lime concretions. The dry surfaces of the exposures of gumbotil are distinctly checked by sun cracks. It contains only a few small, scattered pebbles, which consist predominantly of quartz and chert and subordinately of crystallines and quartzites. A striking feature of the quartz and chert pebbles is their remarkably smooth surfaces. The gumbotil grades down- ward into yellowish to chocolate-colored till, in many places with numerous pebbles, few if any of which are calcareous. This oxidized and non-calcareous till, in turn, merges into unleached George F. Kay, ‘‘Some Features of the Kansan Drift in Southern Iowa,” Bull., Geol. Soc. Amer., Vol. XXVII, pp. 115-17; reprinted in Iowa Geol. Surv., Vol. XXV, pp- 612-15. THE ORIGIN OF GUMBOTIL 95 till, oxidized yellowish for several feet vertically, below which is the normal, unoxidized and unleached, dark grayish to bluish-black Kansan till. An impressive feature of the unleached, oxidized till is the presence of numerous concretions of calcium carbonate, the lime of the concretions having been dissolved in connection _ with the formation of the overlying gumbotil and leached till, carried downward and later precipitated. Some sections showing the relations of Kansan gumbotil to under- lying Kansan drift—The following sections are given as typical of many sections that have been studied at widely separated places in the Kansan-drift areas of Iowa. They are intended to show the intimate field relations of the Kansan gumbotil to the under- ' lying till. | Section in cut on the Chicago, Milwaukee & St. Paul Railway about one mile east of Foster Station, in the southeast corner of Monroe County, Lowa: Feet Inches FeSO DIAGKMOLOUS.. of). ccfew cs ais le cases 6 5 2 4. Loesslike clay, chocolate-colored, leached.... 1 -6 3. Loesslike clay, light-colored, grayish; on dry surface looks like gumbotil; has chocolate- colored stains; sticky when wet; contains a few small siliceous pebbles; leached......... 5 6 2. Gumbotil (Kansan), gray-colored, in lower part chocolate-colored; few pebbles; starch- liketracture when wet; leached: .:.:........ 12 1. Glacial till (Kansan), brown color, with very irregular patches of gray-colored till resem- bling gumbotil; dry surface of the till is brownish yellow; damp surface is chocolate- colored; few pebbles; leached to base of cut 5 Section in cut on the Chicago, Burlington & Quincy Railway at mile 372, one mile west of Murray Station, Clarke County, Iowa: Feet 4. Loesslike clay, gray to pale-yellowish color, with irregular lines of brown on dry surface; when damp it is grayish with mottling of yellow to brown colors; stands vertically, uppettewsteeu mealye. om... nt - deus stage 15 96 GEORGE F. KAY AND J. NEWTON PEARCE Feet 3. Gumbotil (Kansan), gray to drab in color, sticky when wet, hard and tenacious when dry; contains a few siliceous pebbles; leached 11 2. Glacial till (Kansan), oxidized and leached.. 4 1. Glacial till (Kansan), oxidized and unleached; has many lime concretions................. II Section in cut on Chicago, Milwaukee & St. Paul Railway no the divide about three miles west of Templeton, Carroll County, Iowa: Feet Inches 4. Loess Butt-coloredleachedyy temas: ae cea 15 Buti-colored, unleached)). 7 3. a, 42-47 ite) 3. Gumbotil (Kansan), gray to dark-drab to chocolate-colored, upper few feet reddish, a fewasmallesiiceousipepolesaemane? es eee 20 6 2. Glacial till (Kansan), oxidized yellow to buff, -leached, closely related to No. 3............ " 1. Glacial till (Kansan), oxidized, unleached; many calcareous concretions..............+ 8 Section in cut on Santa Fe Railway east of New Boston, Lee County, Iowa: Feet 4. Loesslike clay, top 2 feet very light gray; . below, yellow to light-brown on dry surface; when freshly cut into, more chocolate-colored; ajolnticlayce Stacdes: 1ntou Nossa eee ee 12 3. Gumbotil (Kansan), typical; gray on dry surface and has a checked appearance; when freshly cut into, has a more drab color; very sticky; contains some spots of brown; con- tains small siliceous pebbles; leached; grades IMtOANO M9: Hatae pintal er ee Net oh ey tener eer 12 2. Glacial till (Kansan), oxidized and leached; contains patches of gray similar to the gum- botil; many pebbles; grades into No. 1.... 5 1. Glacial till (Kansan), oxidized and unleached; contains many pebbles and small bowlders; many calcareous concretions; to the bottom of the cut, exposed:...7-ce ee Reeth cee cea 27 THE ORIGIN OF GUMBOTIL 97 In all of these sections the zone of oxidized and leached till beneath the gumbotil is narrow. A study of thirty-five sections widely separated as to location shows that in eighteen of them the zone of oxidized and leached Kansan till beneath the Kansan gumbotil is 5 feet. In twelve of them it is 5 feet 6 inches; in the - remaining sections the zone is somewhat more than 5 feet 6 inches or slightly less than 5 feet. The uniform thickness of the leached zone is impressive. The thickness of the oxidized, unleached zone of Kansan till is about 4o feet. The distribution of Kansan gumbotil in Iowa.—The relations of Kansan gumbotil to the underlying Kansan till have been seen at scores of places in southern Iowa and at many places in other parts of the state. In fact the Kansan gumbotil has been studied in every county of three tiers of counties in southern Iowa as well as in many of the counties which are farther north.* Moreover, within the Iowan-drift area the Kansan gumbotil has been found beneath Iowan drift at numerous places.? It will be of interest to state that the Kansan gumbotil is now known at a sufficient number of places in Iowa to permit the restoration of the Kansan gumbotil plain, that is, the original plain surface of the weathered Kansan till, as it was in Iowa before any great erosion was accom- plished. Some sections showing the relations of Nebraskan gumbotil to underlying Nebraskan drift—tThe field relations of the Nebraskan gumbotil to the underlying Nebraskan till are similar to the rela- tions that have been described as existing between the Kansan gumbotil and the underlying Kansan till. The two tills, the Nebraskan and the Kansan, are much alike lithologically and both appear to have undergone similar changes under similar conditions. Below the Nebraskan gumbotil there is, as in the case of the Kansan gumbotil, a narrow zone of leached, oxidized till which grades downward into unleached, oxidized till with many concretions. * George F. Kay, ‘Pleistocene Deposits between Manilla in Crawford County and Coon Rapids in Carroll County, Iowa,” Jowa Geol. Surv., Vol. XXVI (191 Hi) PPIs ask 2 W.C. Alden and M. M. Leighton, ‘‘The Iowan Drift, a Review of the Evidences ~~ of the Iowan Stage of Glaciation,” Zowa Geol. Surv., Vol. XXVI (1917), pp. 92-109. 98 GEORGE F. KAY AND J~NEWTON PEARCE A good section to show the field relations of the Nebraskan gumbotil to the underlying Nebraskan -till is a railroad cut just east of a viaduct 13 miles west of Manning, Carroll County, Iowa.* From the surface the cut shows loess, Kansan till, soil band, Nebraskan gumbotil, and Nebraskan till. follows: 6. Loess: Leached, yellowish-gray on dry surface; yellowish-brown® to buff-brown on damp surface; no shells or concretions............ Unleached, lighter-colored on dry surface than the leached loess, and when damp is buff with gray streaks; contains shells ands Coneretions'y aware a aha tere . Glacial till (Kansan), yellow, unleached, with calcareous concretions; numerous pebbles including granites, quartzites, etc. Below the oxidized, unleached till is gray till with a few pebbles. It is gumbotil-like, but effer- vesces freely. It was probably picked up from the gumbotil zone below.............. . Soil band containing carbonaceous material. . . Gumbotil (Nebraskan), gray to drab-colored, few pebbles. The upper 6 feet is fine grained, gray, and is less sticky and gumbotil-like than the lower 7 feet, which is leached but has some calcareous concretions................ . Glacial till (Nebraskan), oxidized, apparently leached, but has calcareous concretions, upon which are films of manganese dioxide....... . Glacial till (Nebraskan), unleached, oxidized, light-yellowish color on dry surface, mottled brownish with gray when damp; many cal- careous concretions, especially in upper 10 The section is as Feet Inches 13 In Taylor County, Iowa, at a stream crossing just west of Conway Station on the Chicago, Burlington & Quincy Railway, is an exposure at which the following section was observed: t George F. Kay, ‘Pleistocene Deposits between Manilla in Crawford County and Coon Rapids in Carroll County, Iowa,” Iowa Geol. Surv., Vol. XXVI (z917), pee2ee THE ORIGIN OF GUMBOTIL 99 Feet Inches PIG CSSINEON CLAY I Accel ower Geis wales oats we vi eo I 3. Gumbotil (Nebraskan), grayish to drab on dry surface; when damp, is grayish to brown- ish; a few siliceous pebbles; leached........ II 2. Glacial till (Nebraskan), oxidized and leached, many disintegrated bowlders............... 4 6 1. Glacial till (Nebraskan), oxidized, unleached, EGNCTEEIONSVEXPOSEMs. hones see cree sees s B Near this exposure Kansan till overlies the Nebraskan gumbotil. Another interesting cut which shows Nebraskan gumbotil and underlying drift is along the wagon road west of Osceola, Clarke County, southwest corner of section 13, Ward Township, in front of a schoolhouse. The section is as follows: Feet Inches 5. Loesslike clay, gray to light-yellow......... 8 4. Glacial till (Kansan), oxidized, leached, fer- RE LLOPZONECIONPLOPS: siiais stents ales" I 3. Gumbotil (Kansan), drab to dark color, starchlike fracture, some calcareous concre- tions; few pebbles; leached; gradesinto No.2 8 6 2. Glacial till (Kansan), oxidized, pebbles and Howilders leaChedey se chico: cio cs aes ees oe 5 1. Glacial till (Kansan), oxidized, unleached, many concretions, breaks with irregular frac- BULEEOXPOSCU eer sabre son a heise % arate w aus o 12 x The transition zone between gumbotil and the base of the oxidized leached drift—Many interesting sections might be given to show disintegrated and decomposed bowlders in the transition zone between gumbotil and the base of underlying leached and oxidized till. For example, on the east-west wagon road in section 1, Otter Creek Township, Lucas County, Iowa, there is in the transi- tion zone between the Kansan gumbotil and the base of the leached, oxidized Kansan till a granite bowlder with dimensions of 4 X 2 feet on the slope. It is so thoroughly weathered that its outlines are discerned only with difficulty. Again, in a cut through the upland + mile north of Forbush on the interurban railway between Center- ville and Moravia, Appanoose County, Iowa, there is in the transi- tion zone between Kansan gumbotil and the base of the oxidized, leached Kansan till a completely disintegrated granite bowlder 5 feet long by 2 feet wide as exposed on the surface. The pebble content of gumbotil and underlying drift_—The gumbo- tils and underlying tills were studied also with regard to their pebble content to ascertain whether or not additional evidence could be obtained to strengthen the view that the gumbotils are the result of changes in what was originally till. 102 GEORGE F. KAY AND J. NEWTON PEARCE The average pebble content of Kansan gumbotil gained from eight analyses of pebbles made in widely separated areas in Iowa is as follows: Percentage Quarta. Fo" ae Pee Metin ce gah ae nee ances 1 Chert, flint ete. arr Rig ov alte Sn ee eS Quartzite) 0 Bio eR eee ea ty Oe ees eR Granite . : : : 7.8 Basalt and greenstone TA rere cal UMN he ih oe) 18) Feldspar 1.0 Sandstone On5 It will be seen that more than 87 per cent of the pebbles are of siliceous material; the highest percentage of siliceous pebbles shown by any of the exposures of Kansan gumbotil subjected to study was 98 per cent, the lowest 75 per cent. The average pebble content of the leached and oxidized Kansan till beneath the Kansan gumbotil is as follows: Percentage Quartz gc! OR ae. nS ee ee eS Chert, fmt eter cc 0" 22 es aire a Se OPE Quartzite 4 se ee oe a PO eee Ts Ne ee RT Giramiterh, Pe a PAE ee acl as ee OO Basalt and greenstone em Tec arenes eel. 2, 5 Feldspar. TO Felsite 7.0 Sandstone ae kim) iO Ghalet oie Medel be Vc eekee | SERS os nea) ah Ae OO Quartz porphyry . Ons Schist : 2n3 Gneiss ong The average content of siliceous pebbles is here only about 42 per cent, compared with 87 per cent in the Kansan gumbotil; the highest siliceous content was about 55 per cent, the lowest about 25 per cent. A study of the pebble content of the unleached and oxidized Kansan till beneath the leached and oxidized Kansan till gave an average result as follows, seven analyses being used: Percentage Quartz: 6) ke see a Gee aie ee er ‘ Chert, fling €lG..o5. 0 25) hatte ee et a Quartaite yo oe ee eR Were tee ee THE ORIGIN OF GUMBOTIL 103 Percentage MES OUCH ier eetAr ee. ag eee. Oro GTEDTDOUE yi as a a mR: 27 etc Basalt and greenstone LGR tp set ath oe ER elsvier pate can a5 re fe SOA By ee ae Sea a O Sandstone SRR As oh oe on 2k, EO SIZtCePaaaee Une eo) ; BSir fee elk ie ie eS 0 yt Schist 0.4 It is clearly seen that the content of siliceous pebbles of the un- leached and oxidized Kansan till is considerably less than that of the leached and oxidized Kansan till, the average siliceous content of the former being less than 20 per cent. It is of interest to note that the kinds of pebbles found in the different zones of material are much alike. Of course limestone is in the unleached zone only. Similar results were obtained when the Nebraskan and Illinoian gumbotils and their respective underlying tills were studied. For instance, the average pebble content obtained from several analyses of Nebraskan gumbotil is as follows: Percentage Quartz . ‘ ; : : 3 : : : : WN BOR75 Heres melCurmm ne ne Oo ee 2 pols CRORIR EAS SiGe OT a alee eee ee MS [0 PS Granite ME MMe daira Scr eta ck) Sak oe A gaol 2G ibasdiéramcdvoreenstones (Sea Mod coe ete A) SEO Feldspar. eae 2 Ae, ‘ac P ; ie ene 1G eIsicc mua ete eee se i OEO The content of siliceous pebbles is here more than 78 per cent; none of the Nebraskan gumbotil examined gave less than 72 per cent, and the highest gave 88 per cent. The studies of leached and oxidized Nebraskan till gave about 38 per cent of siliceous pebbles, and the unleached and oxidized Nebraskan till gave about 15 per cent of siliceous pebbles. Analyses of the pebbles of Illinoian gumbotil gave an average result as follows: Percentage Quartz ; ‘ ; 2 ; : : : 5 », 43 herpwuimtectOnes ss cy Oe oS OL Re eae Quartzite ‘ot 10) eee Granite . eae Fe oh ets. : 5 OYE ga Meh eS a ene Basalt and greenstone SEE AGE Me Nee 2 Sandstone . : : : : ; ; j ; . 2 104 GEORGE F. KAY AND J. NEWTON PEARCE Analyses of pebbles of leached and oxidized Illinoian till beneath Illinoian gumbotil gave results as follows: _ Percentage Quartz oh) See ae ce ae ora eee ge aS Chert,, flint "ete 207 yee ek ie Ses ee een RS Quartzite Granite . i : Ph Yas : . oy a Basalt and crenerones Han cater mae fics, foe, TE Peldspater iia siiihee tis lene Wik ert we engi 8G eine ee eae 2 Pelsite: "23 ) ie nee ico ee READ oak ee er 5 Sandstone. 22 oq cas he thy eee le peer a ae a I These analyses show also that the percentage of siliceous pebbles in the gumbotil is much higher than that in the oxidized and leached till which underlies it. The sizes of pebbles in gumbotil and underlying drifi—When pebbles from the gumbotils and underlying tills were being taken in the field the only purpose in mind was to ascertain the percentage content of the different constituents. Later, when these studies were being made in the laboratory, it became evident that the pebbles might also be used in estimating the relative sizes of the pebbles in the different horizons, and in determining the shapes of the pebbles. One hundred pebbles collected from the Nebraskan gumbotil at one locality had dimensions as follows: largest pebble 2.4X1.4XI cm., smallest pebble 2mm., and average pebble 8X5mm. ‘The shapes of these pebbles were subangular to spheroidal. The unleached and oxidized Nebraskan drift beneath the Nebraskan gumbotil had pebbles with dimensions as follows: largest pebble 6.25X4.5X1.75 cm., smallest pebble 3X4mm., and average pebble 1.75X1.5X1cem. The pebbles were chiefly flat and subangular; some were slightly rounded. The largest pebble in the Nebraskan gumbotil from another locality was 10X75 mm., the smallest 1.5X2mm., and the average 3 mm.; the shapes were subangular to spheroidal. Here the underlying Nebraskan till had pebbles, the largest pebble of which was 5.5X3.5X3cm., the smallest pebble 3X2 mm., and the average pebble 1. Be wis oo, Ihlave snnpee of the pebbles were subangular to more or less flat. THE ORIGIN OF GUMBOTIL 105 Similar studies made of pebbles taken from Kansan gumbotil and from underlying Kansan drift gave results as follows: largest pebble in gumbotil 3.2 cm., smallest pebble 3 mm., and average pebble 7mm. In the underlying drift there are many pebbles 10-12 cm. in diameter, a few more than 3 cm. in diameter; the smallest pebble seen was 7 mm., and the average of one hundred pebbles collected was about 1.8cm. ‘The shapes of the pebbles were similar to corresponding horizons in the Nebraskan materials. CHEMICAL STUDIES OF GUMBOTIL AND RELATED MATERIALS In addition to a study of the field relations of Nebraskan gumbotil, Kansan gumbotil, and Illnoian gumbotil, the tills which underlie these gumbotils, and laboratory studies of the physical properties of these materials, there were made detailed chemi- cal analyses of gumbotils and related materials. The speci- mens were taken from exposures which had been studied carefully in the field, and the materials selected were thought to represent satisfactorily the compositions of the zones from which they were taken. The analyses were made from i-gm. samples of fine material which had been separated carefully from pebbles and concretions. Only the material which could be sifted through a “‘twenty-mesh”’ copper-gauze sieve was pulverized and subjected to chemical analysis. Accurate determinations were limited to the oxides of aluminum, silicon, iron, calcium, and magnesium, since deductions as to the nature of the chemical processes involved in the transformation of the drift can be made only upon the pro- portions of these less mobile constituents now present. The analyses were made in strict accord with the preferred methods and the recommendations prescribed by Hildebrand.* Before referring in detail to the kinds of materials which were analyzed, the localities from which they were taken, and the results of the analyses, it seems well to discuss somewhat fully some of the geo-physico-chemical factors which need to be under- stood in order to interpret correctly whether or not a material such as gumbotil is the product of weathering of till. Chemical evi- dence will be presented to support the field evidence that the * Hildebrand, ‘‘The Analysis of Silicate and Carbonate Rocks,” U.S. Geol. Surv., Bulletin 422. 106 GEORGE F. KAY AND J. NEWTON PEARCE gumbotils and underlying oxidized and leached tills have been formed by chemical weathering and leaching of till which was originally unoxidized and unleached. The abundant field evidence supporting this theory has been presented. Emphasis has been put upon the gradation of gum- botil into underlying till, the variations in the sizes of pebbles in the related zones, and the presence of remnants of thoroughly disintegrated and decomposed bowlders in the transition zone between gumbotil and oxidized and unleached till. There have been profound changes involving chemical processes which operated during immense lengths of time, and which occurred long ages ago. These chemical processes are subject to a few definite, general physical laws of nature which are independent of time or place. The laws of stress and strain, of the degradation of energy, of hydrolysis, of mass-action, or of solution in general are as lasting as the universe itself. The réle of water in geochemical changes——The dominant factor in all of these geo-physico-chemical changes is water, more especially the aerated water. When the rain falls upon the ground one part, the “run-off,’™ flows over the surface and escapes by way of the natural drainage channels. It is this form which pro- duces erosion. A second part, the ‘‘fly-off,” immediately evapo- rates into the air, while the third part, the “cut-off,” penetrating the soil by way of the soil interstices, flows downward under the influence of gravity. Of these the cut-off water is the only form which is directly effective in geochemical transformations. It moves through the soil and its substrata with comparative rapidity, reappearing elsewhere as seepage water or as springs. The rain and surface waters contain dissolved oxygen, nitrogen, and carbon dioxide, each in proportion to its partial pressure in the atmosphere.’ The chemical and physical processes which are con- tinually taking place below the surface involve the absorption and formation of carbon dioxide and the disappearance of oxygen and nitrogen. These gases impart to the soil an atmosphere, and their concentrations in the soil solution follow more or less slowly the barometric changes above the surface. The soil bacteria and 1 This terminology was proposed by McGee. THE ORIGIN OF GUMBOTIL 107 other lower forms of life are likewise producing change and under- going change continually. According to Cameron’ the water within the soil is in reality of two kinds, namely, ‘“‘film”’ water and ‘‘free’’ water. When a relatively small quantity of water is added to an absolutely dry soil or powdered solid there is some shrinkage in the apparent volume of the solid; the water spreads over the surface in the form ofafilm. With further addition of water the apparent volume of the solid material increases until a maximum is reached. The optimum water content which gives the maximum volume is a definite, critical, physical, characteristic property for a given soil or solid. A further addition of water will not increase the thickness of the soil film but will produce free-water in the soil interstices. These two kinds of water play an important réle. The film- water is tenaciously held by the soil and subsoil particles. In dry seasons it is practically a saturated solution of the dissolved rock and soil-‘materials. When the surface of the ground becomes flooded, as in wet seasons or during heavy rains, the downward- moving free-water extracts and carries away a part of the mineral content of the film solution. Obviously the proportions of water- soluble materials in soils containing moving free-water should be, and are, less than those in soils containing only film-water. This was conclusively shown by Hilgard.* In the humid regions there is a greater amount of rainfall, hence a greater amount of downward- moving free-water, consequently a greater amount of leaching. Once the free-water is removed the saturation of the film-water repeats. In this film-water are the dissolved rock materials, the carbon dioxide, the oxygen, the nitrogen, the soluble humous material, and the soil bacteria. Between these are evolved all of the processes leading to the disintegration of the rock material and the formation of the soil and its substrata. Chemical nature of glacial materials—The rock materials transported by the glaciers consisted chiefly of silicates, quartz, some clays, and other previously weathered materials. The com- plex silicates are salts of a very weak acid—silicic acid, with various Ra Cameron, Jour. Phys. Chem., Vol. XIV (1910), p. 340. 2 See Merrill’s Rock Weathering and Soils, p. 368. 108 GEORGE F. KAY AND J. NEWTON PEARCE base-forming metals, like sodium, potassium, calcium, magnesium, iron, and aluminium. ‘These silicates are only very slightly soluble in water. While the amount dissolved may be exceedingly slight, it is nevertheless sufficient for the purposes here involved, if the time allowed is sufficiently long. For a given rock material the solubility, and hence the speed, of the weathering process increases with increasing fineness of the particles. Hence the fineness of the glacier flour renders it peculiarly suitable for rapid chemical conversion. Like all salts of weak acids with strong bases, these silicates when dissolved react chemically with water, that is, hydrolyze, to form the free more or less ionized bases—the soluble hydroxides of sodium and potassium, the less soluble hydroxides of calcium and magnesium, the relatively insoluble hydroxide of iron, and either the free un-ionized silicic acid or some-simple silicates. These simpler silicates continue to hydrolyze, if the reaction products are removed. ‘There results the liberation of still other bases and in the end still simpler silicates, possibly kaolin, or even silica as quartz or sand. . If the reaction products are not removed by leaching, the dis- solved materials soon attain a state of solution equilibrium. Under these conditions the decomposition products of one rock material may react with those of another to form more or less complex silicates of a secondary origin. Let the saturated solution be removed and fresh water added, the various solution equilibria are disturbed and the solution processes begin again. The solution and subsequent hydrolysis of rock matertal, and the chemical reactions imvolved—In the soil solution thus formed the dissolved materials will at the proper concentrations react with the carbon dioxide of the soil atmosphere to form the soluble, easily hydrolyzable carbonates of sodium and potassium and the slightly soluble carbonates of calcium and magnesium. ‘These slightly soluble carbonates crystallize out as calcareous concretions, so frequently found in the clay subsoils. In the presence of an excess of carbon dioxide these insoluble carbonates pass into the soluble acid-carbonates and are leached away by the downward- moving free-water to lower depths, where they are again deposited in the form of concretions. THE ORIGIN OF GUMBOTIL 10g Ferrous silicates upon hydrolysis give ferrous ions. These may react with other ions of the soil solution to form the slightly soluble ferrous hydroxide or carbonate, or, what is more probable still, they may be immediately oxidized to ferric ions by the dis- solved oxygen and then precipitated as the insoluble hydrated ferric oxide or basic carbonate. Some may raise the question as to the possibility of the oxida- tion of the iron below the surface and out of contact with the air above. It must not be forgotten that by virtue of diffusion not only oxygen but also other dissolved gases tend to go just as far as the water does. Except at points where organic matter is undergoing decay, it is very doubtful if more than minimal traces of secondary ferrous compounds exist within or beneath the soil. Crystalloids and colloids—The resulting so-called weathering products may be grouped into two great general classes, namely, crystalloids and colloids. Crystalloids include all of the soluble acids, bases, and salts. They are simple in structure, that is, they exist in the dissolved state either as molecules or asions. They are characterized by a relatively high diffusion speed and by the power to pass, though more or less slowly, through colloidal mem- branes. Nor is the colloidal membrane lacking in the soil; the clay itself may be considered as such a membrane. Generally speaking, the term colloid is applied to all glue- like, gelatinous, amorphous substances. Strictly speaking, colloids represent suspensions of matter in an extremely fine state of subdivision, the suspended particles having diameters varying from 1 4 to 100 wu. The properties of colloids are primarily surface properties. The extent of surface development, and hence the magnification of specific properties, such as solubility, adsorption power, etc., may be seen from the following illustration: A centi- meter cube of any solid substance, say platinum, exposes a surface of 6 square centimeters. Let this cube be divided successively and decimally to the dimensions of colloid particles. The total surface then exposed by the platinum will vary between 60 and 6,000 square meters. Some colloidal properties involved in this problem.—Since colloids play an important role in the formation of soil strata and since they 110 GEORGE F. KAY AND J. NEWTON PEARCE impart to these strata many of their most important properties, it will be necessary to mention in detail some of the more important ‘colloid properties and phenomena. Colloids are divided into two general classes, namely, suspen- -soids and emulsoids. Briefly stated, suspensoids are suspensions of solid particles, chiefly inorganic, in a fluid medium; emulsoids are suspensions of fluid or semi-fluid particles. The emulsoids ‘found in the soil are chiefly of organic origin, resulting from the -secretions of animals, the exudations of plant roots, the humus, and other products of decaying organic matter brought about ‘through the assistance of bacteria and fungi. _ Colloids are also classified as reversible and irreversible. Most -of the suspension colloids when desiccated, sometimes when frozen, or when in the presence of electrolytes, coagulate into a solid or semi-solid water-insoluble precipitate. When this solid is placed in water it does not again pass into suspension. It is therefore said to be irreversible. ‘To this class belong the hydrated ferric oxide, the gelatinous silicic acid, the gelatinous, hydrated aluminum silicates, the clays, kaolin, etc. In rare cases one may find alumi- num hydroxide.* Suspended colloid particles are either positively or negatively charged. Thus silicic acid, kaolin, and clay particles are nega- tive; the basic hydrated ferric oxide is positive. These charged particles are precipitated by electrolytes, and it has been found that the precipitating power of the electrolyte is specifically a property of the ion bearing a charge opposite in sign to that of the particle. Further, the precipitating effect is greatest for those ions carrying the greatest number of charges. Factors determining the stability of colloidal clay suspensions.— The effective properties of any colloid suspension depend upon its stability—its power to exist in the colloidally suspended state. The stability is also one factor in the slow transportation of the colloid particles through the soil capillaries. This stability depends not only upon the absence of precipitating ions but also upon the potential difference between the charged particles and the oppo- : Cameron-and Bell, ‘The Mineral Constituents of Soil Solution,” U.S. Dept. of Agric., Bull. 30 (1905), p. 22. THE ORIGIN OF GUMBOTIL Tit sitely charged solvent. The stability is greatest when this potential difference is greatest; its instability, or its tendency to coagulate, is greatest when this difference approaches zero. Electrolytic coagulation or precipitation is preceded by the electrical adsorption of ions by the oppositely charged colloid particles. ‘The precipitate carries with it the adsorbed ion or its salt, which in its adsorbed state is more or less difficultly removed by washing. Thus certain salts like those of potassium and ammonium are specifically and tenaciously held by the soil and clays, while the more toxic, less firmly adsorbed sodium salts are leached away. When present in traces the singly charged H* and OH™ ions not only do not coagulate colloidal material but may even increase the stability of similarly charged colloid particles. In the hydrolytic decomposition of the alkaline silicates free OH~ ions are formed. ‘These tend to increase the negative potential, likewise the stability of the negative colloids. Thus in the alkaline- soil solutions silicic acid and the colloidal hydrated silicates are kept to a slight extent, at least, in a state of pseudo-solution. Under the influence of the free carbonic acid and of mineral acids formed by adsorption cleavage of the dissolved salts there may be, as in acid soils, a slight excess of H* ions. These stabilize those colloidal “‘sols’’ containing positively charged particles. Thus colloidal hydrated ferric oxide is, to a very slight extent at least, rendered capable of transportation by the downward-moving free-water. In the initial stages of the leaching of an original silicate material, where the solution is distinctly alkaline, only the soluble salts and the transportable negative colloidal silicic acid are removed by leaching. Not until the alkalinity has disappeared would it be possible for the positively charged colloidal ferric hydroxide to exist in suspension. Iron in the colloidal form would be, therefore, almost the last colloidal material to undergo leaching. A soil or clay colloid when once coagulated may again pass into the soluble hydrosol condition.. Numerous experiments have been made dealing with this particular problem. Van Bemmelen has found that when finely divided clay is washed upon a filter the loosely bound coagulating salts are washed away. Upon further iL? GEORGE F. KAY AND J. NEWTON PEARCE washing, the clay becomes still more finely divided and finally passes through the filter, giving a turbid non-settling suspension. On adding a small amount of an electrolyte the milk-white liquid . coagulatés and settles. Upon washing again another point is reached at which the particles become infinitely fine and pass through. So it is in clay soils: an excess of water percolating downward removes the excess of coagulating electrolyte from the leached clay. ‘This permits first a swelling of the reversible col-. loidal material and finally, to a slight extent at least, the gradual re-formation of the colloidal “‘sol.’”’ The suspended particles are thus permitted to pass slowly downward, where they are again coagulated at some lower level. _ The inorganic colloids of soils and clays exhibit a marked tendency to adsorb upon their surfaces the various organic emul- soids formed from plant and animal débris. The humus is full of these. The adsorbed emulsoidal material forms an oil-like film about the suspended particles and imparts to them its own reversibility and stability. Hence when a mixture of the emul- soid and suspensoid materials are evaporated to dryness and the dry material is again placed in water the whole mass again passes into colloidal suspension. Furthermore, if emulsoidal material of any sort is added to a coagulated hydrogel, such as clay, the emulsoid possesses the power to peptizate or deflocculate the clay hydrogel, thus rendering it capable of colloidal suspension. By their reversible and protective influence humous materials hinder the coagulation of clay colloids; by their deflocculating influence they tend to make the hard, dry, sun-baked clays again reversible. The terms humus and humic acid have been mentioned. The latter is a very complex substance of doubtful composition; it is an acid and possesses a colloidal nature. It dissolves in 8,337 parts of water at 6°. Its ammonium and magnesium salts are rather easily soluble; calcium humate dissolves in 3,125 parts of water, while the least soluble ferric humate dissolves in 5,000 parts. The humic acids are solvents for silica. Humic acid has the property* t Julien, ‘“‘On the Geological Action of the Humic Acids,” Proc. Amer. Assoc. Adv. Sci. (1879), pp. 311-410. —— THE ORIGIN OF GUMBOTIL TE} of gluing together vegetable earths into layers impervious to water. Their action consists mainly in the removal of calcium, magnesium, and iron, which are again precipitated at the lower limit of action, either by soluble salts, by exchange of bases, or by loss of water. The precipitated organic humus is finally oxidized and disappears, depositing the base metals as hydroxides or carbonates. Summary of the mineral constituents of the soil solution—tin summarizing, the mineral matter of the soil solution may be divided into two classes. The more easily diffusible are those comprising the soluble alkali salts, the soluble acid-carbonates of calcium and magnesium, the slightly soluble ferrous compounds, the ferric humates, and the semi-colloidal sodium and potassium silicates. The less easily diffusible are the colloidal gelatinous silicic acid, the gelatinous hydrated silicates, and the colloidal hydrated ferric oxide. The solvent action of the alkaline-soil solution and of the humic acids, aided by the abrasive effects of the earth’s displacements, slowly but surely transform the quartz pebbles into colloidal silica. Under the influence of the decaying organic matter the ferric compounds are reduced, temporarily at least, to ferrous compounds. While the existence of ferrous com- pounds in contact with the oxygenated soil atmosphere must obviously. be a short one, the alternate oxidation and reduction permit the slow downward transportation of iron. The decompo- sition of the original complex aluminum silicates leads ultimately to the formation of the colloidal hydrated aluminum silicates. These are the most complex, most resistant, and the least soluble of all of the decomposition products produced by the disintegration of silicate rocks. Hence in the leaching of the so-called weathering products of the original glacial till one should expect to find a gradual relative increase in the proportion of the soluble diffusible materials from the surface downward. On the contrary, conditions permitting, a gradual decrease in the proportion of alumina should be observed. This is exactly what is found from a study of the results of the chemical analyses of a complete series of strata in any single complete cut. II4 GEORGE F. KAY AND J. NEWTON PEARCE The kinds of materials analyzed and the localities from which they were taken.—The kinds of materials analyzed and the localities from which the materials were taken are as follows: A. Kansan gumbotil, and oxidized and leached Kansan till, from cut on the Chicago, Milwaukee & St. Paul Railway, about one mile east of Foster in the southeast corner of Monroe County, Iowa. B. Kansan gumbotil, oxidized and leached Kansan till, and oxidized and unleached Kansan till, from cut on the Chicago, Milwaukee & St. Paul Railway, at mile 372, one mile west of Murray, Clarke County, Iowa. C. Nebraskan gumbotil, Nebraskan oxidized and leached till, and Nebraskan oxidized and unleached till, from cut on the Chicago, Milwaukee & St. Paul Railway, one and one-half miles west of Manning, in the southwest one- quarter of section 18, Warren Township, Carroll County. D. Illinoian gumbotil, oxidized and leached Illinoian till, and oxidized and unleached Illinoian till from bluff north of Fort Madison, Lee County, Iowa. The complete sections at each of the foregoing localities have already been given in this paper, but it seems well to bring them together here in relation to a discussion of the chemical analyses of materials: A. Section in cut on the Chicago, Milwaukee & St. Paul Railway, about one mile east of Foster, in the southeast corner of Monroe County, Iowa: Feet Inches ge OO, DLA CK sMOROUS eee tiers caret eeterce eects 2 4. Loesslike clay, chocolate-colored, leached.... 1 6 3. Loesslike clay, light-colored, grayish, on dry surface looks like gumbotil, has chocolate- colored stains, sticky when wet, contains a few small siliceous pebbles, leached............ 5 6 2. Gumbotil (Kansan), gray-colored, in lower part chocolate-colored; few pebbles; starchlike fracture swmaemewets led c We Cress ieee er 12 1. Glacial till (Kansan),. brown in color, with very irregular patches of gray-colored till resembling gumbotil; dry surface of the till is brownish-yellow; damp surface is chocolate- colored; few pebbles; leached to base of cut.. 5 THE ORIGIN OF GUMBOTIL IIs B. Section in cut on the Chicago, Milwaukee & St. Paul Railway at mile 372, one mile west of Murray, Clarke County, Iowa: Feet 4. Loesslike clay, gray to pale-yellowish color on dry surface with irregular lines of brown; when damp it is grayish with mottling of yellow to brown colors; stands vertically, MPPELTeWwMeet MICALY. is a. nce s see sas ea Ls 3. Gumbotil (Kansan), gray to drab in color, sticky when wet, hard and tenacious when dry; contains a few siliceous pebbles; leached 11 2. Glacial till (Kansan), oxidized and leached.. 4 1. Glacial till (Kansan), oxidized and unleached; hasmnany, lime concretions. ..... 042. 0.0. 2.4 II C. Section in cut on Chicago, Milwaukee & St. Paul Railway, one and one-half miles west of Manning, in the southwest one-quarter of section 18, Warren Township, Carroll County, Iowa: Feet Inches 6. Loess: Leached, yellowish-gray on dry surface; yellowish-brown to bufi-brown on damp surface; no shells or concretions........... 7 Unleached loess, lighter-colored on dry surface than the leached loess, and when damp it is buff with gray streaks. Contains shells AUNCCONETEMIOUS tenn. oa tee ew apa Sa hela 5 5. Glacial till (Kansan), yellow, unleached, with calcareous concretions; numerous pebbles including granites, quartzites, etc. Below the oxidized, unleached till is gray drift with a few pebbles. It is gumbotil-like but effer- vesces freely. It was probably picked up from the gumbotil zone below.............. 5 4. Soil band containing carbonaceous material. . 4 3. Gumbotil (Nebraskan), gray to drab-colored, few pebbles. The upper six feet is fine- grained, gray, and is less sticky and gumbotil- like than the lower seven feet, which is leached, but has some calcareous concretions......... 13 I16 2 GEORGE F. KAY AND J. NEWTON PEARCE i Feet _ Inches Glacial till (Nebraskan), oxidized, apparently leached, but has calcareous concretions, upon which are films of manganese dioxide....... 2 . Glacial till (Nebraskan), unleached, oxidized, light-yellowish color on dry surface, mottled brownish with gray when damp, many cal- careous concretions, expecially in upper ten D. Section in bluff north of Fort Madison, Lee County, Iowa: 4. Feet Inehes Loess and loesslike clay, grayish-yellow to bui-yellowar raisin one eee qi . Gumbotil (Illinoian), drab to dark color; starchlike fracture; few pebbles; leached.... 4 6 . Glacial till (Illinoian), oxidized, leached.... 6 . Glacial till (Illinoian), oxidized, unleached to DASE Le thir cone ic ap bares cayere te ceri enn ieee oo aee 15 Chemical analyses of the Nebraskan, Kansan, and Illinoian gumbottls and their subsirata.—The analytical data obtained from the chemical analyses of samples taken from localities at “A,” “B,” “C,” “D” have been collected in the following tables. The results are given in two forms, (a) the percentage composition with respect to the less mobile constituents, (6) the parts by weight of these constituents per 100 parts of the more resistant Al,O;. A. Chemical analyses of Kansan gumbotil and oxidized and leached Kansan till from cut on the Chicago, Milwaukee & St. Paul Railway about one mile east of TABLE I Foster, in the southeast corner of Monroe County, Iowa. @) PERCENTAGE COMPOSITION SiOz Fe.0; Al,03 CaO Gumbotil (Kansan)...........| 72.03 4.18 12.27 1.33 Glacial till, oxidized, leached. .| 73.11 4.62 IL.57 1.66 b) Parts PER too Parts AI.0, SiO. FeO; -. CaO MgO Gumbotil (Kansan).......... 587.0 34.10 10.84 18.68 Glacial till, oxidized, leached...) 631.9 39.92 14.35 22.18 THE ORIGIN OF GUMBOTIL amo TABLE II B. Chemical analyses of Kansan gumbotil, oxidized and leached Kansan till, and oxidized and unleached Kansan till, from cut on the Chicago, Milwaukee & St. Paul Railway, at mile 372, one mile west of Murray, Clarke County, Iowa. a) PERCENTAGE COMPOSITION SiO: Fe.0; . A103; CaO MgO Gumbotil (Kansan).......... 70.46 4.17 12.04 T.21 0.55 Glacial till, oxidized, leached. .| 71.84 4.62 10.86 1.29 0.72 Glacial till, oxidized, unleached| 68.56 4.40 IE.13 4.48 °.79 b) Parts PER toc Parts Al,0; SiOz Fe,0; CaO MgO Gumbotil (Kansan)........... 585.2 34.7 TO.05 4.57 Glacial till, oxidized, leached ..| 661.5 42.5 11.87 6.66 ' Glacial till, oxidized, unleached} 616.0 30.5 40.30 7.16 TABLE III C. Chemical analyses of Nebraskan gumbotil, Nebraskan oxidized and leached till, and Nebraskan oxidized and unleached till, from cut on the Chicago, Milwaukee & St. Paul Railway one and one-half miles west of Manning, in the southwest one- quarter of section 18, Warren Township, Carroll County. a@) PERCENTAGE COMPOSITION SiO. Fe.0; ALO; CaO MgO Gumbotil (Nebraskan)........ 71.50 Aes 12.70 1.26 0.93 Glacial till, oxidized, leached...) 66.85 5.92 11.65 B07) 0.78 Glacial till, oxidized, unleached| 66.52 4.80 Ir.18 4.28 1.43 b) Parts PER 100 Parts ALO; SiOz Fe.0; CaO MgO Gumbotil (Nebraskan) ....... 559-7 34.0 9.90 7.31 Glacial till, oxidized, leached...) 573.8 0) Bit Git 6.71 Glacial till, oxidized, unleached| 594.5 42.0 38.30 12.81 Table V shows the comparative results of the chemical analyses of cae Bi GN 50 Deed Discussion of the chemical data.—The three localities “A,” “B,” and “D” show the gumbotil underlying loess or loesslike clay covered by a thin layer of soil. At “C” the Nebraskan gumbotil lies below soil with no intervening loess or loesslike clay. The T18 GEORGE F. KAY AND J. NEWTON PEARCE TABLE IV D. Chemical analyses of Illinoian gumbotil, oxidized and leached Illinoian till, oxidized and unleached Illinoian till, from bluff north of Fort Madison, Lee County, Iowa. a) PERCENTAGE COMPOSITION Si02 Fe.0; Al.O3 CaO MgO Gumbotil (Illinoian).......... 71.07 4.24 14.91 °.79 0.85 Glacial till, oxidized, leached..} 72.24 7.43 11.65 o.61 ©.95 Glacial till, oxidized, unleached| 72.30 3.47 8.59 Ante TDs) b) Parts PER roo Parts Al.0; SiOz Fe.0; CaO MgO Gumbotil (Illinoian).......... 476.0 28.44 5.34 5.68 Glacial till, oxidized, leached. .| 620.1 63.80 5.26 8.22 Glacial till, oxidized, unleached| 841.2 40.39 48.10 T4.95 TABLE V A | B | GS D SiOz Gumibotiles vai elongate eee i eee 72.03 70.46 71.50 71.07 Glacial till, oxidized, leached............ 7B. Hert 71.84 66.85 72.24 Glacial till, oxidized, unleached..........]......... 68.56 66.52 72.30 Fe.0; Gumbotile ye See oh eevee teats eee: 4.18 4.17 4.35 4.24 Glacial till, oxidized, leached............ 4.62 4.62 5.02 TA Glacial oxidized srumaleate tec Bae eee 4.40 4.80 3.47 Al.O; Gumbotih ye eee ee ane ac TD Oy 12.04 12.70 I4.Q1 Glacial till, oxidized, leached............ ene] 10.86 Ir.65 Ir.65 Glacial till? oxidized; wnleached= 22.225. 4| 4... as. II.13 11.18 8.59 CaO Gumbotilis. iso eee ee eee 1.33 1.21 1.26 ©.79 Glacial till, oxidized, leached............ 1.66 1.29 3.67 o.61 Glaciallitilloxidizedsiunleached- es pee ee 4.48 4.28 4.13 MgO Gumbotil.. 6: ajo owe Pecks Leryn cece res 2.20 0.55 0.93 0.85 Glacial till, oxidized, leached............ 2.506 0.72 0.78 0.96 Glacial till, oxidized, unleached.......... nczegee °.79 1.43 1.28 THE ORIGIN OF GUMBOTIL 119 loess and loesslike clay are of great interest but are not being considered except incidentally in this paper. In places these materials are clearly eolian in origin. In other places the loesslike clay may be related closely in origin to the gumbotil. The presence of loess or loesslike clay above the gumbotil might be expected to have had some slight effect upon the present chemical composition of the gumbotil, and in fact may explain the few percentages in the analyses which might seem to contradict the theory proposed. It is assumed that the composition of the flour of the unoxidized and unleached till is now the same as it was when laid down by the glacier. The possibility exists, however, that it may have received a small amount of leached material from above, or that it may have lost to the strata above by capillary flow slight quantities of the more easily diffusible dissolved materials. It is not to be expected that its composition will be similar to that of overlying materials which have been subjected to marked chemical changes, to leaching, or to infiltration. Attention should be called again to the fact that the gumbotils occupy definite topographic positions and that, as a result of erosion subsequent to the formation of the respective gumbotils, the areas of gumbotil are now very limited compared with the extent of the former gumbotil plains. A study of Tables I to IV will bring out several interesting facts. In all of the series here represented the percentage of Al,O, decreases downward from the gumbotil through the oxidized and leached zone. Except in the case of “‘B”’ (Table II), this decrease continues also through the oxidized and unleached zone. Perhaps the most important evidence in favor of the leaching theory is to be gained from a study of the relative proportions of CaO and MgO in the various horizons. In practically every series the proportions of these two constituents show a pronounced increase downward. Apparent contradictions for both might be considered for MgO in Table III and CaO in Table IV. Field relations will show in these instances either that the gumbotil is overlaid by material containing a higher proportion of these con- stituents, or that erosion began before the leaching process in the gumbotil was completed. Assuming that the loess or the loesslike I20 GEORGE F. KAY AND J. NEWTON PEARCE clay is a subsequent formation, it also will have been leached of some of its CaO and MgO. This means a slight increase in the proportions of these two in the stratum below, the leaching of which has not been completed. The leaching process.—The silica and iron are less diffusible and hence less subject to leaching than are the carbonates of calcium and magnesium, and a much longer time is required for complete leaching. In every instance the proportion of iron in the gumbotil is less than it is in the oxidized and leached stratum just below. Except for the case of the Nebraskan (Table III), the proportion of S10, is greater in the oxidized and leached stratum than in the gumbotil. It should be observed that the Nebraskan at locality “C” underlies the Kansan, and the apparent discrepancy may be accounted for in the transfusion of the alkaline silicates from the Kansan into the upper strata of the Nebraskan below. On the basis of parts per too parts of Al,O; not only CaO and MgO but also SiO, show distinct evidence of leaching even in the oxidiz_ | and leached zone, and this is true for practically every series. On the same basis the evidence points to a leaching of the iron into the oxidized and leached zone from the gumbotil, but the time allowed was not sufficient for the subsequent leaching of the iron from the oxidized and leached zone into the one below. This is not surprising, since, as will be shown later, the iron is the last constituent to be leached away. The oxygenated and carbonated water falls upon a uniformly level, more or less uniformly constituted, blue to blue-black drift. Percolating downward, it dissolves a portion of the rock material. Hydrolysis follows, and there are liberated successively the hydrox- ides of sodium or potassium, then of calcium or magnesium, and finally the more or less difficultly soluble hydroxides of ferrous and ferric iron, depending upon the nature of the iron in the original silicate. The ferrous iron throughout the depth penetrated by the dissolved oxygen is immediately oxidized and the deposit assumes the typical iron color of the yellow clays. The calcium and magnesium hydroxides combine with the carbon dioxide of the soil atmosphere to form the insoluble car- _ bonates. These crystallize out as calcareous concretions. The ; THE ORIGIN OF GUMBOTIL 121 soluble alkalies and the alkaline silicates are carried downward by the moving free-water. The negative colloidal silicates and silicic acid are coagulated and thus rendered motionless by the positively charged calcium and magnesium ions. As hydrolysis proceeds the mass of the insoluble material thus formed increases and probably does continue to increase until all of the easily available, hydrolyzable materials are used up or removed. The second stage of the leaching process now follows, Obvi- ously those insoluble substances which are most easily attacked will be the first to be leached away. ‘These are the carbonates of calcium and magnesium. Although only very slightly soluble, the dissolved portions of these combine with the carbonic acid of the soil solution to form the soluble acid carbonates. These are carried downward to lower levels, where in fissures and crevices they again: crystallize as irregular concretions. In this way were formed all of those calcareous concretions which are found in the _jnzed zone. According to the law of mass action, the activity, or the solvent effect, of the carbon dioxide will be greatest at points where its concentration is a maximum. ‘This obviously will be at the upper level of the initially unleached calcareous zone. Owing to its diffusion power some of the carbon dioxide may escape combina- tion at the upper level only to combine at a slightly lower level. Ultimately there will be a lower limit beyond which the carbon dioxide entering from the atmosphere will not penetrate, or its concentration in the soil solution will be too slight to produce any appreciable chemical effect. These two limits of maximum and minimum activity represent the boundaries of the dynamic zone of carbonic acid activity—the oxidized and leached zone. As time goes on the concretions at the upper level disappear, and the levels of maximum and minimum activity move downward simultaneously. This dynamic zone has played an important rdle in all drift transformations. It spreads horizontally like a continuous sheet of more or less uniform thickness. | It is found always directly upon the oxidized and unleached drift and directly below the gum- botil. In the Nebraskan drift it is thin, less than two feet to some- what more than four feet. The oxidized leached zone of the 127, GEORGE F. KAY AND J. NEWTON PEARCE °~ Kansan drift averages about five feet, attaining in a few places a thickness of about seven feet. That of the Illinoian has an average thickness not to exceed six feet. After the leaching of the calcium and magnesium carbonates there follows a third step. When the concentrations of the pre- cipitating calctum and magnesium ions have been reduced below their critical coagulating values new processes occur within the leaching zone. ‘The coagulated iron passes into solution either as colloidal ferric hydroxide by peptization or deflocculation by the emulsoidal organic humous material or through the influence of peptizing ions, or as ferrous compounds by reduction by organic matter. Thus, either by colloidal flow, by alternate reduction and oxidation, or through the medium of its soluble or slightly soluble salts, the iron is leached and slowly passes downward. The silica either in the form of the colloidal gelatinous silicic acid or as the alkaline silicates also moves downward. Likewise, through various peptizing influences the colloidal clays and the simpler colloidal silicates begin to swell and deflocculate. Ulti- mately some of these pass into suspensions of colloidal particles. They are caught also in the downward current and carried by it to lower levels, where they are again coagulated. Only a very slight amount of this kind of materialis leached away. ‘There is left above only the more resistant, less mobile, complex colloidal aluminum silicates. The stratum now forming, deprived of practically all of its sodium and potassium, of most of its calcium and magnesium, and some of its iron and silica, is the present residuum of the whole chemical leaching process. ‘This is the gumbotil. Physical and chemical properties of the gumbotits —The properties of the gumbotil are largely those which one might predict from a knowledge of the colloidal chemistry of clays. Like certain colloids it becomes very hard and tenacious when dry; it swells when wetted and then to some extent passes spontaneously into colloidal sus- pension. It becomes sticky and sometimes so slippery that under the pressure of the earth above it oozes or slides out of thé sides of the hills. It is gray when dry, dark when wet. The char- acteristic color changes of the gumbotil are those imparted to it THE ORIGIN OF GUMBOTIL 123 by the colloidal clays, perhaps the kaolin contained in it, the color of the colloidal material being sufficiently strong to mask the reddish-yellow color of any oxidized iron which may be present. This doubtless is responsible for the belief held by some persons that the iron in the gumbotil is deoxidized or reduced, a condition which could hardly be possible in the presence of the oxygenated soil solution. The chemical analyses of the gumbotils from different drifts and localities show, with respect to certain constituents, a striking similarity. ‘This is*especially true for the iron and silica and, as we might expect, for the calcium. Slight fluctuations may be expected due to differences in the nature of the original rock materi- als, to the amount of rainfall, or to leaching from above. It may be concluded, therefore, that all gumbotils have a common origin— the chemical modification by weathering of glacial till. Similarities of the gumbotils and the adjacent yellow oxidized and leached zones.—Furthermore, the chemical analyses, as arranged in Table V, show a slightly less striking similarity between the gumbotil and the yellow oxidized-leached clay. Naturally one should expect to find a slightly greater concentration of the dif- fusible material in the leached zone. One should expect also to find a slight variation in the proportion of any one constituent between the top and the bottom of any single zone. Each level in any one zone is still slightly unleached with respect to another level close to and above it. The proportions of most of the constituents present in the oxidized-leached and gumbotil zones differ in most of the series by only a few tenths of 1 per cent. When greater deviations occur it can be shown that one or more of the upper strata have been removed before the leaching process was completed. The dis- tinguishing features between these two strata are, therefore, due primarily to differences in ‘the physical properties, and these properties are chiefly. the colloidal properties of the clay itself. It is possible that two forms of the same material are being dealt with, namely, the gumbotil, a highly colloidalized form, and the oxidized-leached clay, the non-colloidalized form, that is, a form which in the presence of electrolytes is incapable of assuming 124 GEORGE F.'KAY AND J. NEWTON PEARCE certain colloidal properties. In fact recent uncompleted work by Mr. L. B. Miller’ gives evidence which not only supports the idea that these two zones differ chiefly in respect to differences in colloidal properties but also strongly confirms the theory that the gumbotil is directly related to the drift. It has already been stated that colloidal properties are primarily properties of the surface. A given colloid material of different degrees of sub- division will adsorb varying amounts of a given substance, and the amounts adsorbed by a given mass of the colloid will be approxi- mately proportional to the specific colloidal surface. Assuming that hygroscopic water adsorbed by clays may be taken as a measure of colloidality and of surface development, Miller has determined the amount of hygroscopic water taken up by each of these clays at 25.° He has found that, beginning with the original drift material the specific surface increases gradually, but at an increasing rate upward to the gumbotil. He has also determined the ‘“‘total-water capacity,” that is, the amount of water per gram which is just sufficient to cause the clays to “run.” This likewise is - greatest for the gumbotil in any drift, and it decreases gradually downward through the lower layers. The high “total-water capacity” of the gumbotil accounts for the ease with which it slides in exposed cuts. SUMMARY The aim of this paper has been to show by field and laboratory evidence that the gumbotils on Nebraskan, Kansan, and IIlinoian glacial tills are the result chiefly of chemical weathering of drift. Thus far no distinctive evidence has been found in Iowa to indicate that the bowlder clay from which gumbotil is thought to have been derived differed to any great extent from typical bowlder clay. In the case of the Iowan and Wisconsin glacial tills, which are too young to have had a gumbotil developed on them, the till at and near the surface does not appear to differ in any important respects from the till which is deeper below the surface. In this connection it should be stated that Mr. E. W. Shaw, as a result of his studies of the Illinoian drift in southern Illinois, the Kansan drift in tL. B. Miller, The Colloidal Properties of Clays. THE ORIGIN OF GUMBOTIL LAS northern Missouri, and other till sheets elsewhere, believes that the upper parts of the tills have been, from the times of deposition of the drifts, somewhat different from the middle and lower portions of the drift." ‘ It should be stated that, although it is believed that gumbotil is essentially the result of chemical weathering of glacial till, it is recognized that wind action, freezing and thawing, burrowing of animals, slope wash, and other factors may have contributed to the formation of these gumbotils. In a subsequent paper attention will be directed to the fact that the gumbotils, on account of their distinctive characters, wide distribution, and topographic positions, are the most satisfactory criteria that have yet been found for differentiating the older drifts. Furthermore, since the gumbotils are the result of changes which took place in interglacial times, they may be considered in relation to the problem of the relative durations of the interglacial epochs. The gumbotils strengthen the view now generally accepted that the history of the Glacial Period involves, not a few thousand years but probably hundreds of thousands, and possibly millions, of years. tE. W. Shaw, ‘‘Characteristics of the Upper Part of the Till of Southern Illinois and Elsewhere,”’ Abstract, Bull. Geol. Soc. Amer., Vol. XXIX (1918), p. 76. DIASTROPHISM AND THE FORMATIVE PROCESSES XI. SELECTIVE SEGREGATION OF MATERIAL IN THE FORMA- TION OF THE EARTH AND ITS NEIGHBORS T. C. CHAMBERLIN University of Chicago In the last number of this Journal" I endeavored to deduce from a comparison of the earth with Mars, Venus, and the moon, the order of magnitude of the total shrinkage suffered by the earth. The results proved surprisingly large. Not only that but they seemed to show that the shrinkage per unit of mass-increase be- came greater as the total mass grew. Since small bodies have but feeble gravitative ability to gather and hold the lighter order of molecules in a free state, it seemed probable that the moon and Mars contain higher proportion of the heavy molecules than Venus and the earth. This seemed to add emphasis to the high densities of Venus and the earth compared with the moon and Mars. It appeared to strengthen the presumption that, in this group of bodies at least, the degree of density was due to the concentrating effect of superior mass rather than original heaviness of material. However, final conclusions were held in abeyance until the modes of organization of the four bodies could be studied with a view to detecting the probable laws of their segregation in so far as these affect the proportions of inherently light and inherently heavy materials. It is this inquiry that forms the theme of the present paper. ; The subject necessarily reaches back to the genesis of this group of bodies, and the discussion will need to concern itself quite as much with the dynamic environment that influenced their formation as with the material that entered into it. The «The Order of Magnitude of the Shrinkage of the Earth Deduced from a Com- parison with Mars, Venus, and the Moon,” Jour. of Geol., XXVIII (1920), pp. 1-17 126 DIASTROPHISM AND THE FORMATIVE PROCESSES 127 study need not, however, go much beyond the conditions that determined the amount and nature of the material, the mode of aggregation, and its physical state. It must be specific enough, however, to touch the conditions that controlled the relative proportions of inherently heavy and inherently light material. It is therefore necessary to consider with some care the basic laws of organization of such bodies so far as these bear on selective segre- gation. It will save time and help toward clear treatment to give at the outset what seem to me to be the more essential principles that control the formation of cosmic bodies. It will not be amiss if these are made rather sweeping, provided they are definite enough to apply to the particulars required by our problem. PRELIMINARY CONSIDERATIONS While the four bodies named will usually be in mind when other bodies are not specified, there will be occasion to use certain terms in other than their commonest senses, and so let us agree ‘upon these at the outset. By cosmic units let us understand, not simply celestial bodies, but organized bodies of any kind, whether large or small, whether ‘“‘organic” or ‘“‘inorganic,’”’ provided they serve a unitary function in natural processes. Let us recognize that these range from the atom, whose organization is now being pursued with a skill and success worthy of the highest admiration, up through the molecule, the crystal, the chondrus, the colloidal unit, the cell, the biologic organism, the planet, the star, the star cluster, to the stellar galaxy, at least. Let these more salient types stand for the multitude of intermediary and divergent species of divers sorts that make up the full series. Let them also stand for the unknown extensions of the series downward and upward. Let us agree that the only essential of a cosmic unit is an individual- ized organization which has its own material center and its own dynamic province. These organized units of course enter into varied relations with one another and form a great complex super- series, but let us consider merely the features that are common to all because essential to all. Among these essentials should appear, in their true relations, those particular features we. need to apply to the solution of our special problem. 128 T. C. CHAMBERLIN The wide range of the category thus recognized implies that the study of the organization of cosmic units is not a theme which falls solely within the province of any one of the natural sciences as we now know them; it is common ground, belonging to each and all in so far as their problems reached back into it. The biologic student of the genesis and career of the iron bacteria may invade, of his own right and to any extent that serves his purpose, the Proterozoic terranes or any other terrestrial field that promises him evidence, and we of the geological school may not say him nay, however much we may claim the province as peculiarly our own in respect to our own problems. Each particular science is best suited to make certain inquiries into the origin of organisms and of organizations, and to yield certain contributions to the common science of cos- mology, using the term in its broadest sense as the science of cosmic organization, in distinction from cosmogony, which in its original sense—the birth of the cosmos—belongs to philosophy, theology, and mythology. The cosmic systems mount up by hierarchies from what seem simpler to ‘what seem more complex, but in ultimate analysis all, even the atom, and perhaps even the electron, are themselves complex, and the depth of such complexity is at present unfathom- able. This pervasive complexity puts all in a common class and, in some sense, simplifies the common cosmologic problem, for its essence lies in finding those principles of organization that are so far essential to any organization as to be common to all. FUNDAMENTALS OF COSMIC ORGANIZATION _I. Every cosmic unit bears a dual aspect, a material organiza- tion, and a dynamic organization. In ultimate analysis these may be merely different phases of the same fundamental entity, whatever that may be, but in their sensible aspects they are distinctly diverse. The one is very tangible and impressive and has almost monopolized attention; the other is invisible in itself and has fallen much short of the recognition it deserves as an essential element in every cosmic organization. The first is too familiar to need emphasis here; the second requires all the em- phasis which a neglected essential can well receive. DIASTROPHISM AND THE FORMATIVE PROCESSES °129 II. The material factor of every cosmic unit is not only per- vaded internally but is surrounded by a field of force, domi- nantly attractive but partially repellant. In the present study, the sphere of dominating attraction—the sphere of gravitative con- trol—will, for brevity, usually be spoken of as though it alone represented the dynamic element of the dual organization, for it is chiefly the outlying field of gravitative control that functions actively in the selective segregation of material in the formation of cosmic bodies. The extreme theoretical reach of gravitation is indefinite, but within a certain portion of this indefinite field, the attractive force of the particular body under study is sufficient to give it immediate control over bodies inferior to itself, unless they carry kinetic energy of their own in sufficient amount, and properly directed, to insure their escape. This field of superior force constitutes the body’s sphere of control.‘ It is necessary to note that this control is merely zmmediate control; there are usually, perhaps always, higher types of control which hold ulterior sway over these, but this superior sway is exercised in such a concur- rent way as not to prevent the immediate control essential to the | minor body as a condition of its own existence and perpetuity. These higher controls may spring from some single more massive body or from some composite organization, as a star group. The superior spheres of control envelop the minor spheres of control; thus the moon’s sphere of control revolves within that of the earth; the earth’s sphere of control revolves within that of the sun; the sun’s sphere of control revolves within the sphere of control of the “local cluster” of stars, and this in turn within the sphere of ™The recognition that cosmic bodies are surrounded by spheres of gravitative control is not at all new but merely neglected; Laplace worked out “‘the spheres of activity” of the planets—here called spheres of control to avoid confusion with the indefinite outward extension of the influence of the cosmic body. The spheres of control of the planets have been worked out more recently on a different basis by Moulton (Popular Astronomy, No. 60, May 15, 1899). The spheres of equal attrac- tion, a different matter, have been worked out by the senior Asaph Hall (Popular Astronomy, April, 1899). The concept of a sphere of control is herein given a wider application than is common, and is assigned an essential function in the organization and maintenance of cosmic bodies. The concept is regarded as a helpful means of research, particularly as an aid to visualization. 130 T. C. CHAMBERLIN control of the stellar galaxy. The sphere of control of the atom enters into the sphere of control of the molecule, and that into higher orders in succession up to the earth and beyond. The whole cosmic scheme seems to be a system of such hierarchies whose limits in either direction are unknown. III. The dynamic value of each sphere of control dies rapidly away from the mass in which it centers to its outer border. Not only this, but each sphere of control that revolves within a superior sphere of control is larger or smaller, more effective or less effective, according to its position within such superior sphere of control. It is likely to be either increasing or diminishing as the body in which it centers swings through its orbit. If it were made to constantly approach the controlling body, its extent and efficiency of control would grade entirely away to extinction before such superior mass was reached. IV. In such spheres of control as center in single great masses, the differential pull of the controlling mass becomes so great relatively, in its innermost portion, that bodies of a minor order intruding upon it are liable to be disrupted.. When the approaching minor bodies are gaseous, their spontaneous tendency to dis- persion insures their dissipation. When they are solid, the degree of fragmentation to which they are subject is likely to be limited to certain sizes, for as the fragments grow small the strength of their cohesion increases relative to their mass. Cohesion is not likely to be important in large bodies because they are usually self- compressed and hot within to such a degree that their tendencies to expand, when pressure is relieved, usually surpass their coherence. The outer border of the disruptive zone is known as the Roche limit. Its determination by Roche was based on an ideal homo- geneous fluid satellite approaching an ideal homogeneous fluid planet of equal uniform density, cohesion being neglected. He — fixed the limit of disruption at 2.44 times the radius of the planet." This limit is in close accordance with what seems to be the realized result in the case of Saturn’s rings which stand as the classic ex- ample of minute division and distribution in response to this dis- ruptive effect. ‘The mathematical conclusions of Roche were amply * Edward Roche, Memoirs de Académie Montpelier, 1, p. 243. DIASTROPHISM AND THE FORMATIVE PROCESSES 131 supported some years later by the studies of Clerk-Maxwell,’ and their common results were afterward verified by the spectroscopic observations of Keeler, who demonstrated that the rings are formed not of gas, as once supposed, but of discrete particles revolving in independent orbits, in other words are minute satellites or satel- litesimals. From a recent study of the albedo of the rings, Bell has concluded that the largest masses in them probably do not exceed three meters in diameter—at least they are not more than a few meters across—while the majority of the visible particles are very much smaller, ranging down to the dimensions of wave- lengths of light.’ ; For the immediate purposes of our study, the important point is not so much that bodies entering this zone of disruption either from without or within are reduced to relatively small particles, though this is important, as that these conditions of stress from the controlling body stand in the way of the organization of any new body within this zone. - So far as the aggregation of independent bodies of any notable mass is concerned, this is an inhibitive zone. Considered with reference to the controlling body, it may perhaps be said to be a protective zone, tending to preserve its isolation, independence, and undivided sovereignty. In view of uncertainties as to the precise qualifications the Roche limit might require in the case of a rotating nebular spheroid, Moulton worked out a limit of similar nature but on a different basis, the purpose of which was merely to fix a more certain limit within which the organization of nebulous matter would be in- hibited. This limit was placed at 1.38 times the radius of the nebular spheroid, or a little more than half the radial extent of the Roche limit. V. Accepting as a working basis the Newtonian doctrine of the unlimited penetration of the force of gravitation, it is a logical deduction that all space is under the immediate domination of t*On the Stability of Motion of Saturn’s Rings,” Scientific Papers of James Clerk-Maxwell, Vol. I, pp. 288-375. 2Louis Bell, ‘‘The Physical Interpretation of Albedo, II. Saturn’s Rings,” Astrophysical Journal, L (July, 1919), pp. 1-22. 3 F. R. Moulton, “An Attempt to Test the Nebular Hypothesis by an Appeal to the Laws of Dynamics,” Astrophys. Jour., XI (1900), pp. 122-26. 132 T. C. CHAMBERLIN some dynamic organization or some combination of such organiza- tions, except perhaps that theoretical neutrality may arise momen- tarily on the border lines of spheres of control where exact balancings of attraction may obtain for an instant; but as all celestial bodies are moving relative to one another this can only be transient and unimportant. The concept of such pervasiveness of stress throughout all cosmic space has the merit of dismissing from serious consideration certain inherited notions, for example, that somewhere in space there may be regions where “ primordial” matter may have lurked in idleness from a supposed beginning, or where nebulous matter, or dissipated particles of any sort, may somehow assemble purely under their own attraction and later: drift into the active cosmic world as quasi-primordial matter, and similar notions that seem to be but reshaped vestiges of oriental concepts of primitive chaos. A much more important function of the deduction, however, is its amplification of the doctrine of dynamic encounter. | VI. From the preceding generalizations it follows that the spheres of control of cosmic bodies are either perpetually plowing through the higher orders of spheres of control that envelop them, or are impinging upon other spheres of control of their own type. In either case, their own domains, as well as those of the bodies with which they interact, are perpetually suffering encroachments. Innumerable dynamic encounters of widely varying types and moment thus spring from cosmic movements. As an incident of these innumerable encroachments of one domain upon another, transfers of the minuter class of units from the field of dominance of one controlling center to that of another are almost perpetual occurrences. They constitute a system of exchange of the first order of extent and seem to be a vital factor in cosmic life. In the case of the earth, revolving in the sphere of control of the sun, this system of exchange is regarded as having a very high order of importance in the maintenance of our atmosphere and the in- crease of our hydrosphere. Since encounters of some order of importance are almost in- finitely frequent, it is necessary to specify explicitly the nature of the encounter in any argument that is based on frequency of DIASTROPHISM AND THE FORMATIVE PROCESSES — 133 encounter. Furthermore, in determining the frequency of a given class of encounters, it is not sufficient to postulate an artificial case convenient for computation, for actual cases usually involve a natural selective adjustment of associated bodies with reference to one another. And further, the present deployment of stars may not be identical with that of earlier ages. And still further, if the frequency has for its criterion a given effect or is to be considered with reference to a given effect—explosive action for example— susceptibility to such effect is as important as the nature of the encounter. ‘These requirements are commonly neglected. VII. All cosmic organizations seem to be the products of oppos- ing elements. The balance between these opposing factors seems to form the critical issue on which their endurance depends. These opposing factors vary with age, state of growth, environment, and other conditions. Out of these variations of balance arise stages of increase and depletion, of partial or total disorganization and regeneration. ‘The history of the cosmos seems to be essentially a succession of cycles arising from either internal or external dis- turbances of balance. Interestingly enough, the atom happens just now to afford one of the best illustrations of internal disturbance leading to transition in organization. Thought until recently to be beyond the utmost resources of disintegration, it is now known that some of the heaviest atoms are undergoing ‘‘spontaneous’”’ dis- organization. By way of offset for the old error, as it were, the dictum now is that no known device, appliance, or force can stop this disintegration. Future inquiries will probably disclose the golden mean between these extremes. In spite of all disintegrating influences, the integrity of atomic organization, in the main, is maintained to an extraordinary degree. In the larger cosmic world there are intimations of analogous “‘spontaneous”’ disin- tegrations standing over against similar persistency. The erup- tivity of the sun, on which the planetesimal hypothesis is founded, is revealing striking analogies to the partially disintegrating atom as will be detailed later. Some of the great hot stars give intima- tions of a very delicate balance of internal forces. Over against the enormous concentrating force of gravity and its allies, stands the scarcely less potent alliance of the forces of dispersion. These 134 T. C. CHAMBERLIN latter seem, on the whole, to be overmatched by their opponents, as implied by the very existence of the stars; and yet the forces of dispersion are clearly successful in particular ways, as, for example, in the matter of radiations and in some loss of high-speed mole- cules and of electrons. In this state of wavering balance between internal contending forces, those great seething bodies are plunging at high velocities through what, in a material sense, is an approximate vacuum, but what, in a dynamic sense, is an approximate plenum, a plexus of lines of force of almost infinite complexity. They are thus speeding through a perpetual succession of contingencies of external dis- turbance. Their hold upon their own material hangs on the perpetu- ated superiority of their concentrative forces, expressed typically but not wholly in their spheres of gravitative control over their dispersive forces. If the controlling spheres are invaded in a shallow way there is likely to be only trivial loss; if they are invaded deeply, serious disintegration is the logical effect. It is important to note that this disintegration is a joint effect, as much due to the approximate balance of the internal forces as to the disturbing power of the external forces. And so in dealing with phenomena of this class, it is not more important to inquire into the direct action of the external agencies than into the state of balance of the powerful forces within these supremely active organizations themselves. This is the more im- perative because there is growing reason to believe that certain orders of stars are at or near the limit where growth in mass is over- matched by concurrent growth in dispersive forces. If this belief is well founded, such nearly balanced giants of the skies may be regarded as peculiarly susceptible to disturbances of equilibrium arising from the intrusion of foreign dynamic influences into their domain, or perhaps equally their own penetration into areas of concentrated stress arising from special marshallings of other great bodies. These considerations are deployed at some length here because the stellar conditions that render dynamic encounter effective are too commonly overlooked, and because these conditions are vital in considering cosmic disorganization which in turn is regarded as a step necessarily precedent to cosmic reorganization. DIASTROPHISM AND THE FORMATIVE PROCESSES 135 VIII. There is pressing need for rectified concepts of cosmic time and stellar endurance. The recent deep penetration of the stellar field by improved instruments, increased skill, and new methods has forced a revolutionary enlargement of concepts of interstellar space, while co-ordinate enlargements of concepts of cosmic time have not kept apace. And yet time and space are necessary correlatives in stellar movements and in stellar organiza- tion. Time concepts must keep pace with space concepts if consistent views of organizing processes are to be entertained. Inadequate concepts of time retained from old estimates of the sun’s longevity and like sources now embarrass the free acceptance of cosmic views—if these imply great intervals of time—much as they restrained geological interpretations during the last century. It may be wholesome therefore to inquire specifically: What length of life is implicitly assigned to stars when they are made integers in the evolution of a globular cluster or of a galaxy? What intergenetic periods mark off the generations of stars? What careers appropriately fit them into the vast cosmos that is now revealing itself? And then, subordinate perhaps to the life of a star, what intergenetic periods mark off the generations of planets ? It is to be recognized, of course, that the career of a planet may belong to a different order of magnitude from the career of a star, or of a star cluster, or of the galactic system, and no doubt these differ among themselves. And so our immediate problem may not be more than remotely concerned with these immense questions, but yet it is related to them, and its answer should be consistent with them. Perhaps all that need be said here is that when the estimates of the longevity of stars, star clusters, and the stellar galaxy are brought into harmony with the time requirements of their own processes of organization and their own normal careers, students of the evolution of our little planet will probably feel quite as much call to amplify as to repress their interpretations of the terrestrial time factors in order to bring them into harmony with those of the higher systems. IX. As a matter of scientific conservatism, it should be taken for granted that the sole source of material and of energy for the formation of new organizations is to be sought in the dissolution 136 T. C. CHAMBERLIN of pre-existent organizations. In most cases, if not in all, how- ever, the dissolution of such previous organizations is not ulterior dissolution; it is usually only the disintegration of one order of organization into elements of a somewhat lower order. Neither dissolution into chaos, in any strict sense, or into ultimate factors, nor generation from chaos or from ultimate factors or from any- thing that is absolutely new, seems to have any naturalistic warrant so far as present cosmic processes are concerned. Nor is there much more warrant for bringing into play any really unknown force or agency, though the discovery of new ways of action of known forces and agencies is to be expected. Scientific inquiry in genetic lines appears therefore to have for its appropriate field of study little else than partial disorganization followed by corresponding reorganization, though not necessarily of the same type, order, or extent. The scientific student therefore hesitates to call into service any material or energy which he cannot trace back to some known source. X. In the formation of new cosmic bodies, even of the “ organic”’ class, an organizing germ or nucleus, inherited from some parent organization, seems to have much the same function, and to be about as necessary as the seed or the ovum of an organism of the “organic” type. In either case the germ must apparently have both a material and a dynamic factor. This necessity appears to be chiefly due to the essential part which a collecting field of force and a retaining sphere of control play in organizing a cosmic body. In the concrete discussion to which we shall now turn, the strength and the reach of the organizing field of force will be held to be the chief criterion that determines whether dissevered or disorganized masses of matter shall reorganize as single bodies and pursue independent careers, or shall continue to be merely scattered food to be picked up by bodies already cheeged and endowed with: effective collec Ling fields of force. THE SELECTIVE SEGREGATION OF PLANETARY MATERIAL In this discussion there will be no occasion to consider any hypothesis of the origin of the four bodies under study that has not been worked out into such definite terms as to bear specifically DIASTROPHISM AND THE FORMATIVE PROCESSES 137 on the question of the segregation of inherently heavy from in- herently light material, the crux of our problem. Two postulated origins may clearly have such bearings, both of which are now familiar: (1) derivation from a rotating nebular spheroid by centrifugal separation brought into effective action by cooling and consequent acceleration of rotation, and (2) derivation from solar material ejected either spontaneously or under the stimulus of a passing body. ‘The principles involved in these two types will probably serve to cover any other origin for which good reasons may be assigned. While I am unable to see how planets such as form our system can have arisen from a rotating spheroidal nebula by centrifugal action, it yet seems best, out of deference to any who may still think that some view of this general type is tenable, to discuss this postulated mode of genesis in so far as it bears on our problem. It will only be necessary, however, to review the phase of the theory most dependent on the dynamic environment which con- trolled the evolution, for that touches the soul of the subject. THE CENTRIFUGAL EVOLUTION OF A GASEOUS SPHEROID UNDER ITS OWN DYNAMIC ENVIRONMENT Every organized nebula, like every other organized body, must have an adequate enveloping field of force and sphere of control as a necessity of its organized existence (I and II, above). The pos- tulate that there was once a spheroidal nebula of the mass of the solar system which in contracting shed secondaries at various dis- tances from its center as far out as 23 billion miles and yet was able to hold them then and afterward, carries the implicit assumption that it had a distinctly effective sphere of control. The shedding of the four little bodies under study took place only after the postulated nebula had shrunk ‘to about one-twentieth of the radius it had when it displayed its effective holding power by its control over the material shed for the planet Neptune, while the outermost reach of its holding power must have extended much beyond this. At the relatively concentrated stage when the shedding of the substance for our little group of bodies took place, the inner zone of control must have grown relatively intense; the 138 T. C. CHAMBERLIN disruptive belt just outside the rim of the rotating spheroid (IV above) must have had a notable development. If it were quite safe to assign it the full breadth of the Roche limit, which holds so well in the case of Saturn’s rings, it would have had, taking the earth stage as a mean, an outward reach of 133,000,000 miles. But let us follow the safer course of using the conservative criterion of Moulton which gives a zone of 35,000,000 miles (IV above). Let it be recalled that the fragments of a disrupted mass revolving about the controlling body under the conditions of this case take orbital courses more or less parallel with one another. If the gaseous rim of the nebula could have been “thrown off” as a coherent body, it would not only have been disrupted into minute constituents, but these would have been given orbits of a type much like those pursued by the particles in Saturn’s rings, all the more because the constituents of a gas tend to disperse themseives by their own interaction. This is equivalent to saying that the dynamic conditions within this zone were such as to inhibit any aggregation of this material into a common large body like the earth, Mars, or Venus, or into a lesser number of bodies of any considerable size. Even when such scattered orbital matter was left by the withdrawal of the nebula in the less intense horizons of the nebular field of force, its aggregation would still be greatly embarrassed by the superior control of the central mass. It is a common error to think of such scattered matter as though it were in neutral space entirely free from all forces except its own mutual attractions. The control of the central body so far embarrasses the assemblage of minute particles under the actual conditions of’ such a case as this as to render their aggregation into a single body improbable, as Moulton has so effectually shown." However, for the sake of seeing its bearings on the problem in hand, let us waive this improbability and try to follow the aggre- gation ofthe minute constituents of a quasi-Saturnian ring “thrown off”? from the postulated rotating nebula.’ *F. R. Moulton, ‘‘An Attempt to Test the Nebular Hypothesis by an Appeal to the Laws of Dynamics,” Astrophys. Jour., XI (1900), p. 115. 2 The deduction that the molecules shed by centrifugal action from a rotating gaseous spheroid would pass into individual orbits and form a planetesimal system does not depend solely on the Roche effect, as shown in ‘‘The Bearing of Molecular Activity on the Spontaneous Fission of Gaseous Spheroids,” Publication No. 107, Carnegie Institution of Washington, 1909, pp. 161-67. DIASTROPHISM AND THE FORMATIVE PROCESSES 139 Let it be noted at the outset that the material to be aggre- gated was planetesimal in a very strict sense of that term. Each integer, whether it be a molecule, a particle, or any such more considerable aggregate as might be formed under the conditions of the case, was pursuing a nearly circular orbit around the central con- trolling body, the nebula at first, the sun later. This precisely fits the definition of a planetesimal. This means that the parti- cles were in a dynamic, not a static state; they were under con- trol, not free. Whatever aggregation followed was therefore of the planetesimal type, that is, particle joined particle in an individual way as their orbits and orbital forces permitted. Their orbital velocities hovered about that of the earth (18.6 miles per second) let us say, as a mean, the inner faster, the outer slower, the orbits of those equally distant from the center slightly inclined toward one another. Beside these differences of velocity and inclination— that arose from the nature of the case—the planetesimals inherited diverging courses from mutual collisions and rebounds as they emerged from the gaseous into the orbital state. To overcome these divergencies of orbit and these differences of speed and develop aggregates of one kind or another in place of the molecules inherited from the gaseous state, there were two classes of forces: (1) the collective attraction of the whole ring or disk or some bunched portion of it, and (2) the aggregating influences of indi- vidual molecules upon one another. ‘The first would tend to make a single planet, if the whole were drawn together, or a few planet- oids, if there was aggregation by bunching; the second would make at first a multitude of minute particles which would grow to larger sizes In proportion as the agencies of later aggregation proved superior to the effects of fragmentation, exfoliation, tritura- tion, and friction in other forms after the particles had grown large enough to give these notable efficiency. Only the salient features can be touched here. t. The ground of the first class of agencies has already been covered. No general nucleus nor any effective bunching was in- herited from the nebula; indeed concentration was definitely inhibited up to the time of the withdrawal of the Roche limit. There might be, to be sure, a certain kind of transient bunching of the planetesimals in their orbits, such as affects the present planets, 140 T. C. CHAMBERLIN but the relative rates of revolution that brought this about would destroy it. Collectively, the planetesimals would be so distributed that they would have almost no concentrating force of their own that was not later reversed or neutralized by their own orbital motions. 2. Practically the whole aggregation, then, would be that of the formation of discrete particles such as started with the joining of molecules and were built up thence into crystals, pellets, nodules, or whatever these might grow into later. The chemical combina- tion of molecules would take place at proper temperatures readily enough by simple contact, whether this arose from collision while in the state of a gas, or from contacts brought about by planetesimal motion, or otherwise. Such refractory chemical compounds as now form the main mass of the solid bodies of the moon, Mars, Venus, and the earth, would probably be formed at high tempera- tures while they were still a part of the postulated nebula. The critical feature of the case lies in the way these complex refractory molecules would be gathered together after they were formed. While they remained in a free state as molecules they would normally tend to rebound on collision as molecules do and so maintain their free state. Even if they were brought together under conditions favorable to remaining together, their rotations or vibrations would tend to throw them apart, as would also sub- sequent collisions. To overcome these adverse influences, there was need for some special uniting agency, as is well recognized in the familiar case of the formation of the globules of fogs and _ clouds from water vapor in the atmosphere. It was long supposed that there must be a dust particle or some similar aggregate to serve as a collecting center (the ‘‘seed,” X, above); but it was later found that molecules electrically charged serve this function also. In our problem, the formation of the first minute aggregates is the very crux of the question, and we cannot assume the existence of any such dustlike aggregates as the means of starting the process. But molecules electrically charged would probably be freely developed by friction, by the action of ultra-violet light, and by other means, and such charged molecules might well serve as the “‘seed”’ for starting the minute aggregates. DIASTROPHISM AND THE FORMATIVE PROCESSES 141 How far would such aggregation be likely to go, as a rule? There is no need to consider exceptional possibilities, for our problem relates to the common average result. The attraction between two molecules oppositely charged is many billion times greater than their gravitative attraction" and may be large compared with the inertia of their relative motion. Charged molecules might then serve as very efficient centers for the gathering-in of molecules, as also very small particles. But an electric charge is - confined to the surface of a particle, which increases as the square of its radius, while gravitation varies as the mass which increases as the cube of the radius. And so, after a certain amount of growth, the charges carried on the particles would have less attractive power than the masses into which the particles had grown. Buta more important practical consideration lies in the fact that electric charges of like kind repel one another and thus limit the total charge likely to be gathered on a given mass under natural condi- tions; for example, any electric charges which a forty-pound bolide would probably pick up naturally would lend little aid in gathering in other forty-pound bolides to form a forty-ton bolide. There is thus an obvious limitation to the range of effective electric aggregation, however efficient it may be as an originating agency. A beautiful illustration at once of such effective aggregation and of its limitation is presented by the formation of snow crystals from vapor in the air. These form and grow with great facility up to a certain’ size when the temperature of moist air falls below the freezing-point; but after a certain moderate growth, the limiting and adverse conditions increase in relative efficiency and arrest further growth; not infrequently it is reversed. In the case of cosmic particles probably the most effective ‘preventive of indefinite growth is the friction and collision of the masses, themselves. As the particles grow into nodules of notable size and mass, their cohesion is less effective relative to their moving force, and they more readily go to pieces on impact. ‘Trituration and other lesser effects of moving contact would be more frequent *R. A. Millikan, in a personal communication, states: “The attraction of two opposite electric charges is 1077 times as great as the gravitative attraction of two atoms of hydrogen.” 142 T. C. CHAMBERLIN and perhaps more effective on the whole than fragmentation. This would quite surely be true of the minute particles. According ‘to the interpretation of Bell,t a milling process of this triturative sort has proved very effective in reducing to minute sizes the satellitesimals of the Saturnian rings. When masses of any con- siderable size were reached, probably exfoliation from the effects of rotation in the unscreened rays of the sun would give rise to flaking and thus prepare new matter for the milling process. On the other hand, if magnetic particles were much developed, it is probable that their special attraction would aid in building up masses, so far as the supply of such material went. Probably such malleable substances as iron, nickel, and the other metals would weld rather freely by impact. Metallic particles might thus unite nearly to the extent they were permitted to come into collision. Such stony substances as brought crystalline or con- cretionary forces into play would probably build up more readily than other matter and more effectively resist destructive agencies afterward. But it must be noted in all these cases that as the molecules were more or less heterogeneously mixed originally, the opportunities for assembling homogeneous matter to form aggregates of any one kind would have natural limitations. The logic of the case, taken all together, seems to lead to the conclusion that aggregates arising under these ideal planetesimal conditions would be limited to small sizes as a general rule. This is in harmony with the results realized in the Saturnian rings, and also in the zodiacal planetesimals to be more fully discussed in the next article. It is also in harmony with the dimensions attained by the chondri and chondrules that form characteristic constituents of go per cent of the stony meteorites. These range in size from a walnut down to spherules of dustlike minuteness.’ The point of most critical importance to our inquiry is the effect on selective action introduced by these growths so far as the went. In the first place, all such matter as continued in a free molecular state, whether aggregated as gases or deployed as planetesimals, would not be gathered about these small aggregates OG Cite 20. C. Farrington, Meteorites (1915), p. 102. ( DIASTROPHISM AND THE FORMATIVE PROCESSES 143 by their own gravitative power. Incidentally molecules might be entrapped or occluded within these small bodies, or chemically united with them, or possibly even held against them by surface adhesion; but, otherwise, free molecules would rebound and escape control. Those solid particles that were highly elastic would also largely escape by rebound; those that were inelastic would more largely remain adherent after impact. Malleable substances like the metals would be likely to weld and cohere by collision. In general, these cohering bodies belong to the heavier order of sub- stances, and so bodies formed in this way would be for the greater part selectively heavy. If we could be sure that the chondrules of meteorites represent accretion of the foregoing type—a hypothesis to be seriously con- sidered—it would give a specific insight into the cosmic aggregates of this order, for then they might be said to be dominantly formed of ferro-magnesian silicates, nickel-iron, and metallic sulphides, but it would be premature to draw this conclusion. At any rate, since, on the one hand, these small bodies could not hold free molecules of the lighter order, and, on the other hand, _ the conditions were favorable for the aggregation of metallic sub- stances, heavy silicates, sulphides, and so forth, it seems safe to conclude that these small aggregates contained a relatively high proportion of inherently heavy matter. It seems to follow then that, if Mars, Venus, the earth, and the ‘moon could have been gradually built up by the assemblage of particles, crystals, pellets, nodules, or even more considerable masses formed in this selective way, the percentage of heavier con- stituents could scarcely have been less in the earlier stages and in the smaller bodies than in the later stages and in the larger bodies, while they were probably somewhat distinctly greater; for in so far as these bodies succeeded in becoming large, they could then, but only then, have held the lighter order of molecules in a free state and thus have reduced their mean density. When atmospheres and hydrospheres were thus added, the processes of oxidation, hydration, and carbonation became important and the groundwork was laid for petrological derivatives from the products of these processes. A large class of the rocks and minerals of the outer I44 T. C. CHAMBERLIN part of the earth, which usually have rather low specific gravities, are probably wholly dependent on the presence of the atmosphere and hydrosphere at the time of their formation. And so, if the moon, Mars, Venus, and the earth, could have been built up in this way, the moon should have the highest percentage of heavy material and the others should follow in the order of their masses, atmospheres, and hydrospheres. But it was previously shown that the conditions were very adverse to the building up of these four bodies in this way and there is no likelihood that they had such an origin. It is perhaps worth while to add that, even if we were wrong in concluding that the planetesimal aggregates would be small, the result would be little different in density or in physical state, for as the aggregation of the particles in their orbits proceeded, the resulting aggregates would become more widely separated and further aggregation would take place only at correspondingly longer intervals. There would be little change in the kind of material or in the heat effects. The material would be a little more bunched before infall, but the bunches would be more scat- tered in space and successive infalls more distant in time. Pre- cipitate aggregation is ans out of the question under these conditions. As remarked at the beginning of this section, the material discharge from the rim of a rotating spheroid of gas by centrifugal action should form an ideal system of planetesimals, and so the method of their growth may be taken as a type of such action where the molecules are given subparallel orbits from the start and there is no commanding nucleus to gather them into bodies of the planetary order. If this analysis is correct, it will be seen that the chance of developing a molten earth from a rotating gaseous nebula by centrifugal separation is about as remote as could well be imagined. THE SELECTIVE SEGREGATION OF MATERIAL UNDER THE PLANETESIMAL HYPOTHESIS The particular form of the planetesimal hypothesis which has been most fully worked out and tested postulates that the material of the planets was derived from the sun by means of its own DIASTROPHISM AND THE FORMATIVE PROCESSES 145 ejective activity stimulated to special intensity by the differential attraction of some passing body. Two essential factors are involved: (1) the ejection to the requisite distance of the requisite matter— only a fraction of 1 per cent of the sun’s mass; and (2) the addition of sufficient transverse momentum to cause the ejected matter to revolve about the sun. The latter is the more critical factor, for the eruptivity of the sun is known to have such a high degree of efficiency, even at the present time, that only a relatively slight increase is required to project the small fraction of the sun’s substance to the distance of the planets. Such projections would, however, fall back to the sun, unless they were given a transverse motion by some agency other than radial projection. A passing star or other body has been postulated to meet this requirement. The tidal stresses developed in the sun by such passing body would stimulate eruptivity and give direction to the projections while the pull of the passing body would draw the ejected matter into orbital courses. In about a half-hundred cases worked out pee Urea by Moulton to test the validity of the postulated effects, a star of medium size passing at from one to five astronomical units’ distance was taken as the parent of the orbital motion; its diverting com- petency was found to be unexpectedly effective.t Later it was suggested that a much smaller body—passing how- ever much nearer to the sun—might serve as well to give both the tidal stimulus and the transverse motion required, but this has not yet been worked out mathematically.? Still more recent studies have led to the belief that there is a wide range of possibilities respecting the co-operating body, as will be specified later. RECENT DISCLOSURES BEARING ON THE SOLAR PARENTAGE OF THE PLANETS Respecting the projectile power of the sun, important light has been shed by very recent discoveries. Remarkable eruptions of the sun took place on May 29 and on July 15, roro. A fine series of spectroheliographic photographs were taken at the Yerkes * Carnegie Year Book, No. 5 (1906), pp. 166 and 168. 2 The Origin of the Earth (1916), p. 118. 146 T. C. CHAMBERLIN Observatory by Pettit and associates.t These disclose motions of quite an astonishing nature. Fortunately for our purposes, the photographs were taken with one of the calcium lines, and as cal- cium has the atomic weight 40 and enters widely into the constitu- tion of the earth body, its projection in this effective way has more significance than if it were one of the lightest elements. The gen- eral facts of the two cases are shown by Plates I and II, reproduced here from Plates IV and VI of Pettit’s article. Both eruptions gave rise to archlike forms. The apex of each arch vanished while it was still ascending, the first disappearing at a height of 760,000 km. above the surface of the sun, the second at 720,000 km., that is, at heights somewhat more than the radius of the sun in each case. At the time of disappearance the first had an ascensive velocity of 60 km. per sec., the second, 163.9 km. per sec. Perhaps the most remarkable features disclosed were suddenly increased rates of ascent taken on at intervals, while the rates of ascent between these stages of increase were essentially uniform... Thus in the eruption of May 29, there was a rise, for 50,000 km., from a point 150,000 km. above the surface of the sun at a rate of 5.5 km. per sec., when the rate changed to 9.2 km. per sec., which was main- tained for 119,000 km., when the velocity again changed to 27.9 km. per sec., which was held for 91,000 km., when the rate again increased to 60 km. per sec., which was maintained for 230,000 km: and was still being held when the prominence vanished, doubtless either by cooling or dispersion or both. In the eruption of July 15, it was found that from 200,000 km. to 294,000 km. above the sun’s surface the rate of ascent of the center of the arch was 37 km. per sec.; at the latter height the velocity abruptly increased to 163.9 km. per sec., which was held for no less than 426,000 km., and was still retained at the disappearance of the projected mass. However these extraordinary phenomena are to be explained, two significant things are implied: (1) that in some way the gravi- tation of the sun is offset or neutralized sufficiently to permit uniform motion, so far at least as the projected matter was con- t Edison Pettit, ‘The Great Eruptive Prominences of May 29 and July 15, 1919,” Astrophys. Jour., L (October, 1919), pp. 206-10. 7 JourRNAL oF GroLocy, VoL. XXVIII, No. 2 PLATE I C G.M.T. RBe amy Ts b 2h56™568 a thyr™76s THE GREAT PROMINENCE OF MAy 29, IgI9Q Scale: for a, 1mm. = 9,326 km.; for 6 and c, 1 mm. = 8,416 km. (After Pettit) JoURNAL oF GEoLocy, VoL. XXVIII, No. 2 PLATE IT h G.M.T. Ap 9s § 3h 7568 f ghgmas THE PROMINENCE OF JULY 15, I9I9 Scale: Imm. = 9,572 km. (After Pettit) DIASTROPHISM AND THE FORMATIVE PROCESSES 147 cerned; and (2) that this matter received successive strong outward impulses, amounting in the last case to an increase of projection no less than 126.9 km. per sec. It is to be noted that this impulse was received after the arch had reached a height of 290,000 km. above the sun’s surface. It is further to be observed that at the time the projections became invisible, in each case, more than half the restraining power of the sun, measured in terms of velocity, had been overcome. In the second case, the speed, at the time the projected matter became invisible, was competent to overcome more than half the remainder of the sun’s restraining power. Pettit finds data confirming these strange modes of movement, but of a less conclusive kind, in the photographs of certain other prominences already taken. He regards this singular mode of ascent by sudden accessions of speed with uniform motion between as the common one. However, none of these cases are sufficiently complete in themselves to show the full nature of these remarkable phenomena, for the ejected material passed out of sight while still under the highest observed uniform motion, and the extent to which this uniform motion may have continued and what followed it are left undetermined. While further disclosures are required, and can only be awaited with ‘eagerness, enough has already been revealed to give radical suggestiveness to these phenomena. They show that even at present and without obvious external stimulus there come into action, in addition to the internal eruptive forces, projectile forces of a high order which became effective at horizons high above the sun’s surface, and that the combined projectile effect of these had overcome a large fraction of the restraining power of the sun before they passed out of sight. Still another feature of the solar eruptions of May 29 and July 15, 191g, is scarcely less remarkable than their singular increments of motion. The projections on each of the two dates took the form of arches whose centers, at first low, rose impulsively into the forms shown on PlatesI and II. The arch of May 29 was transverse to the sun’s equator and had a chord of 584,000 km.; that of July 15 stood obliquely across the equator and had a chord of 363,000 km. as seen in perspective. The two ends of the arches appear 148 T. C. CHAMBERLIN to have functioned quite differently. At one end, the arches seem to have sprung from short stumplike prominences that had been present for some time before the special eruptions of the dates named and remained for some time afterward. The movements — of the calcium clouds near these seemingly originating ends were not only upward but inward toward the center of the arches. At the other end, the arches were apparently related to sun-spots. In these ends of the arches the calcium clouds seemed to be shooting swiftly toward the sun-spots. The photographs appear to show that special features in the upper part of the arches were drifting more or less from the prominence at the originating ends toward the sun-spot ends, though the motion of the central part of the arches was mainly upward. The total motions of the individual calcium molecules seem thus to have embraced a notable lateral component as well as the dominant ascensive one. The discovery by Hale and © his associates that the cyclonic whirls associated with the sun-spots are negatively charged may perhaps be made to throw light on this. When ionization takes place by the discharge of an electrical ele- ment, it is usually the electron that is shot away, and the residual matter is then commonly positive. If, therefore, it be assumed that the calcium molecules shot forth from the stump prominences were positively charged, they would be drawn toward the negative charges of the sun-spot whirls. A further feature of much interest is the suggestion of a rota- tional component in the projectile motion. This is implied in what has just been noted, a lateral movement combined with a vertical movement. A rather distinct expression of rotation seems to be shown in the spiraloid form of the upper central mass shown in Fig. C, Plate I. The value of this rotatory movement may be inferred from the fact that this spiraloid cloud had a volume more than 1,000 times that of the earth, assuming that its diameter in the line of vision was equal to the shorter of the two visible diameters. While all these disclosures must remain sub judice until they have been amply verified and their interpretation made sure, it is permissible to bring their suggestiveness into service at once to mitigate the force of old views that always act as a drag upon new DIASTROPHISM AND THE FORMATIVE PROCESSES 149 views. ‘These disclosures should help to lessen all hesitation in accepting the view that our atmosphere is even now receiving solar contributions. ‘They should also lessen doubts as to the possibility of the projection of great masses of sun substance to planetary distances whenever stimulus is added to the spontaneous eruptivity of the sun. MULTIPLE PHASES OF THE PLANETESIMAL HYPOTHESIS In these new disclosures there is the germ of a new phase of the planetesimal hypothesis, a phase that may possibly dispense with aid from outside the solar system—heretofore supposed to be necessary—and so make the origin of planets a wholly domestic affair, though at present the suggestion does not look very promis- ing. These disclosures seem to make the radial projection of the matter requisite for a planetary system possible without stimulus from outside. At the same time a transverse component of the projection is indicated in what has just been said about the lateral movement from the originating end of the arch toward the sun- spot end. If this lateral movement should prove adequate, and also be found to be preponderant in the right direction, it would contribute the component necessary to the revolution of the projected matter. The lateral movements in the two cases observed were chiefly across the equator of the sun, that is, normal to the required effect instead of coincident with it; but the position of the arches transverse to the line of sight in these two cases may be the essential reason why they were observed to such good effect. Similar movements in the line of sight, and in the direction of the sun’s rotation, that is, the direction of planetary revolution, may not only be as common as these, but even be the predominant ones. As the suggestion has only limited plausibility at present, it need not be further deployed here, but as it offers the possibility of a monoecious development of the planetary family in distinction from the dioecious origin heretofore postulated, even the shadow of such hypothesis is welcomed to a place among the multiple work- ing sub-hypotheses that make up the planetesimal genus. Arranged with other developments of like order, the group of such sub- hypotheses embraces the following: 150 T. C. CHAMBERLIN 1. Stimulus to eruptivity, as well as generation of tangential motion in the projected matter assigned to a passing star. a) Star of medium mass and distance (one to five astronomical units, more or less). Original type, tested mathematically by hali-hundred concrete trial cases. b) Giant star (in mass); distance great, tidal effect small, tangential effect large. c) Diminutive star, distance relatively small, tidal effect relatively large, tangential effect relatively small. 2. Stimulus to eruptivity, as well as generation of tangential motion in the projected matter, assigned to some non-stellar body, a stray planet for example. Approach to sun quite close, perhaps pene- trating its Roche limit; tidal effect relatively great; adequacy of tangential effect less obvious, but assigned to the closeness with which the solar projections were shot out behind the passing body. 3. Stimulus to eruptivity, as well as generation of tangential motion in the projected matter, assigned to a special concentration of gravitative stress in open space arising from two or more related bodies of large mass, for example, the center of gravity between two stars, or the concentrated gravity-stress of star clusters in certain forms of ar- rangement. 4. Actuating forces arising wholly within the solar system. Pro- jectile effects assigned to eruptive and projective forces within the sun; the tangential effects assigned to co-operative action of positive and negative centers in the sun as suggested above. Little more than a suggested possibility. CRITICAL PHASES OF THE EVOLUTIONARY PROCESS Let us follow that form of the planetesimal hypothesis whose working competency has been most fully tested. According to this the nuclei of the planets and satellites arose from solar eruptions— those of the planets from the central masses of such eruptions, those of the satellites from subsidiary masses that closely accom- panied these and kept within their spheres of control. It is our special task to follow the nuclei of the four little bodies under study from their source in the sun to their organized states, having especially in mind those features that bear on the segregation of DIASTROPHISM AND THE FORMATIVE PROCESSES 151 heavy from light material, but it will be helpful to keep in mind | bodies of both the larger and the smaller orders. So far as the selection and the segregation of matter is concerned, there was no essential difference between the planets and the satellites as such, for each arose from independent portions of the erupted sun substance. The critical elements were the spheres of control dependent on mass and dynamic environment. While we must await further light on the precise modes in which solar gas-masses are shot forth and the circumstances that induce them, we may be quite sure that, as they passed away from the sun into the outer field of its control, certain influences inevi- tably affected them. ‘They must have been under the control of a projectile force sufficient to overcome the larger portion of the sun’s total attraction. In this controlling force we may safely assume that there were conjoined (1) an original projectile force having its origin in the interior of the sun, (2) radiation pressure from the sun after the mass had left its surface, and (3) electrical effects, attractive and repellent, as also ballistic, that is, due to the mo- mentum of electrons and alpha particles shot forth from the sun and caught by the escaping mass. Just what proportionate parts these co-operating agencies played in the total work of projection, we need not now inquire. It is taken for granted, since it is almost inevitable, that in escaping from the sun the gas- masses acquired some measure of rotatory motion, in addition to the rotation they already had as parts of the sun; there may have been included some measure of vortex motion as most eruptions generate such motion. There can be no question that practically all the constituents of the outbursts were in the gaseous state as they emerged from the sun and that they carried in to the sub- sequent evolution the molecular activities common to hot gases. The several projectile motions were more or less independently imposed on the emerging mass, and later these underwent more or less independent increases and declines, so that an important part of the ensuing evolution consisted in their mutual adjustment to one another. At the outset, the projectile velocity greatly pre- ponderated over the velocities of all other motions, and until this became adjusted so as to be approximately proportionate to all 152 T. C. CHAMBERLIN integers of the mass it was a prime source of turbulence and danger of dispersion. In so far as molecules were driven by it beyond the sphere of control of the common mass, they took courses of their own and, in the main, either returned by elliptical paths to the sun or became planetesimals. The danger of dispersion at this stage was a serious menace to all the minor masses whose self- control was feeble; at the same time it was a prolific source of planetesimal food for the growth of the nuclei later. The hypothesis assumes that a part of the out-shot masses contained matter enough to give them self-control. The effective- ness of their control increased as they passed from the more intense to the less intense field of the sun’s control; except as they lost material. But such self-control is not postulated of more than a part—probably less than hali—the matter projected from the sun, the more scattered portion becoming planetesimals at once or else returning to the sun. Effective self-control is only assigned to a few of the greater eruptions, or, to be specific, to four of the major order, to form the nuclei of the four giant planets, and to four of the minor order, to form the nuclei of the terrestrial planets. But subordinate to these, perhaps a thousand or so little knots succeeded in holding themselves together and later grew into planet- oids and satellites. It is thus assumed that there arose, as a result of eruptive projection and partial dispersion, a graded series of knots ranging from those that were massive enough to form the nuclei of the great planets down through medium and smaller knots to masses too small to hold themselves together in the face of the dissipating influences. It was of course in the lower ranges of this graded series that there arose the more critical questions of self- control and of permanent maintenance. The answers to these critical questions hung, in each individual case, very largely upon gravitative competency. Now we need not dwell on the largest order of knots, for they do not enter our problem. Their strong attractions enabled them to hold their own, except for a small percentage of molecules that attained exceptional cumulative speeds, while, on the other hand, they were able to pick up stray planetesimals that came within their spheres of control in a favorable way. And so in the end, DIASTROPHISM AND THE FORMATIVE PROCESSES 153 between their ability to hold all sorts of molecules as well as pick them up, they came to have a much larger proportion of light molecules than the smaller bodies which could only hold the heavier ones, and so they now have relatively low specific gravities. Their great size helped them to remain hot, which also contributed to their low specific gravities. The largest of these is now more than three hundred times as massive as the largest of the terrestrial group; the mean mass of the giant planets is more than two hundred times the mean mass of the terrestrial planets. Our problem then is with a group of distinctly small bodies in which the balance between holding and non-holding power was more critical, and we need to enter somewhat more into detail. 1. As the material was shot forth from the sun, it was an intimate mixture of solar molecules of various kinds in a very hot gaseous form, and the molecules were interacting upon one another at speeds inversely proportional to the square roots of their molec- ular weights. In the process of forming definite knots under self-control out of the less-defined solar outbursts, of which a considerable part was dissipated into planetesimals or fell back to the sun, the lighter molecules of high speed would be more likely to be dissipated than the heavy ones of lesser speed, but we need not insist on that, as the dispersing danger came from the projectile force and probably was not very selective. 2. But when that contingency was passed and each knot began its own independent evolution, there arose a very definite selective process within the knot itself. We assume that for a short time the knots would still be hot and diffuse, and that during this stage there would be larger chances for molecules to escape from the con- trol of the knot than later (a) because their velocities were highest on account of temperature, and (b) because their deployment was relatively open so that the molecules were less in one another’s way when they happened to accumulate velocity enough to escape from control. We assume that this was the most crucial stage of the knot, and that selective loss was then its greatest danger. The lightest molecules, because they had the highest mean speeds and most frequently encountered and divided energies with other molecules, were those that most often acquired cumulative speeds 154 T. C. CHAMBERLIN enough to enable them to escape. In each encounter there was an equi-partition of energy and the light molecules were given superior speeds in compensation for their lack of mass. The action was thus a highly selective process. But lest we seem to overstress this depleting process, let it be noted that for every reaction that gave exceptional speed to a light molecule there was a reaction in a counter-direction that gave to the heavier partner in the encounter a lower velocity. And further, it was only in the outer border of the knot that the lighter molecule rebounding outward could usually find a way of escape without another collision and a rebound in the wrong direction, and so the effect of the counter-reaction was to drive a heavier molecule inward for every case in which a lighter molecule escaped. This tended to herd in the heavier molecules and make their mutual attraction more effective, while they were inherently more amenable to control. There was a loss of mass, to be sure, but there was a more than compensating loss of dissipating activity and the residue was more congenial to control. Taking the knot as a whole, then, there was a steady progress toward a higher average of heavier molecules, and toward an as- semblage more amenable to control. Those knots which had been given masses enough to endure this process soon reached a stage of safety and then began to build up by capturing such planetesimals as they could control. Those knots that could not endure the process dispersed into planetesimals or erratic wanderers. The hypothesis, of course, assumes that the nuclei of our four bodies had original masses enough to live through this critical stage, as did also those of all the planetoids and satellites, but it favors the belief that the smallest planetoids and satellites represent the lower limit of successful knots, for if still smaller ones were successful we might expect to see their representatives in the heavens about us. The giant hot stars are, of course, the greatest known examples of success in holding light, hot gases by self-gravity. The multitude of these are our assurance that the principle of gaseous self-control is sound and that it has realization of oe highest order in the great cosmos. DIASTROPHISM AND THE FORMATIVE PROCESSES 155 It scarcely need be added that the severest selection of heavy molecules would be realized in the smallest knots where the struggle to maintain themselves was most strenuous, and that in the inter- mediate class the percentage of heavy molecules would be inversely proportional to mass. All this relates to the purely molecular state assumed to prevail while the knots were organizing themselves out of solar ejections and were beginning their careers as the nuclei of growth into mature planets, planetoids and satellites. 3. Let us turn now to the inner evolution af these nuclei. Let us recall that immediately on the emergence of the gas-masses from the sun there was great expansion and much cooling in con- sequence. Rapid radiation must have followed as the expanded '. mass shot out into the relatively cold space of the outer regions. It seems inevitable therefore that the condensation temperatures of the refractory material that now makes up most of the solid body of the earth, and doubtless of its neighbors, would be reached at a succession of stages relatively early in the history of the medium and smaller order of knots. We may assume that the condensation into minute spherules was started by electric charges and followed essentially the lines already sketched in the study of the derivatives from a rotating spheroidal nebula. There was this difference, however. The centrifugal derivatives from the rotating nebula were planetesimals each following its own orbit. The condensa- tions within the nuclei were, at first at least, scattered through the uncondensed portion that was still gaseous. The condensed spherules or crystals were like cloud particles or dust particles in an atmosphere. Dynamically they were like Brownian particles, jostled about by the impacts of the molecules that still remained in the gaseous state. They would naturally develop earliest in the outer parts of the nuclei and later in the inner parts. They would constitute a class of bodies heavier than the molecules and would tend to damp molecular action, while they themselves would tend to fall toward the center of the nuclei, but their fall would be re- sisted by the part that remained gaseous. It is obvious that the condensation of the refractory heavy material into spherules or 156 T. C. CHAMBERLIN | crystals in the midst of the gaseous molecules would mark the turn of the tide in the history of the smaller nuclei for the molecular losses would speedily grow less and a definite centralizing move- ment would set in which would increase the power of self-control. Molecular losses would be lessened and the capture of planetesimals increased relatively. The question of the temperatures and the physical states that would follow this stage in the smaller nucleiis important and difficult but must be deferred. 4. It remains only to consider the selective action of the success- ful nuclei in the process of gathering in planetesimals, but this need not detain us for the principles would be essentially those already emphasized sufficiently. The smaller order of nuclei could not capture and hold the lighter free molecules as such, though ~ they could perhaps capture and retain the very heavy molecules. They could quite certainly hold most of the planetesimal aggregates that they encountered but would have little power to draw them to themselves. The nuclei of medium mass could hold some of the free molecules but not the lightest, and so their accessions would be greater in mass, but lower in mean specific gravity. SUMMARY It seems clear, then, from the foregoing considerations that, 7 general, the planets, planetoids, and satellites, if built up by the planet- esimal method, would be composed of inherenily heavy material im inverse proportion to their masses, and hence that the inherent specific gravity of the matter of the moon would be somewhat greater than that of Mars, that of Mars somewhat greater than that of Venus, and that of Venus greater than that of the earth. There is still need to consider (1) what were-the physical states of the nuclei while they were gathering in the planetesimals, (2) what masses the planetesimals attained, and (3) what was the effect of their infall on the later stages of the growing bodies. This last will obviously involve the frequency of the fall of the planet- esimals upon the nuclei. The discussion of these points must be deferred to the next article. DIASTROPHISM AND THE FORMATIVE PROCESSES 157 It may be noted, however, that the physical state of the matter, whatever it may be, will not radically affect the mean constitution of the bodies, though it is liable to affect greatly the distribution of the material. Reserving judgment on any shrinkage effect that may arise from such difference of distribution, we may note that the inquiries of this paper, in harmony with the suggestions of the previous paper, very distinctly imply that, if the moon, Mars, Venus, and the earth were built up normally by the planetesi- mal method, they should contain proportions of inherently heavy material in the inverse order of mass. There is therefore cor- responding reason to think that the estimate of the total shrinkage of the earth deduced in the preceding article will need to be some- what increased, as anticipated, on account of the differences of material that make up the four bodies compared. A QUANTITATIVE MINERALOGICAL CLASSIFICATION OF IGNEOUS ROCKS—REVISED ALBERT JOHANNSEN University of Chicago PART II CLASS 2, ORDER I (210) Meso-silexite. See note under (110). (211) Moyite. The rocks of this family are quartz-ortho- granites. Since three of the rocks described by Daly* from the Moyie sill along the Forty-ninth Parallel are of this uncommon type, the term moyite is here suggested for the family name. (215) Rockallite Jupp. The name rockallite is derived from the Island of Rockall, off the coast of Ireland. Judd? describes the rock as composed of aegirite, quartz, and albite. While the presence of aegirite makes the rock abnormal, the term may be used, at least temporarily, for the family name. Rockallite is a quartz-rich aegirite-albite-tonalite. (216) Orthogranite. This is a comparatively rare rock, only differing from normal granite in containing no plagioclase. In this group also belong anorthoclase-granite, potash-rhyolite (orthorhyolite), comendite, and some grorudites and solvsbergites. Many of the granites belonging here are aegirite- or riebeckite- bearing. For the use of the prefix ‘‘ortho-” see (111). Microcline-granite. Anorthoclase-granite. Orthogranitite. t Reginald A. Daly, “‘Geology of the North American Cordillera at the Forty- ninth Parallel,” Geol. Surv. Canada, Mem. 38, Part I, pp. 229, 230, 231. 2 J. W. Judd, “‘On the Petrology of Rockall,” Trans. Royal Irish Acad. Dublin, XXII (1897), 49-57. 158 CLASSIFICATION OF IGNEOUS ROCKS 159 Orthorhyolite is the extrusive equivalent of this family. Comendite BERTOLI0' is an aegirite-rhyolite. Paisanite OSANN. Paisanite was named from Paisano Pass, Texas, by Osann.? It is a dike rock composed of quartz, microperthite, and riebechite. (217) Albite-granite. The term albite-granite is here applied to a normal granite containing orthoclase and plagioclase between the proportions 95:5 and 65:35. The plagioclase, however, is albite, and not the usual oligoclase. A new name should be chosen, since the terms soda-granite, albite-granite, albite-syenite, etc., have been used for rocks with albite as the only feldspar, that is, for rocks which properly are called albite-tonalite, albite- diorite, etc., in this classification. Albite-rhyolite. The extrusive equivalent of the pre- ceding. The objection to the term albite-granite applies also to albite-rhyolite. (218) Albite-adamellite. No confusion can result from the use of this term, since adamellite conveys the idea of a quartz- monzonitic rock, and the prefix indicates that the plagioclase of the orthoclase-plagioclase combination is albite. A number of grorudites fall here, as does also a lindoite, and numerous arfvedson- ite- and riebeckite-granites, so called. Albite-dellenite. The extrusive of the preceding. (219) Albite-granodiorite. There is no possibility of misunder- standing this name. See notes under (218) and (229). Albite-rhyodacite. See note under (229). (2110) Albite-tonalite. The meaning of this term also'is unmis- takable. See note under (217). For the use of tonalite for quartz- diorite see (2210). Albite-dacite. The extrusive of preceding. tS. Bertolio, ‘‘Sulle comenditi, nuovo gruppo di rioliti con aegirina,” Ait. della Reale Accad. dei Lincei, Cl. scienze fisiche, mat. e nat., IV (1895), 2 semestre, pp. 49-50. 2 A. Osann, “Report on the Rocks of Trans-Pecos, Texas,” gih Ann. Rept. Geol. Surv. Texas (1892), p. 132. 160 ALBERT JOHANNSEN (2111) Orthosyenite. This family includes syenites with ortho- clase, microcline, microperthite, or anorthoclase, but with less than 5 per cent plagioclase. Orthotrachyte. The extrusive equivalent of the pre- ceding. (2112) Albite-syenite. A term less likely to be misunderstood should be chosen for this family. See note under (217). Albite-trachyte. ‘See note'under (217). (2113) Albite-monzonite. This term cannot be misunderstood. See note under (218). Albite-latite. The extrusive equivalent of the preceding. (2114) Albite-monzodiorite. See note under (218). For the use of monzodiorite see (2214). Albite-andelatite. The extrusive equivalent of the preceding. See note under (2214). (2115) Albite-diorite. This term also is unmistakable. The word diorite conveys the impression of a plagioclase rock, and the prefix suggests that this feldspar is albite. The term soda- syenite has been used for this rock, but it was badly chosen. Soda- syenite naturally suggests a syenite rich in soda feldspar, but not to the exclusion of orthoclase. Albite-diorite is a much better term. Albite-andesite. The extrusive equivalent of the pre- ceding. (2116) Pulaskite Wittiams. This term is used in the sense of Williams” original definition: ‘‘a rock made up of ortho- clase, pyroxene, amphibole, and a little eleolite or its decom- position product, analcite.” In another place’ he says that the orthoclase is “similar to Brogger’s kryptoperthite, although the amount of soda is somewhat less than is usually found in this.” The rock, therefore, clearly falls into (2116), for while the dark constituents are not prominent they are greater than 5 per cent, as was shown by the examination of various thin sections from the type locality. Brégger’s type laurvikite belongs here, but the tJ. Francis Williams, ‘‘The Igneous Rocks of Arkansas,” Ann. Rept. Geol. Surv. Arkansas for r8go, II, 20. 2 Tbid., p. 60. 3W. C. Brégger, Die Eruptivgesteine des Kristianiagebietes. III: Das Gang- gefolge des Laurdalits (Kristiania, 1898), p. 30. CLASSIFICATION OF IGNEOUS ROCKS 161 characteristic texture of this rock excludes it from representing the family. (2121) Ortho-nephelite-syenite, ortho-leucite-syenite. This is nephelite- (or leucite-) syenite without plagioclase. Orthofoyaite. See note under (2222). Orthoditroite. See note under (2222). Orthophonolite. The extrusive equivalent of the pre- ceding. (2122) Albite-nephelite-syenite, albite-leucite-syenite. Nephe- lite-(leucite-) syenites carry orthoclase either as the only feldspar or in combination with plagioclase, which may be albite or some other member of the series. In this family (2122) belong those carrying albite. Albite-foyaite. See note under (2222). Albite-ditroite. See note under (2222). Albite-phonolite. The extrusive corresponding to the preceding. (2123) Albite-nephelite-monzonite. Here belong numerous tinguaites and so-called nephelite-syenites, also a sodalite-foyaite and a cancrinite-syenite. Covite WASHINGTON, according to the calculated mode, belongs here, but under the definition Washington’ says the rock is a “‘leucocratic combination of orthoclase and less nephelite with hornblende and aegirite-augite,”’ which would place it in Family 21. (2124) Litchfieldite Baytey. In his description of litch- fieldite Bayley? says: “‘As was indicated by the microscopic study, no plagioclase other than albite is present, and this, as seen (from a calculated mode), is largely in excess of the orthoclase.” The mode computed from the analysis indicates 27 per cent ortho- clase and microcline, 47 per cent albite, 17 per cent nephelite, and 2 per cent cancrinite. These proportions are about the same as shown by the thin section and Thoulet separation. While cancri- nite occurs in the type rock, Bayley’ says: “I do not regard * Henry S. Washington, ‘‘The Foyaite-Ijolite Series of Magnet Cove: A Chemical Study of Differentiation,” Jour. Geol., TX (1901), 615. 2W. S. Bayley, ‘Eleolite-Syenite of Litchfield, Maine,” Bull. Geol. Soc. Amer., TIT (1892), 242. 3 Ibid., In litteris, June 14, 1919. 162 ALBERT JOHANNSEN cancrinite as an essential constituent . . . . and many of the speci- mens show none. .... Some of it is certainly secondary.” The rocks of this family may therefore be divided into Litchfieldite (normal). Cancrinite-litchfieldite. | The calculated mode of canadite QUENSEL' places it in this family, but the actual mineral composition may be quite different. Quensel? says that ‘‘a rock entirely consisting of nephelite, albite, and a mafic mineral may or may not be a canadite, depending upon the presence or absence of a certain amount of normative lime- feldspar.” Depending thus upon chemical composition. for its name, it is excluded here. (2125) Mariupolite Morozewicz. Morozewicz’ described cer- tain rocks from the shores of the Sea of Azof which consist of much albite, less nephelite, and some ferromagnesian constituents and zircon: He gives two modal analyses, determined microscopically, both of which fall in this family somewhat above the center point. (2126) Naujaite Ussinc. Several nephelite- and _ leucite- syenites belong here, as does also arkite WASHINGTON.4 The latter, however, is too poor in feldspar to be typical of the family, and it carries garnet as an essential. Naujaite Ussinc® is the rock origi- nally described by Steenstrup® as a sodalite-syenite. Pending the location of a nephelite-rich orthofoyaite or orthoditroite, naujaite may represent the family. Syn.: Sodalite-syenite STEENSTRUP. Arkite WASHINGTON. t Percy Quensel, ‘‘The Alkaline Rocks of Almunge,” Bull. Geol. Inst. Upsala, XII (1914), 135, 176-77. 2 In litteris, May 16, 1915. 3 J. Morozewicz, “‘Uber Mariupolit, ein extremes Glied der Eleolithsyenite,” Tscherm. Min. Petr. Mitth., XXI (1902), 245. 4 Henry S. Washington, ‘‘The Foyaite-Ijolite Series of Magnet Cove: A Chemical Study in Differentiation,” Jour. Geol., TX (1901), 617. 5N. V. Ussing, “‘Geology of the Country around Julianehaab, Greenland,” Meddel. om Grénl., XX XVIII (1911), 32, 143-56. 6K. J. V. Steenstrup, ‘“‘Bemaerkninger til et geognostisk Oversigtskaart over en Del af Julianehaabs Distrikt (den 20 Juni 1878),” Meddel. om Grénl., II (1881), 35. CLASSIFICATION OF IGNEOUS ROCKS 163 (2127) Beloeilite. O’Neill' described a ‘“‘feldspathic tawite’’ from Mount St. Hilaire (Beloeil), Quebec, intermediate in composi- tion between sodalite-syenite and tawite. This rock may be called beloeilite and may serve as the type of the sodalite rocks of the family. (21209) A soda-sussexite of Hackman? and a nephelite-sodalite- syenite of O’ Neill belong here. (2130) Toryhillite. Adams and Barlow‘ described a nephelite- rich albite rock from Toryhill, Monmouth Township, Ontario- The rock falls near the center point of this family; therefore a new name is here suggested. (2131) Urtite Ramsay. Urtite is.a feldspar-free nephelite rock, described by Ramsay.’ The name is derived from the second part of Lujavr-Urt, where the rock was first found. It consists of 82 to 86 per cent nephelite, 12 to 18 per cent aegirite-augite, and 2 per cent apatite. Ijolite Ramsay and BERGHELL,* with 51.6 per cent nephelite, and monmouthite ApAMs and Bartow,’ with 77.6 per cent feldspathoid (of which 72 is nephelite), also belong here. The leucite rock of this family is fergusite Prrsson.6 Only a single occurrence is known. It consists of 65 per cent pseudoleucite and 35 per cent mafites, chiefly diopside. The melilite rock of the family is represented by uncompahgrite LARSEN? It is a tJ. J. ONeill, ‘St. Hilaire (Beloeil) and Rougemont Mountains, Quebec,’’ Geol. Surv. Canada, Mem. 43 (Ottawa, 1914), p. 46. 2V. Hackman, ‘“‘Neue Mitteilungen iiber das Ijolithmassiv in Kuusamo,” Bull. d.1. Comm. Geol. d. Finlande, No. 11 (Helsingfors, 1899), p. 24. 3 J. J. O'Neill, op. cit., p. 41. 4 Frank D. Adams and Alfred E. Barlow, ‘‘Geology of the Haliburton and Ban- croft Areas, Province of Ontario,”’ Geol. Surv. Canada, Mem. 6 (1910), p. 273. 5 Wilhelm Ramsay, “Urtit, ein basisches Endglied der AuiisyertNepielasy a Serie,” Geol. Foren. i. Stockh. Férhandl., XVIII (1896), 463. 6 Wilhelm Ramsay and Hugo Berghell, “‘Das Gestein vom Jiwaara in Finland,” Geol. Foren. i. Stockh. Forhandl., XIII (1891), 304; also V. Hackman, op. cit., supra, Pp. 20. 7 Frank D. Adams and Alfred E. Barlow, op. cit., p. 277- 8 Louis Valentine Pirsson, ‘‘Petrography and Geology of the Igneous Rocks of the Highwood Mountains, Montana,” U.S. Geol. Surv., Bull. 237 (1905), p. 88. 9 Esper S. Larsen, “‘Melilite and Other Minerals from Gunnison County, Colo- rado,” Jour. Wash. Acad. Sci., IV (1914), 473- 164 ALBERT JOHANNSEN deep-seated rock consisting of two-thirds melilite and one-third dark constituents (magnetite and pyroxene) with some accessory perofskite and apatite. The extrusive equivalent is melilite- basalt STELZNER' (Syn. Melilithit Lozwrnson-LeEssiINc?). CLASS 2, ORDER 2 (222) Quartz-granite. Since the term granite in itself carries the idea of a quartz-bearing rock, the term quartz-granite will indicate a granite that is rich in quartz. All other rocks which, by definition, carry quartz may be similarly qualified, namely, adamellite, granodiorite, tonalite, etc. (223) Quartz-adamellite. See note under (222). Here belongs, though far from the center point, the Hauksuo, Kisko, aplite, described by Eskola,3 which consists of quartz 48.9 per cent, plagioclase (average Abs;An;;) 22.9 per cent, microcline 20.4 per cent, biotite 4.4 per cent, magnetite 1.9 per cent, and epidote 1.5 — per cent. (224) Quartz-granodiorite. See note under (222). (225) Quartz-tonalite. See note under (222). For the use of tonalite for quartz-diorite see (2210). : (227) Granite. This term is of very old date. It is not found in Pliny. Breislak says it was first used by Caesalpinus? in 1596. The name may be derived from the Italian granito, ‘“‘grained”’ (Lat. granum), but its origin is uncertain. The word is similar in sound in many languages, for example, gwenith faen (“wheat stone’’) in Welsh, and it is possible that the name was brought from Wales by the Romans who built roads and worked the mines there about 78 A.D. Family (227) is that of the normal granites. These rocks consist of quartz, orthoclase, less oligoclase or andesine, and a moderate « Alfred Stelzner, “‘Mittheilungen an den Redaction,” Newes Jahrb., I (1882), 230-31; also “Uber Melilith und Melilithbasalte,” Neues Jahrb., B.B., I (1882), 364-440. 2 F. Loewinson-Lessing, “‘Kritische Beitrage zur Systematik der Eruptivgesteine, TVET PU PxeXe (GO) wats 3 Pentti Eskola, “‘On the Petrology of the Orijarvi Region in Southwestern Finland,” Bull. d.1. com. géol. d. Finlande, No. 40 (1914), p. 83. 4 Andreas Caesalpinus (Cesalpino), De Metallicis (1596), II, cap. 11. CLASSIFICATION OF IGNEOUS ROCKS 165 amount of mafic minerals. There are many subfamilies, depending upon the accessories or dark constituents. Granitite. This is an old term which has been used with a great variety of meanings. So long ago as 1823 von Leonhard'* said that the proposal to call granites with accessory minerals granitite was superfluous and unfitting. Rose? used the word for muscovite-free biotite-bearing rocks with quartz, orthoclase, and considerable oligoclase. It was used in the same sense by Senft* and Rosenbusch,‘ and is here also used as a synonym for biotite- granite. Cathrein’s’ suggestion that it be used for plagioclase-rich hornblende-bearing granites has never been followed. Amphibole-granite. Arfvedsonite-granite BROGGER.® Augite-granite. Binary-granite KryEs’ contains both muscovite and biotite. It is also called two-mica granite. Hypersthene-granite. Syn.: Charnockite HoLianp.® Microcline-granite. Muscovite-granite. Tourmaline-granite, etc. Rhyolite von RIcHTHOFEN. The name rhyolite (from pba, “lava stream,” “torrent”) was given by von Richthofen? to the extrusive equivalent of granite on account of its usual fluidal =K. C. von Leonhard, Charakteristik der Felsarten (Heidelberg, 1823), I, 54. 2 Gustav Rose, ‘Ueber die zur Granitgruppe gehérenden Gebirgsarten,” Zeitschr. d. d. geol. Ges., I (1849), 386; also pp. 363-66. 3 Ferdinand Senft, Classification und Beschreibung der Felsarten (Breslau, 1857), p- 297. 4H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (1st ed.; Stuttgart, 1877), II, 20-21. 5A. Cathrein, “Zur Diinnschlifisammlung der Tiroler Eruptivgesteine,” Newes Jahrb., I (1890), 72. 6 W. C. Brégger, “Die Mineralien der Syenitpegmatitginge der siidnorwegischen Augit- und Nephelinsyenite,” Zezt. f. Kryst., XVI (1890), 68. 7Charles Rollin Keyes, “Origin and Relations of Central Maryland Granites,” U.S. Geol. Surv., Ann. Rept., XV (1895), 714. 8T. H. Holland, “The Charnockite Series, a Group of Archean Hypersthenic Rocks in Peninsular India,” Mem. Geol. Surv. India, XXVII (1900), 2, 119. 9 Ferdinand Freiherrn v. Richthofen, “Studien aus den ungarisch-siebenbiirgeri- schen Trachytgebirgen,” Jahrb. d. k. k. geol. Reichsanst., XI (1860), 156. 166 ALBERT JOHANNSEN character. Roth' called the same rocks liparites, from the Lipari Islands, where they occur. Syn.: Liparite Rot. Nevadite von RICHTHOFEN. Nevadite is a variety of rhyolite with very abundant phenocrysts and very little ground- mass. It was named by von Richthofen? from its occurrence in Nevada. (228) Adamellite CATHREIN-BROGGER. The term adamellite is here given to quartz-monzonites. It was originally used by Cathrein? for rocks from Adamello which contain both orthoclase and plagioclase. Speaking of these rocks he says: ‘Er ist ein lichtes, schon im Aussehen mehr den Graniten als Dioriten sich niherndes Gestein.” Brogger* uses the name for acid quartz- monzonites, and Ball5 and Hatch® use it for quartz-monzonite. Syn.: Quartz-monzonite BROGGER. Windsorite Daty.? This is a quartz-poor quartz- monzonite dike-rock. Dellenite Br6ccER. The rock intermediate between dacite and rhyolite from Dellen, Helsingland, Sweden, was named dellenite by Brégger,® and the term is here used for the extrusive equivalent of quartz-monzonite: Since it consists of a single word it is preferable to quartz-latite. Syn.: Quartz-latite RANsomME,? Dacite-liparite BROGGER.” Justus Roth, Gesteinsanalysen, xxxiv (1861). 2 F, Baron von Richthofen, “‘ Principles of the Natural System of Volcanic Rocks,” Mem. Cal. Acad. Sci., I (1868), Part II, p. 16. 3A, Cathrein, “Zur Diinnschliffsammlung der Tiroler Eruptivgesteine,” Neues Jahrb., I (1890), 74. 4W.C. Broégger, Die Eruptivgesteine des Kristianiagebietes. II: Die Eruptionsfolge der triadischen Eruptivgesteine bei Predazzo in Stidtyrol (Kristiania, 1895), p. 6. s Sydney H. Ball, “General Geology of the Georgetown Quadrangle, Colorado,” U.S. Geol. Surv. Wash., Prof. Paper 63 (1908), p. 51. 6F.H. Hatch, The Petrology of the Igneous Rocks (London, 1914), p. 176. 7R. A. Daly, “The Geology of Ascutney Mountain, Vermont,” U.S. Geol. Surv., Bull. 209 (1903), p. 46. 8 W. C. Brégger, Die Eruptivgesteine des Kristianiagebietes. II: Die Eruptionsfolge der triadischen Eruptivgesteine bei Predazzo in Stidtyrol (Kristiania, 1895), p. 59, note. 9 F. Leslie Ransome, “Some Lava Flows of the Western Slope of the Sierra Nevada, California,” Amer. Jour. Sci., V (1898), 372. 10 W. C. Brégger, op. cit., p. 60. CLASSIFICATION OF IGNEOUS ROCKS 167 (229) Granodiorite BECKER. The term granodiorite was sug- gested by Becker™ for plutonic rocks containing both orthoclase and acid plagioclase, the latter in excess of the former. The name was first published in a paper by Lindgren? in 1893 and described in more detail later. Lindgren* says: ‘‘The truly characteristic feature of the granodiorites is that the soda-lime feldspar, which always is a calcareous oligoclase or an andesine, is at least equal to double the amount of the alkali feldspar. The latter may be taken to vary from 8 per cent to 20 per cent” in a rock with an assumed feldspar content of 60 per cent.5 ‘‘ Below the lower limit the rock becomes a quartz-diorite: above the upper a quartz- monzonite.” ‘In the quartz-monzonite,” he continues, “I would give this mineral a range from 20 per cent to 4o per cent,” again assuming a total of 60 per cent feldspar. His divisions, therefore, based on the orthoclase-plagioclase ratio are 0-133-333-0063 for quartz-diorite, granodiorite, and quartz-monzonite. In the present classification the divisions are taken at o-5-35-65, the writer believing that more than 5 per cent of orthoclase changes the char- acter of a quartz-diorite too much for it to retain that name. Setting apart as orthogranite the rock with 95 to roo per cent of its feldspar orthoclase, and as quartz-diorite that -having 95-100 per cent plagioclase, the remaining 90 parts are divided into three t Waldemar Lindgren, ‘‘Granodiorite and Other Intermediate Rocks,” Amer. Jour. Sci., TX (1900), 270. 2 Tbid., “The Auriferous Veins of Meadow Lake, California,” Amer. Jour. Sci., XLIV (1893), 202. 3 Sacramento Folio, U.S. Geol. Surv., No. 5, 1894; Placerville Folio, No. 3, 1894; Smartsville Folio, No. 18, 1895; Nevada City Folio, No. 29, 1896; Pyramid Peak Folio, No. 31, 1896; Truckee Folio, No. 39, 1897; H. W. Turner, “The Rocks of the Sierra Nevada,” U.S. Geol. Surv., An. 14, pp. 478, 482; Waldemar Lindgren, ‘The Gold- Silver Veins of Ophir, California,’ An. 14, U.S. Geol. Surv., Part II (1804), pp. 252, 255-56; also “The Gold Quartz Veins of Nevada City and Grass Valley Districts, California,” An. 17, U.S. Geol. Surv., Part II (1896), p. 35; “‘The Granitic Rocks of the Pyramid Peak District, Sierra Nevada, California,’ Amer. Jour. Sci., III (1897), 308; ‘The Granitic Rocks of Sierra Nevada,” Abstr. in Science, New Series, V (1897), 361; ‘Granodiorite and Other Intermediate Rocks,’ Amer. Jour. Sci., IX (1900), 269-82. 4 Waldemar Lindgren, “Granodiorite and Other Intermediate Rocks,” Amer. Jour. Sci., TX (1900), 277. 5 [bid., p. 279. 168 . ALBERT JOHANNSEN equal portions, one each for normal granite, quartz-monzonite, and granodiorite. The term granodiorite is not the most satisfactory, from the standpoint of the construction of the word, for a rock inter- mediate between quartz-monzonite (adamellite) and quartz-diorite (tonalite), for it suggests a rock intermediate between a granite (an orthoclase-quartz-bearing rock) and a diorite (a plagioclase-quartz- free rock), that is, a quartz-monzonite. Monzonite in the sense in which we now use it, however, was not introduced by Brogger until 1895, while granodiorite was first used in 1892 or 1893. The term. was intended to convey the idea of a diorite with granitic characters, that is, with quartz and a certain amount of orthoclase, but upon the introduction of the term monzonite Lindgren’ found it advis- able to restrict the “definition of granodiorite to rocks considerably nearer quartz-diorite than originally intended.” A name derived from the rocks between which granodiorite, as now defined, actually stands would give adam-tonalite, which is hardly euphonious, to say the least. Furthermore, the term granodiorite is so firmly estab- lished and was so clearly defined that it should not be changed. The term banatite was used by Brégger? for his intermediate quartz-monzonites, and under this term are included rocks which are probably to be classed as quartz-poor granodiorites (see under 2213). The objection to the construction of the word granodiorite as applied to rocks under the present definition applies also to the terms introduced by the present writer’ in 1917, namely grano- gabbro, syenodiorite, and syenogabbro. But since granodiorite is retained for the reasons stated above, and adam-gabbro is as objectionable as adam-diorite, granogabbro (239) also will be retained to make analogous terms. The names of the extrusive rocks should naturally conform in construction to their deep-seated equivalents; consequently rhyodacite WINCHELL is used for the t Waldemar Lindgren, In litteris, June 17, 1919. 2W.C. Briégger, Die Eruptivgesteine des Kristianiagebietes. II: Die Eruptionsfolge der triadischen Eruptivgesteine bei Predazzo in Siidtyrol (Kristiania, 1895), p. 60. 3 Albert Johannsen, “Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,” Jour. Geol., XXV (1917), p- 80. CLASSIFICATION OF IGNEOUS ROCKS 169 granodioritic extrusive, and rhyobasalt is here proposed for the granogabbroic. Syenodiorite and syenogabbro will be replaced by the better terms monzodiorite and monzogabbro. For the corre- sponding extrusives, ande-latite and basa-latite are here proposed. Rhyodacite WINCHELL.* See note under granodiorite above. ‘This is the extrusive equivalent of granodiorite. (2210) Tonalite vom RaTH-BROGGER-SPURR. Tonalite was ap- plied by vom Rath? to a certain rock from Monte Tonale, in the Tyrol. It consists of abundant quartz, andesine, very small amounts of orthoclase as an accessory, and a dark mineral. In the definition, however, vom Rath omitted the orthoclase, which occurs as micropegmatite and was regarded simply as an accessory, and said: “Der Tonalit enthalt in kérnigem Gemenge als wesentliche Bestandtheile: eine trikline Feldspath-Species, Quarz, Magnesia- glimmer und Hornblende.” In his tabulation Brégger? used tona- lite for acid quartz-diorite, and in another place’ he said: ‘‘Der Name Tonalit ware dann den Quarzdioriten vorbehalten.’ Since Brogger does not use the subdivision granodiorite, his tonalite includes some of these rocks. Simply as quartz-diorite the term was used by Spurr,> who adopted it as a group name instead of quartz-diorite, because it “‘is shorter and more characterizing.” In the same sense it was used by Winchell® and Hatch,’ and it is so used here. Incidentally, the average igneous rock of Clarke,® computed into common minerals, falls in this family. It contains quartz * Alexander N. Winchell, “Rock Classification on Three Co-ordinates,” Jour. Geol., XXI (1913), 214. 2G. vom Rath, “Beitrige zur Kenntniss der eruptiven Gesteine der Alpen,’ Zeitschr. d. d. geol. Ges., XVI (1864), 250. W.C. Brogger, Die Eruptiugesteine des Kristianiagebietes. II: Die Eruptionsfolge der triadischen Eruptivgesteine bet Predazzo in Siidtyrol (Kristiania, 1895), p. 60. 4 Ibid., p. 61. SJ. E. Spurr, “Reconnaissance in Southwestern Alaska in 1898,” U.S. Geol. Surv., Ann. Rept., XX, Part VII (1900), p. 190. 6 Alexander N. Winchell, “‘Rock Classification on Three Co-ordinates,” Jour. Geol., XXI (1913), 214. 7F. H. Hatch, The Petrology of the Igneous Rocks (London, 1914), p. 195. 8 F. W. Clarke, “Data of Geochemistry,” U.S. Geol. Surv., Bull. 491 (1911), p. 31. 170 ALBERT JOHANNSEN 12 per cent, andesine 59.2 per cent, pyribole 16.8 per cent, mica 3.8 per cent, and other accessories 7.9 per cent. Syn.: Quartz-diorite. Dacite STAcHE. The name dacite was given by Stache" to certain older quartz-trachytes, found in Siebenbiirgen (Dacia), which contain oligoclase and pyroxene and which were thus thought to be separated from the younger quartz-trachytes or rhyolites which contain sanidine and biotite. The term has been variously used by different petrographers, some applying it only to extrusive plagioclase rocks with quartz phenocrysts, others includ- ing rocks whose only quartz is in the groundmass; some requiring the presence of pyroxene, others including rocks with biotite. In general the name has come to mean quartz-andesite, the extrusive equivalent of quartz-diorite, and it is so used here. Syn.: Quartz-andesite. (2212) Syenite. The name syenite was used at least as long ago as the time of Pliny,? and was apparently in common use in his day for the rocks from Syene (Assuan), Egypt. “Circa Syenen vero Thebaidis Syenites,” he says, ‘““que ante pyrrhopoecilon voca- bant,” that is, spotted with red. Werner? applied the term Sienit to the hornblende-bearing rock from near Dresden, under the impression that the two were the same. It was later found that the Syene rock contains quartz, is, in fact, hornblende-granite, while the Dresden rock is practically quartz-free. Some writers conse- quently used the term syenite for hornblende-granite, but gener- ally it was used for quartz-free orthoclase rocks in the sense of Werner. Rosiére* found that the rock of Mount Sinai is of the quartz-free variety and therefore proposed that the name be changed to sinaite. The older term, however, was in such common use that this suggestion was not followed. ; 1G. Stache, “‘Quarztrachyte Siebenbiirgens,” in Fr. von Hauer and G. Stache, Geologie Siebenbiirgens (1863), pp. 56, 70, 79. 2C. Plinii Secundi Naturalis historiae xxxvi. cap. viii. In the edition Lugd. Batav. Roterodami, Ao. 1668, p. 647. 3 “‘Vermischten Nachrichten,” Bergmdnnisches Journal, II (1788), 824. 4J. F. d’Aubuisson de Voisins, Traité de géognosie (Strasbourg et Paris, 1819), Ds 2, CLASSIFICATION OF IGNEOUS ROCKS Lt There are many varieties of syenite, depending upon the char- acter of the dark or accessory constituents. — Biotite syenite. Syn.: Syenitite Lo—EwInson-LESSING,’ analogous to granitite for biotite-granite. Polenov,? however, had previously used the term in a different sense, namely for syenite-aplite. Hornblende-syenite. Arfvedsonite-syenite. Augite-syenite. Epidote-syenite. Pyroxene-syenite. Hypersthene-syenite (Syn.: Mangerite), etc. Trachyte Haty. The term trachyte, from the Greek Tpaxvs, on account of the rough appearance of the rock, was intro- duced by Haiiy? in his lectures at the Jardin des Plantes to char- acterize the rocks from the Drachenfels on the Rhine. Although the feldspar in these rocks had previously been determined and named sanidine by Nose,’ yet plagioclase rocks of the general appearance of the original trachyte came to be included in the term. Later, von Buch’ separated certain Andean extrusives with plagio- clase from this group and called them andesites. At the present time trachyte is used for the extrusive equivalents of syenite, that is, for rocks consisting of potash-feldspar, a little plagioclase, and a biopyribole. Many trachytes are orthotrachytes (2111). (2213) Monzonite De LApPPARENT-BROGGER. The principal rock of Monzoni, in the Tyrol, is of quite variable character. It * F. Loewinson-Lessing, “‘Kritische Beitrige zur Systematik der Eruptivgesteine, II,” Tscherm. Min. Petr. Mitth., X1X (1900), 173. 2B. Polenov, ‘“‘ Die massigen Gesteine von nordlichen Theile des Witim-Plateau,” Travaux de la Société Impériale des Naturalistes de St. Pétersbourg, XXVII (1899), livr. 5, 464. 3.Haiiy did not publish the name until 1822 in his Traité de minéralogie (2d ed.; Paris, 1822), IV, 579, but it was accredited to him as far back as 1813 in Alexandre Brongniart’s ‘‘Essai d’une classification minéralogique des roches mélangées,” Jour. d. Mines, XXXIV (1813), 43. 4K. W. Nose, Orographische Briefe, I, 26, 113. 5 Leopold von Buch, “Ueber Erhebungscratere und Vulcane’’ (read in Berlin Akad., March 26, 1835), Pogg. Ann., XXXVII (1836), 190. 17D . ALBERT JOHANNSEN was called Monzon-syenit by von Buch,’ and later monzonit by De Lapparent.? The latter term served as a collective name for some years, but in 1881 Reyer used it as a specific name for rocks of the Monzoni type, which he thought to be augite-syenite. Earlier Tschermak* had recognized the principal rock to be’ an orthoclase-plagioclase rock, but he applied the term monzonite to the whole series. Lemberg® characterized the rock as inter- mediate between syenite and diorite. In those days rocks were classified either as syenites or diorites, and it was not until 1895 that an intermediate family was established. In that year Brogger® introduced the monzonite family. He says: ‘Es ist .... nach meiner Ansicht nothwendig, zwischen den Orthoklasgesteinen und den Plagioklasgesteinen . . . . eine Ubergangsordnung von Alkali- feldspath-Kalknatronfeldspath-Gesteinen einzuschieben.” The plagioclase of the Monzoni rocks, according to Brégger, is generally andesine to labradorite, or even anorthite.’ Rosenbusch® says that andesine is much rarer in all monzonites than more basic plagioclase, and more acid plagioclase than andesine had not been observed by him in the rocks of Monzoni. Elsewhere, it would seem, orthoclase- acid-plagioclase rocks are more common than those with basic plagioclase. In this classification, therefore, the more acid rocks, that is, those whose feldspar is oligoclase or andesine (sodi- monzonites), will be considered normal monzonites, while those with labradorite or bytownite will be melas under the calci- monzonites (2313). tLeopold von Buch, ‘‘Ueber geognostische Erscheinungen im Fassathale,” v. Leonh. Mineralogisches Taschenbuch fiir das Jahr 1824, p. 347- 2M. de Lapparent, “‘La constitution géologique du Tyrol, méridional,” Am. d. Mines, 6 sér., VI (1864), 250. 3 Ed. Reyer, ‘‘Predazzo,” Jahrb. d. k. k. geol. Reichsanst., XXXI (1881), 36. 4 Gustav Tschermak, Die Porphyrgesteine Osterreichs (1869), p. 110. 5 J. Lemberg, “Uber die Contactbildungen bei Predazzo,’’ Zeitschr. d. d. geol. Gesell., XXIV (1872), 188, 190. 6W. C. Brégger, Die Eruptivgesteine des Kristianiagebietes. IT: Die Dine. folge der triadischen Eruptivgesteine bei Predazzo in Siidtyrol (Kristiania, 1895), Pp. 22-23. 7 Op. cit., p. 54. ' 8H. Rosenbusch, Elemente der Gesteinslehre (Stuttgart, 1808), p. gage also Mikro- skopische Pissing apie der massigen Gesteine (Stuttgart, 1907), II, 167. CLASSIFICATION OF IGNEOUS ROCKS ie _ Certain intrusive rocks in Banat, cutting through limestones and crystalline schists, were named banatites by von Cotta. They are usually quartz-bearing, and are intermediate between quartz-diorites, quartz-augite-diorites, diorites, and augite-diorites. Brogger? applied the name to quartz-bearing monzonites, inter- mediate between normal monzonites and adamellites, and related to monzonites as quartz-syenites are to syenites. The extrusive equivalent of banatite BROGGER is quartz-trachyte-andesite BRrOGGER? (see note under 2209). Olivine-monzonite. While the Monzoni __ olivine- monzonites carry basic plagioclase, elsewhere olivine-monzonites | with andesine are found. ‘These rocks, therefore, fall in the present subfamily. Kentallenite is stated by Hatch‘ to be “‘iden- tical with Brégger’s olivine-monzonite. ... . The two feld- spars are present in roughly equal proportions.” As a matter of fact, Hill and Kynastons distinctly state that “‘the term (mon- zonite) has come to be associated with the presence of an approxi- mately equal amount of the two feldspars—a feature which cannot be said to be an essential characteristic of our group.” From the-variation in the various specimens described, it would seem that _ kentallenite is an olivine-orthoclase-plagioclase rock, the ratio of the feldspars being quite variable, consequently not limited to Family 13 here. Latite RANsom. The name latite, from the occurrence of such rocks in the Italian province of Latium, was used by Ransom‘* for the extrusive equivalents of the monzonites. Syn.: Trachyte-andesite. (The term trachy-andesite has been used in a different sense from trachyte-andesite by some writers, though others make it synonymous. Rosenbusch’ uses tB. von Cotta, Erzlagerstatien im Banat und in Serbien (Wien, 1864). 2W. C. Brogger, op. cit. (II, 1895), p. 61. 3 Ibid., p. 60. 4F. H. Hatch, The Petrology of the Igneous Rocks (London, 1914),.pp. 206-7. 5 J. B. Hill and H. Kynaston, “On Kentallenite and Its Relation to Other Igneous Rocks in Argyllshire,” Quart. Jour. Geol. Soc., LVI (London, 1900), 532. ° F. Leslie Ransome, “Some Lava Flows of the Western Slope of the Sierra Nevada, California,” Amer. Jour. Sci., V (1898), 372. 7H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (Stuttgart, 1908), II, 1036-37. iA: . ALBERT JOHANNSEN it, by analogy to trachydolerite, for an extrusive rock whose deep- seated equivalent would be intermediate between alkali-syenite and essexite. It represents, consequently, a rock carrying a feld- spathoid or aegirite, riebeckite, etc., and not, as one would suppose from the name, a rock intermediate between trachyte and andesite.) (2274) Monzodiorite. The term syenodiorite was proposed by the writer’ for the quartz-free equivalent of granodiorite. For the reasons stated under granodiorite (229), this term is withdrawn and monzodiorite is substituted for it. | Syn.: Syenite-diorite BROGGER.? Andelatite. For the extrusive equivalent of monzo- diorite, the term andelatite, as intermediate between andesite and latite, is suggested. See note under granodiorite (229). Mugearite Harker. Mugearite, from the village of Mugeary, is the name given by Harker? to certain extrusive rocks resembling basalt but with oligoclase (Ab,;An,) and much olivine. A modal analysis shows the rock to consist of oligoclase 573 per cent, orthoclase 123 per cent, olivine, iron ore, and augite (augite quite subordinate to olivine) 263 per cent, and apatite 33 per cent. It is therefore an olivine andelatite. (2215) Diorite Hatty. The term diorite (from dupitw, “dis- tinct”) was introduced by Haiiy* as a substitute for Werner’s term Griinstein. The name now stands for a rock consisting of acid plagioclase and a dark mineral. Syn.: Griinstein WERNER, Diabase BRONGNIART. Oligoclase-diorite. Andesine-diorite. Andesite von BucH. Formerly there were included, under the name trachyte, not only orthoclase rocks, but those t Albert Johannsen, “‘Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,” Jour. Geol., XXV (1017), 80. 2W. C. Brégger, “‘Die Mineralien der Syenitpegmatitgiange der siidnorwegischen Augit- und Nephelinsyenite,” Part I, Zeit. f. Kryst., XVI (1890), 49. 3 Alfred Harker, ‘‘The Tertiary Igneous Rocks of Skye,” Geol. Surv. Mem. (1904), p. 265; John S. Flett, ““On the Mugearites,” Summary of Progress, Geol. Surv. (1907), p. 110. 4 The term first appeared in the publications of some of Haiiy’s students. Thus see J. F. d’Aubisson de Voisins, Traité de géognosie (Strasbourg and Paris, 1819), p. 146. Later it was used in Abbé Haiiy’s Traité de minéralogie (Paris, 1822), IV, 540. CLASSIFICATION OF IGNEOUS ROCKS 175 containing plagioclase as well. See note under trachyte (2212). In 1835 von Buch’ described certain volcanic rocks from the Andes, consisting of plagioclase (originally thought to be albite) and horn- blende. He named them andesites. Later the term was applied to oligoclase- or andesine-bearing rocks, and it is used in this sense here. The dark mineral may be biotite, hornblende, or augite, or combinations of these, which thus give subfamilies. Biotite-andesite. Mica-andesite. Hornblende-andesite. Syn.: Hungarite LANG,’ named from their wide dis- tribution in Hungary. Augite-andesite. See note under basalt (2315). Hypersthene-andesite BECKE. Syn.: Santorinite BEcKE. Becke* proposed the term santorinite for acid- and alboranite for basic rocks of Santorin. The former are hypersthene-andesites rich in soda (Na:Ca>2). The phenocrysts are labradorite with mantles of oligoclase, while the groundmass is acid oligoclase. The average feldspar, therefore, _ is acidic. Washington’ had previously called the acid rocks pyroxene-andesites and had proposed the term santorinite for the members carrying basic plagioclase. Santorinite, used with two meanings, should therefore be abandoned. (2217) Nephelite- (leucite-) bearing syenite. (2218) Nephelite- (leucite-) bearing monzonite. (2219) Nephelite- (leucite-) bearing monzodiorite. (2220) Nephelite- (leucite-) bearing diorite. (2222) Nephelite-syenite RoseENBUSCH. Rosenbusch’ included, tLeopold von Buch, ‘‘Ueber Erhebungscratere und Vulcane” (read in Berlin Akad., March 26, 1835), Pogg. Ann., XXXVII (1836), 190. 2 Heinr. Otto Lang, Grundriss der Gesteinskunde (Leipzig, 1877), p. 196. 3 F. Becke, ‘‘Der Hypersthen-Andesit der Insel Alboran,” Tscherm. Min. Petr. Mitth., XVIII (1899), 553. AWibide Das 53. 5 Henry S. Washington, “‘Italian Petrological Sketches: V. Summary and Con- clusions,”’ Jour. Geol., V (1897), 368. 6H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (1st ed.; Stuttgart, 1877), II, 204. 176 ALBERT JOHANNSEN under the names Eleolith-Syenit and Nephelin-Syenit, rocks which had previously been described as zircon-syenite, foyaite, miascite, and ditroite. If these terms were not acceptable he suggested that all of the rocks of this class be included under foyaite, since the rocks from Mount Foya, in the Serra de Monchique, province of Algarva, Portugal, as originally described by Blum,’ are most nearly representative of the whole group. The term eleolite has fallen away, more or less, since the age element in mineralogy is no longer considered, and in general, in this country, the rocks have been called nephelite-syenites. Brégger? proposed to subdivide the nephelite-syenites into two groups according to texture. For rocks of this composition and with trachytoid texture he used Blum’s term foyaite, and for those with granitic texture Zirkel’s3 term ditroite. In the present classification, under the term nephelite-syenite, are included those nephelite-bearing rocks in which orthoclase exceeds acid plagioclase (oligoclase or andesine) in amount. The rocks in which the plagioclase is albite are here called albite-nephelite-syenites (2122), and those which contain no _ plagioclase, ortho-nephelite-syenites (212). Foyaite BLUM-BROGGER. Ditroite-ZIRKEL-BROGGER. Syenoid SHanp. As an abbreviation for feldspathoid- syenite, Shand? used syenoid, and said the term would “imply the presence of nephelite.”” - He further suggested that gabbroid, dioroid, and doleroid might be advantageously coined in the same manner. Syenoid could well be used as a family name to include both nephelite and leucite rocks. Leucite-syenite. tR. Blum, ‘“Foyait, ein neues Gestein aus Siid-Portugal,” Neues Jahrb. (1861), p. 426. 2W. C. Brogger, op. cit., Zeitschr. f. Kryst., XVI (1890), pp. 39, 125; also Die Erupiiogesteine des Kristianiagebietes. III: Das Ganggefolge des Laurdalits (Kris- tiania, 1898), pp. 164-65. 3 F. Zirkel, Lehrbuch der Petrographie (1st ed.; Bonn, 1866), I, 595. The term is derived from Ditré, in eastern Siebenbiirgen, Transylvania. 4S. J. Shand, “On Borolanite and Its Associates in Assynt,’’ Trans. Edinburgh Geol. Soc., UX (1910), 377. CLASSIFICATION OF IGNEOUS ROCKS 77 (2223) | Nephelite-monzonite. Under this name Lacroix’ de- scribed rocks from Madagascar consisting essentially of anortho- clase, basic labradorite, nephelite, titaniferous pyroxene, bark- evikite, titaniferous magnetite, and apatite. The mineral pro- portions are’ quite variable. To make it conform with the usual monzonite group, the term nephelite-monzonite is here adopted for a nephelite rock of Class 2 with acid plagioclase about equal in amount to the other feldspars. The rock described by Lacroix is included under the calci-nephelite-monzonite family. Nephelite (leucite-)latite. The extrusive equivalent of the preceding. (2224) Nephelite-(leucite-)monzodiorite. A term here sug- gested for nephelite-(leucite-)bearing rocks comparable to grano- diorites among the quartz-bearing. (2225) Nephelite-(leucite-)diorite. A term, analogous to nephelite-syenite, here suggested for nephelite-acid plagioclase rocks. Under this family would fall Adams and Barlow’s’ rag- lanite, an abnormal rock on account of its large content of corundum, 4.45 per cent. Since it carries only 12 per cent by weight of nephelite, it is also far from the center point. Nephelite-(leucite-)tephrites and nephelite-(leucite-) basanites belong here in part. (2229) No plutonic rock has been located in this family, but there is an extrusive, a leucite-tephrite from the Roman Comag- matic region, described by Washington. (2230) AA leucitite from the Roman Comagmatic region, de- scribed by Washington,’ is the only rock yet found here. [To be concluded] tA. Lacroix, “‘Sur les granites et syénites quartziféres 4 aegirine, arfvedsonite et aenigmatite de Madagascar,” Comptes Rendus, CXXX (1900), 1208; also ‘“‘Sur la’ province pétrographique de Nord-ouest de Madagascar,” ibid., CXXXII (r901); 439- 2 Frank D. Adams and Alfred E. Barlow, “Geology of the Haliburton and Ban- croft Areas, Province of Ontario,’ Mem. Geol. Surv. Canada, No. 6 (1910), p. 314. 3 Henry S. Washington, ‘‘The Roman Comagmatic Region,” Carnegie Publica- tion. No. 57 (Washington, 1906), p. 73. 4 Tbid., p. 136. * REVIE WSs The Environment of Vertebraie Life in the Late Paleozoic of North America; A Paleogeographic Study. By E.C. Case. Carnegie Institution: Washington Publication No. 283, 1919. Pp. 273, figs. 8 and two correlation tables. This publication will be welcomed by all geologists as a signal con- tribution to the interpretation of conditions of the late Paleozoic. Dr. Case is well qualified to summarize the conditions surrounding the verte- brates, and to draw conclusions as to the influence of the environment on their development and distribution. In the first chapter the various methods of attack and the complexity of the factors in any paleogeographic problem are briefly presented. There follows a summary of late Paleozoic rocks in the several prov- inces of North America. For the most part the descriptions are quota- tions from other writers and are not intended to present new material. Quite apart from the obvious purpose, this summary will be of great use to:the student, gathered as it is from a voluminous literature on the subject. The selections are well made; they are representative of current opinion, to the point, and show no evidence of an attempt on the part of the compiler to prove a point. The full usefulness is slightly impaired, perhaps, by the absence of an index. The accompanying correlation tables differ little from generally accepted views. Contrary to what might be expected, the author makes no attempt to fix the lower boundary of the Permian. Throughout the work he refers repeatedly to Permo-Carboniferous times in the sense of a transi- tion period between the Pennsylvanian and the closing events of the Paleozoic. He makes a sharp distinction, however, between Permo- Carboniferous times and Permo-Carboniferous conditions, and em- phatically states that these two things are not necessarily coincident. Perhaps the most important deduction reached, together with the dependent conclusions, is that Permo-Carboniferous conditions pre- vailed in the east, starting with Mid-Conemaugh time, and reached the southwest considerably later. The deposition of red sediments is taken to mark the introduction of important new climatic and physiographic conditions and accompanying changes—Permo-Carboniferous condi- 178 REVIEWS 179 tions. These conditions spread very slowly from east to west and, therefore, have left a record oblique to stratigraphic lines (see Fig. r). The environment of an organism, ‘‘the sum total of all its contacts with the external world,” determines to a great extent its structural changes and the distribution of its kind. The equable, humid climate and topographic uniformity of the typical (lower) Pennsylvanian pro- duced an abundant though fixed food supply—vegetation. Since new forms arise only through isolation (not necessarily geographic isolation) the monotony of this environment acted as a repressive force, checking the expansion of the amphibians and reptiles into new forms. ‘ En- vironmental monotony would result in the persistence of older and simpler types, because the variants, possibly being constantly produced, would not have a chance to develop.” Oktahonan (UII mn Sn Mi is NUT Hit onongahela SSourian H conemaugh Fic. 1.—‘‘ Diagram illustrating in a schematic way the relative position of the sediments formed under Permo-Carboniferous conditions. The land was rising from east to west, but there was continuous sedimentation in the eastern region at the western edge of the rising land of Appalachia. As the land rose slowly the red beds spread toward the west, occupying relatively higher positions in the stratigraphic column. It is difficult to illustrate the actual conditions in the diagram, because the ‘red beds conditions’ were advancing, but the wavy lines indicate the surface of the ground relative to these conditions. In Pennsylvania and West Virginia deposition was continuous during the conditions. In Illinois and Indiana deposition had ceased by the time the conditions reached that far west; in Kansas, Oklahoma, and Texas ‘red-bed conditions’ reached the region in time to affect only the uppermost Paleozoic deposits. The upper limit of the red-bed conditions is not known, and so the upper limit of the wedge is indicated by a dotted line” (from Case, p. 192). Permo-Carboniferous conditions included a cool to cold, arid or semiarid climate, resulting from deformation throughout various parts of the world, the presence of volcanic dust in the air, and a diminution of the carbon-dioxide content of the atmosphere. This made for a great variety in environmental conditions and destroyed the repressive bounds to vertebrate expansion. ‘‘The fauna, long restrained from any expression of evolutionary tendencies, full fed, and in the vigor of its youth, responded at once to the change, and new forms appeared so suddenly as to be unheralded in the preserved remains.”’ 180 REVIEWS These new forms appeared at higher and higher horizons as the Permo- Carboniferous conditions spread slowly westward and ‘“‘to correlate widely separated groups of beds as synchronous in deposition because of a similarity, even approaching identity, in the fauna or flora would be a serious error.” M. G. M. Upper Cretaceous Floras of the Eastern Gulf Region in Tennessee. Mississippi, Alabama, and Georgia. By E. W. BERRY. U.S. Geological Survey, Professional Paper 112, 1919. Pp. 177, pls. 33, figs. 12. Another publication is added to the already considerable list which is making fossil plants such an important part of our geological knowl- edge of the southeastern United States. The Upper Cretaceous of the eastern Gulf region extends in a lunate outcrop around the southern end of the Appalachians. It is subdivided into the Tuscaloosa formation, the Eutaw formation, the Selma chalk, and the Ripley formation. These formations, with the exception of the Selma, are made up largely of cross-bedded sands, with associated clays. The most extensive flora is that of the basal Tuscaloosa formation, _ comprising 151 species of which the majority are dicotyledonous angio- sperms. The place of origin of this dominant element is left unsettled, but the idea of their dispersal from an Arctic area is consistent with the evidence offered by this and other Cretaceous floras. This flora is made up largely of lowland coastal types, and its ecological character is in accord with other evidence of the delta origin of the formation. The plants make up an assemblage which most nearly resembles the modern warm-temperate rain forest. In view of their northward range into Greenland, they may be said to indicate a climate mild over wide areas. The Eutaw flora comprises 43 species, most of which come from the basal portion of the formation and closely resemble those from the Tuscaloosa formation. The physical conditions suggested by this flora are similar to those for the Tuscaloosa. The Selma chalk, which is described as a lithologic rather than a chronologic unit, is entirely marine and contains no plant remains. The Ripley formation contains a few poorly preserved plant fossils. The Tuscaloosa formation may be correlated, on the basis of its contained flora, with the upper part of the Raritan and with the Magothy es oe REVIEWS I8t formations to the north, with the Woodbine sand of southern Texas, and with the Dakota sandstone of the western interior. The Eutaw flora closely resembles the floras of the Black Creek and Magothy formations of the Atlantic Coastal Plain. It cannot be closely related to any of the western floras, but since it is decidedly older than the Montana flora the Eutaw formation may be considered to be syn- chronous with part of the Dakota and with the Colorado series. The . flora of the lower part of the Ripley is related closely to those of the Black Creek, Magothy, Tuscaloosa, and Raritan formations,while that of the upper part shows little relation to any of the earlier Cretaceous floras. RoW. C- RECENT PUBLICATIONS —C ark, F. R. The Farnham Anticline, Carbon County, Utah. [U. S. Geological Survey, Bulletin 711-A. Washington, 1o109.| Geology of the Lost Creek Coal Field, Morgan County, Utah. [U.S. Geological Survey, Bulletin 691-L. Washington, 1918.] —CotLuteR, A. J. Coal South of Mancos, Montezuma County, Colorado. [U.S. Geological Survey, Bulletin 691-K. Washington, roro.| —Compte Rendu Sommaire des Séances de la Société Géologique de France. Année 1916. [Paris: Société Géologique de France, 28, Rue Serpente, vi. 1916.] —Conpit, D. D. Oil Shale in Western Montana, Southeastern Idaho, and Adjacent Parts of Wyoming and Utah. (Contributions to Economic Geology, 1919, Part II, pp. 15-40.) [U.S. Geological Survey, Bulletin. VII-B. Washington, roro.| —Cuerpo de Ingenieros de Minas del Peru. Yacimientos Carboniferos del Distrito de Llapo. Cuenca Carbonifera de Ancos. Boletin del Cuerpo de Ingenieros de Minas del Peru, No. 89. [Lima: Imp. Americana, Santo Toribio, 230-34. 1918.] —DakeE, C. L. The Sand and Gravel Resources of Missouri. [Missouri Bureau of Geology and Mines. Vol. II, Second Series. Rolla, 1918.] The Sand and Gravel Resources of Missouri. [Missouri Bureau of Geology and Mines. Vol. XV, Second Series. Rolla, 1918.] —Denis, T. C. Report on Mining Operations in the Province of Quebec, during the Year 1917. [Providence of Quebec, Canada, Department of Colonization, Mines, and Fisheries, Mines Branch. Quebec, 1918.] Report on Mining Operations in the Province of Quebec, during the Year 1918. [Province of Quebec, Canada, Department of Coloniza- tion, Mines, and Fisheries, Bureau of Mines. Quebec, 1o919.| —DuNELEY, W. A., AND ODELL, W.W. Water-Gas Operating Methods with Central District Bituminous Coals as Generator Fuel. [Illinois Geologi- cal Survey, Bulletin 24, Co-operative Mining Series. Urbana, 1o19.| —Duvntiop, J. P. Secondary Metals in 1017. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part 1:15. Washington, 1919.| —Dykema, W. P. Recent Developments in the Absorption Process for Recovering Gasoline from Natural Gas. [U.S. Bureau of Mines, Bulletin 176. Petroleum Technology 50. Washington, 1919.] —Eastman, E. D., Aanp DuscHax, L. R. The Vapor Pressure of Lead Chloride. [U.S. Bureau of Mines, Technical Paper 225. Washington, T9I9.] 182 RECENT PUBLICATIONS 183 —Europe, the Geography of. Prepared and issued under the auspices of the Division of Geology and Geography of the National Research Council. [New Haven: Yale University Press, 1918.] —Fay, A. H. (Compiler). Monthly Statement of Coal-Mine Fatalities in the United States, June, 1919. [U.S. Bureau of Mines. Washington, ror9.] Monthly Statement of Coal-Mine Fatalities in the United States, July, toro. [U.S. Bureau of Mines. Washington, 1919.] —FIELDNER, A. C., SELvic, W. A., AND TAyLor, G. B. The Determination of Combustible Matter in Silicate and Carbonate Rocks. [U.S. Bureau of Mines, Technical Paper 212. Washington, r1o19.| —Fintay, J. R. Method of Administering Leases of Iron Ore Deposits Belonging to the State of Minnesota. [U.S. Bureau of Mines, Technical Paper 222. Washington, ro19.| —Fucus, F. G. Yacimiento mineral del Cerro de Pasco. [Boletin de la Sociedad Geografica de Lima. Tomo XXXIV. Trim. II. Lima, Junio 30 de 1018.] - —GAUTHIER,H. Road Material Surveys in the City and District of Montreal, Quebec. [Canada Department of Mines, Geological Survey, Memoir 114; No. 95, Geological Series; No. 1755. Ottawa, 1919.| —GiLEs, A.W. The Country about Camp Lee, Virginia. [Virginia Geologi- cal Survey, Bulletin No. 16. Charlottesville, 1918.] —Grecory, H. E. (Superintendent). Eighth Biennial Report of the Com- missioners of the State Geological and Natural History Survey, 1917-1918. [Bulletin No. 28. State of Connecticut Public Document No. 47. Hart- ford, r919.] —Grecory, H. E. (Editor). Military Geology and Topography. Prepared and issued under the Auspices of the Division of Geology and Geography, National Research Council. [New Haven: Yale University Press, 1918.] —Grout, F. F. Clays and Shales of Minnesota. With Contributions by E. K. Soper. (Work done in co-operation with the Minnesota Geological Survey.) [U.S. Geological Survey, Bulletin 678. Washington, r910.] —GRroveR, N. C., et al. Surface Water Supply of the United States, 1915. Part XI. Pacific Slope Basins in California. [U.S. Geological Survey, Water-Supply Paper 411. Washington, 1918.] —Grover, N. C., anp Battey, C. T. Surface Water Supply of Hawaii, July 1, 1917, to June 30, 1918. (Prepared in co-operation with the Ter- ritory of Hawaii.) .[U.S. Geological Survey, Water-Supply Paper 485. Washington, ro19.| : —Harner, E. C., AND Jonnston, A. W. Preliminary Report on the Geology of East Central Minnesota Including the Cuyuna Iron Ore District. [Minnesota Geological Survey, Bulletin No. 15. Minneapolis, 1918.] —Harrincton, G. L. The Anvik-Andreafski Region, Alaska. (Including the Marshall District.) [U.S. Geological Survey, Bulletin 683. Wash- ington, 1918.] 184 RECENT PUBLICATIONS —HavseEn, J. Contribucién al Estudio de la Petrografia del Territorio Nacional de Misiones. Ministerio de Agricultura de la Nacién. Direc- cién General de Minas, Geologia e Hydrologia. Boletin No. 21, Serie B. (Geologia). [Buenos Aires, 1919.] —HEIKEs, V. C. Gold, Silver, Copper, Lead, and Zinc in Montana in 1917. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part 1:16. Washington, ror19.] Gold, Silver, Copper, Lead, and Zinc in Nevada in 1917. Mines Report. [U.S. Geological Survey. Mineral Resources of the United States, 1917. Part I:14. Washington, 1o19.] —HENDERSON, J. The Geology of the Te Kuiti District, with Special Refer- ence to Coal Prospects. The New Zealand Journal of Science and Technology, March, 1918. [Wellington, 1918.] ——. Notes on the Geology of the Cheviot District. Marble in Riwaka- Takaka District. The New Zealand Journal of Science and Technology, May, 1918. [Wellington, 1918.] Notes on the Geology and Mineral Occurrences of the Wakamarina Valley. The New Zealand Journal of Science and Technology, January, 1918. [Wellington, 1918.] ~ Notes on the Geology of the Murchison District. The New Zealand Journal of Science and Technology, March, 1918. [Wellington, 1918.] —Hermitte, E. M. Memoria de la Direccién General de Minas, Geologia — e Hidrologia, Correspondiente al Afio 1916. Republica Argentina. Anales del Ministerio de Agricultura de la Nacién. Seccién Geologia, Mineralogia y Mineria. Tomo XIII, Naim. 5. [Buenos Aires, ror9.] —Honman, C. S. The Geology of the North Coolgardie Goldfield. Part I. The Yerilla District. With Appendixes by R. A. FarquHaRson and J. T. Jutson. [Western Australia Geological Survey, Bulletin No. 73. Perth, 1917.] —Hussakor, L., anp Bryant, W. L. Catalog of the Fossil Fishes in the Museum of the Buffalo Society of Natural Sciences. Bulletin of the Buffalo Society of Natural Sciences, Vol. XII. [Buffalo, N.Y., 1918.] —Institute for Government Research. The United States Geological Survey. Its History, Activities and Organization. [Service Monographs of the United States Government, No.1. New York: D. Appleton & Co., 1918. —TInter-America. Vol. III, No. 1, October, 1919. [New York: Doubleday, Page & Co.] —Jaccar, T. A., AND RomBEerc, ARNOLD. An Experiment in Teleseismic Registration. Bulletin Seismological Society of America, Vol. VIII, Nos. 2-3, June-September, 1918. [Hawaiian Volcano Observatory, 1918.] —Jounson, D. W. Shore Processes and Shoreline Development. [New York: John Wiley & Sons, 1919.] ‘7 hort History of Belgium. By Léon Van per Essen, Professor of History in. the University of Louvain. $1.50, postpaid $1.65. In this new and enlarged edition the author tells in an interesting way of the formation, struggles, and triumphs _ of the little kingdom; presents additional material on the reign of Leopold II; and describes the heroic nation’s part in the Great War. . A Field aid Laboratory Guide in iyscal Nature-Study. By Exzor aN R. Downtnc, Associate Professor of Natural Science in the School of Education, the University of Chicago. Loose-leaf, 110 pages of text, interleaved, paper covers; $1.00, postpaid $1.10. Tells how to make toys that work by air, heat, electricity, light, and other physical agencies. The child learns the principles without effort while he is absorbed in making the toys. Also tells about the common minerals and rocks and the stats that are the wonder of every child. ‘The Geography of the Ozark Highland of Missouri. By Cart Ortwm _.____ Saver, the University of Michigan. $3.00, postpaid $3.20. The purpose of such a ; ; study i is twofold: to furnish an adequate explanation of the conditions of life in a given area, and to contribute proved statements which will aid in working out funda- ‘mental principles. A valuable feature of the volume is the forty-four figures in the text and twenty-six plates. A pawl of Style (Revised Edition). By the STAFF OF THE UNIVERSITY OF CHI- ~ CAGO Press. $1.50, postpaid $1.65. An invaluable handbook for writers, editors, teachers, librarians, ministers, lawyers, printers, oe proofreaders—for all who need guidance on matters of propriety in style. Current Economic Problems (Revised Edition). Edited by Watton H. eS Hamitton, Professor of Economic Institutions, Amherst College. $3.50, postpaid $3.75. A selection of readings from the widest range of sources. The revision pre- sents new materia] made available as a result of the war, and the new point of view - brought about by the war in many lines of thinking on economic subjects. *+oblems of Fertilization. By Frank R. Litxiz, University of Chicago. $1.75, ; postpaid $1.90. As the problem of fertilization deals with the union of male and - female sex elements, which forms the beginning of each individual’s life-history, it has ‘important bearings upon the problems of heredity. It also has close relations to Sass other fundamental biological problems. e RecélaGon of John. Is the Book of Revelation a Mystery to You? By - Surrtery J. CAsE, Professor of Early Church History and New Testament, the Uni- versity of Chicago. $2.00, postpaid $2.15. This is a popular presentation of the subject and not a technical. commentary. The author tells why and when the book was written. a Sih se e ee of Chisstianity i in the Modern World. By Epwarp Carpwet _. Moore, Harvard University. $2.00, postpaid $2.15. The book gives a survey _ of the expansion of Christianity since the beginning of the modern era, about 1757, and depicts the movement against the background of general history. THE =o eats fodapulemaen ss OF CHICAGO PRESS A hi 4 0G Cie Z “ ie ‘ILLINOIS eo PSF GSH SFOS (62) HAUNa}6a)RaNg(03/0a nA NSO AMAT SAIB A ESLESIBAAIE USAR Te FINE INKS 4"> ADHESIVES || ‘For those who KNOW Made; in our own he ory b we are oe to : Drawinginks | Eternal Writing Ink Engrossing Ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Past Office Paste’ letters. ' Vegetable Glue, I Etc. : on i cane the Finest and Best Inks and Adhesives | arms “ oe folder, Prk a | Rule y 18 Glas: Emancipate yourself from the use of corrosive and ill-smelling inks and adhesives and adopt the Higgins Inks and Adhesives. They willbea revelation to you, they are so aint clean, weil Al ? and. give. sharp, ; ‘put up, and withal so efficient. At Dealers Generally f , pressions. ' 4: CHAS. M. HIGGINS & CO., Mfrs.| | | REMINGTON TYPEW! Branches: Chicago, London | 5 374 Broadway ew 271 Ninth Street — ts Brooklyn, N.Y. | ne Branches Everywhere By ARTHUR PEARSON SCOTT Assistant Professor of Gee ae ees of Chicago and the framing ‘of the Tee” i Peace, prehensive explanation oi e ague of N: ations: location of new national boundaries. The author available information of “ por aa to evel citizen. aes $2, 00, pospai $2. 15 chy Mlinois es 2 Re Nady ho + oo on With the Active Collaboration of ROLLIN T. CHAMBERLIN, Dynamic Geology i ASSOCIATE EDITORS D GRIKIE, Great Britain : ; JOHN C. BRANNER, Leland Stanford Junior University RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. : HT PENCK, Germany Doty WILLIAM H. HOBBS, University of Michigan — = oe oe _ | FRANK D. ADAMS, McGill University -D AVID, Auigteatier CHARLES K. LEITH, University of pier amet eland Stanford Junior University WALLACE W. ATWOOD, Harvard University ee Smithsonian Institution WILLIAM H. EMMONS, University of Minnesot ARTHUR L. DAY, Carnegie Institution » APRIL-MAY 1920.04 5 p O AND, COMPOSITION OF BITUMINOUS COALS - REINHARDT THIESSEN | eee as QU: NTI ATIVE MINERALOGICAL CLASSIFICATION OF IGNEOUS CKS—REVISED. PART TI - SiR er sabi tahoe nee ALBERT JOHANNSEN . ‘SETTING OF NEW MEXICO- - - - - +. CHARLES KEves - Frank F. Grout i wm onl 4 iP) 2) Ki z ee t= Leet N — Z @ 2 es Q 3 eo s) Z e) I Q v2) 2 oc E N lo | Z @ = eo i) = > ONT BowEN CHICAGO, ALLINOIS, LOers c743 | ‘THE ‘CAMBRIDGE UNIVERSITY PRESS, Lonpon THE ‘MISSION BOOK COMPANY, SHANGHAI Io. Fic. 44.—Part of a horizontal cleavage plane of coal showing a Medullosa type of woody structure, in which “needles” or “rodlets” form part of the fissue. X<.3. A QUANTITATIVE MINERALOGICAL CLASSIFICATION OF IGNEOUS ROCKS—REVISED ALBERT JOHANNSEN University of Chicago PART III CLASS 2, ORDER 3 (237) Calcigranite. No plutonic rock falling near the center point of this family has yet been located. Quartz-ciminite. Among the extrusives the only rock . found in this family is a quartz-bearing ciminite described by Washington. Ciminite, named from its occurrence on Monti Cimini, Italy, was defined’ as consisting of alkali feldspar, basic plagioclase, augite, and olivine, with accessory magnetite and apatite. From two modal analyses given, it appears that there are quartz-bearing and quartz-free ciminites; consequently the two divisions, quartz-ciminite and ciminite, are here made. A modal analysis of one rock, here called quartz-ciminite, gives orthoclase (OrsAb,) 43.6 per cent, labradorite (Ab,;An,) 16.1 per cent, quartz 4.6 per cent, apatite o.7 per cent, augite 22.4 per cent, olivine 11.7 per cent, and magnetite o.9 per cent. Since this rock contains olivine, it is not representative of the normal extru- sives of the family. (238) Calciadamellite. This is a quartz-monzonite whose plagioclase is labradorite. Here fall four specimens from the Elkhorn district, Montana, on the border of the Butte batholith, described by Barrell The rock, however, is near the border line between Orders 2 and 3. The plagioclase is described as ean * Henry S. Washington, “Italian Petrological Sketches, IT: The Viterbo Region,” Jour. Geol., TV (1896), 838. 2 Tbid., V (1897), 351; ‘‘The Roman Comagmatic Region,” Carnegie Publication No. 57 (Washington, 1906), p. 65. 3 Joseph Barrell, ‘‘Microscopical Petrography of the Elkhorn Mining District, Jefferson County, Montana,” U.S. Geol. Surv., Ann. Rept., XXII, Part II (x90), p. 538. 210 CLASSIFICATION OF IGNEOUS ROCKS 211 ‘‘Ab,An, or more basic’? by Barrell. On the other hand, Cross, Iddings, Pirsson, and Washington’ assume, by calculation from the chemical analysis of a specimen from the Butte region, that the plagioclase has Ab;An, centers but more acid borders. The orthoclase in their analysis, however, was calculated as pure potash feldspar, while as a matter of fact it contains considerable soda. Consequently the plagioclase may be, as Barrell says, more basic, and the C.I.P.W. rock would fall in the calciadamellite family with Barrell’s rock., (239) Granogabbro JOHANNSEN.? See note under grano- diorite (229) and leuco-granogabbro (139). Rhyobasalt. To have a term analogous to granogabbro, the term rhyobasalt is here used. See note under grano- diorite (229). (2310) Quartz-gabbro. Quartz-olivine-gabbro. Quartz-basalt. Quartz-olivine-basalt. (2312) Calcisyenite. No modal analysis of a plutonic rock belonging here has yet been located. ; Vulsinite WASHINGTON. Vulsinites are defined by Washington: as “‘effusive rocks occupying an intermediate position between the trachytes and the andesites. They are characterized mineralogically by the presence of alkali feldspar with a large amount of basic plagioclase (labradorite to anorthite) together with augite and diopside. Hornblende and biotite are not abund- ant in the type specimens, though they may be present in large amounts in other varieties. ... . Olivine is wanting, or if present is so in only accessory amounts.” While the definition would suggest a rock of the latite series, the modal analysis of the Bolsena type’ shows it to belong to Family 12. The mineral percentages 1 Cross, Iddings, Pirsson, and Washington, Quantitative Classification of Igneous Rocks (Chicago, 1903), p. 227. 2 Albert Johannsen, ‘‘Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,” Jour. Geol., XXV (1917), 80. 3 Henry S. Washington, “Italian Petrological Sketches, I: The Bolsena Region,” Jour. Geol., IV (1896), 553. 4 Ibid., ‘The Roman Comagmatic Region,” Carnegie Publication No. 57 (Wash- ington, 1906), p. 65. 212 ALBERT JOHANNSEN are: soda-orthoclase (Or,Ab,) 69.5, basic plagioclase (anorthite phenocrysts 6.1, labradorite groundmass 11.9, average Ab,An,) 18.0, augite 7.4, biotite 1.0, ores 3.6, apatite o.4, titanite o.1. Ciminite WASHINGTON. The quartz-free ciminite has a mode! of soda-orthoclase (Or,,Ab;) 50.7 per cent, labradorite (Ab,An,) 13.1 per cent, augite 23.2 per cent, olivine 11.2 per cent, magnetite o.9 per cent, apatite o.g per cent. See note under quartz-ciminite (237). . (2313) Calcimonzonite. The original Monzoni monzonite, according to Broégger, usually contains a basic plagioclase (see note under 2213). In the present classification normal orthoclase- acid-plagioclase rocks are included under the term monzonite and orthoclase-basic-plagioclase (excluding anorthite) rocks under the term calcimonzonite. ‘ Calcilatite. The extrusive equivalent of the preceding. (2314) Monzogabbro (monzonorite). The rocks which fall in this family are described in the literature as gabbros, norites, or monzonites. The term syenogabbro, originally proposed’ for this family, is here withdrawn, and the term monzogabbro sub- stituted for reasons stated under granodiorite (229). Another reason why syenogabbro should not be used for the plutonic rock of this family is that, by analogy, the extrusive should then be called trachy-basalt. But Boricky? used the term trachy-basalt for rocks which are now called monchiquites. Basalatite. Basalatite, as intermediate between basalt and latite, is here suggested. It corresponds in form with its deep-seated equivalent, monzogabbro. (2315) Gabbro von Bucu. The term gabbro (granito di gabbro) was used by Targioni Tozzetti‘ and other writers for diallage- serpentine and related rocks from Tuscany. Von Buch’ applied «Henry S. Washington, ‘““The Roman Comagmatic Region,” Carnegie Publication No. 57 (Washington, 1906), p. 32. 2 Albert Johannsen, op. cit., p. 89. 3 Emanuel Boticky, “‘Petrographische Studien an den Basaltgesteinen Bohmens,” Arch. f. d. naturw. Landesdurchf. v. Béhmen, II (1874), Abt. ii, Th. ii, p. 44. 4Targioni Tozzetti, Relazioni d’alcuni viaggi fatti in diverse parti della Toscana (Firenze, 1768), IIL, 432. 5 Leopold von Buch, ‘‘Ueber den Gabbro” (read in Akad. d. Wissens., Berlin, October 12, 1809), Magazin d. Gesell. naturf. Freunde z. Berlin, IV (1810), 128-49. CLASSIFICATION OF IGNEOUS ROCKS 213 it to rocks consisting of ‘‘Saussurit oder Jade und Smaragdit, oder haufiger aus Feldspath und Smaragdit . . . . oder auch seltener aus allen diesen Substanzen vereinigt.’”’ His saussurite is altered feldspar, the jade amphibole, and the smaragdite probably green diallage. For a long time the name continued to be applied to plagioclase-diallage rocks, but in recent years the kind of pyroxene has been disregarded, and only the basic character of the plagioclase has been considered essential to the definition. In the present classification, therefore, gabbro is simply a rock consisting essen- tially of labradorite or bytownite and a biopyribole. Olivine may be an accessory, and magnetite is usually present in variable amounts. The orthoclase-gabbro of Streng’ and Irving? is really diallage-monzonite or monzodiorite. Hornblende-gabbro. Olivine-gabbro. Quartz-gabbro. Uralite-gabbro, etc. Norite Esmark. Esmark’ applied the name norite to certain Norwegian rocks which belong in part to the rocks now called norites, but in part to the diorites. Scheerer‘ applied it to rocks related to the gabbros and hyperites. Rosenbusch*® limited it to rocks containing essential hypersthene and plagioclase, and said that unless so limited the term would be meaningless, and later® gave the constituents as basic plagioclase and an orthorhombic pyroxene. In this sense it is now generally used. Hyperite TOrNEBoHM. These rocks, intermediate between gabbros and norites, and containing both orthorhombic and monoclinic pyroxenes, were originally called hypersthenites «Aug. Streng, “Uber die kristallinischen Gesteine von Minnesota in Nord- amerika,’ Neues Jahrb. (1877), pp. 113-38. 2 Roland Duer Irving, “The Copper-bearing Rocks of Lake Superior,’ U.S. Geol. Surv., Mono. 5 (1883), pp. 50-52. 3 Esmark, Magazin fiir Naturvidenskaberne, I, 207. 4Th. Scheerer, Gaea Norvegica, Heft ii, 313; also “Geognostisch-mineralogische Skizzen gesammelt auf einer Reise an der Siid-Kiiste Norwegens,” Mewes Jahrb. (1843), p- 668. 5H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (1st ed.; Stuttgart, 1877), p. 477. 6 Tbid. (4th ed., 1907), p. 348. 214 ALBERT JOHANNSEN by Rose.* The name was too suggestive of a rock entirely com- posed of hypersthene; therefore Térnebohm? gave to them the name hyperite, although this term had previously been used by Senfts for his group composed of eclogites, gabbros, and hypersthenites. Syn.: Hypersthene-syenit, Hypersthenite Rose, Hypersthene-gabbro, Augite-norite, etc. Basalt. The derivation of the term is unknown. It may be derived from the Ethiopian bsalt, “cooked,” suggesting a ~ baked rock, or from barzalien, barzel, the Hebrew for “‘iron.” Pliny’ speaks of “‘basalten” found in Ethiopia, a rock which “ color and hardness resembles iron, and is used in making statuary.” Agricola’ thought that certain rocks in Saxony were identical with the basalt of the older writers, and applied this term to them, and Werner® used it for the same rocks, which he considered sedi- mentary. Basalt is the extrusive equivalent of gabbro, and in general the term is applied to rocks with basic plagioclase and _augite, with or without olivine. Some writers apply the term to plagioclase-augite rocks with olivine, irrespective of the kind of plagioclase, though in general, at the present time, this is not the basis of separation from the andesites. On the basis of the feldspar, therefore, there would be | Hornblende-andesite With acid plagioclase 4 Augite-andesite Olivine-augite-andesite, etc. : Hornblende-basalt Basalt (with pyroxene) = Auganite WINCHELL Olivine-basalt, etc. With basic plagioclase ™G. Rose, ‘‘Uber Hypersthenit,” Pogg. Ann., XXXIV (1835), Io. 2A, E. Tornebohm, “Uber die ee Diabas- u. Gabbro-Gesteine Schwedens,” Neues Jahrb. (1877), p. 379- 3 Ferdinand Senft, Classification und Beschreibung der Felsarten (Becca 1857), Pp. 59- 4C. Plinii Secundi Naturalis Historiae xxxvi. cap. vil. Ed. Lugd. Batay. Rotero- dami, Ao. 1668, p. 645. 5 Georg Agricola (Georg Bauer), De re metallica, 1556. 6 A. G. Werner, ‘‘Bekanntmachung einer . . . . iiber die Entstehung des Basaltes gemachten Entdeckung .... ,”’ Bergménn. Journ., Pt. IL (1788), p. 845. CLASSIFICATION OF IGNEOUS ROCKS ams Winchell’ suggests the word auganite for olivine-free basalt, and uses basalt for the olivine-bearing variety. The writer prefers basalt and olivine-basalt for these rocks. According to their textures, basalts have been divided into three groups: basalts proper, anamesite, and dolerite. Basalt proper. Compact, dense, aphanitic. Anamesite von LEONHARD.? Megascopically crystal- line, but fine-grained. Name derived from dvdpeoos, “in the middle.”’ Dolerite Hatiy,3 from 6dodepos, ‘‘deceptive.’’ Coarse- _ grained. Alboranite BECKE. Becke‘* proposed the term albora- nite for certain extrusive rocks from the island of Alboran. They consist of hypersthene and basic plagioclase. The phenocrysts in the rocks described by him are anorthite and the groundmass microlites labradorite. The rocks thus are olivine-free hypersthene- basalts, and, as extrusive representatives of norite, deserve the new name. See note under santorinite (2215). (2323) Kulaite WASHINGTON. This term was originally applied by Washington’ to certain extrusive rocks from the volcanoes oi . Kula, in Lydia, Asia Minor, under the impression that they were hornblende-basalts. Later they were shown by him® to be nephelite-bearing and to contain approximately equal amounts oi orthoclase and basic plagioclase. No plutonite of this composition has yet been located among modes given in the literature; therefore kulaite is temporarily used as the family name. The term should not be confused with kullait HENNIG. t Alexander N. Winchell, “‘Rock Classification on Three Co-ordinates,” Jour. Geol., XXI (1913), 215; also “‘Geology of the National Mining District, Nevada,” Mining and Scientific Press, CV (1912), 657. 2 Karl Casar von Leonhard, Die Basalt-Gebilde in ihren Beziehungen zu normalen und abnormen Felsmassen (Stuttgart, 1832), I, 151. 3 Ascribed to Haiiy by Alexandre Brongniart, Classification et caractéres minéralo- giques des roches (Paris, 1827), p. 101. 4F. Becke, ‘‘Der Hypersthen-Andesit der Insel Alboran,” Tscherm. Min Petr. Mitth., XVIII (1899), 553. ; 5 Henry S. Washington, ‘‘On the Basalts of Kula,” Amer. Jour. Sci., XLVII (1894), 115. 6 [bid., ““The Composition of Kulaite,’’ Jour. Geol., VIII (1900), 618. 216 ALBERT JOHANNSEN Loewinson-Lessing’ expressed the opinion that kulaite is the extrusive equivalent of heumite, with which the present writer does not agree. Heumite is described by Brégger? as a rock essen- tially of soda-orthoclase or soda-microcline with other feldspars, very small amounts of nephelite and sodalite, and considerable barkevikite and biotite. Furthermore, the leucocratic minerals . in heumite form only 53 per cent of the rock, and the subordinate plagioclase is oligoclase-albite, and not basic plagioclase. (2324) Nephelite-(leucite-)monzogabbro. A term suggested here for the nephelite-(leucite-)bearing rocks of the basic plagioclase series, and comparable to the granogabbros among those bearing quartz. Nephelite-syenogabbro, suggested previously,’ is with- drawn. See notes under (229) and (2314). Essexite SEARS is quite variable, but probably belongs here or to (2325), or to both, although the original description by Sears does not mention the presence of orthoclase. He says it con- tains augite, hornblende, biotite, plagioclase, and nephelite ~ the usual accessories. Washington, describing the same rc says that it is essentially a basic monzonitic rock in which fek - spathoids and both lime-soda and alkali feldspars are present. The feldspar ranges from Ab,An, to Ab,An,, and “an alkali-feldspar is not uncommon... . often microperthitic.” Nephelite is fairly abundant. In another specimen the plagioclase was Ab,An, and only a few grains of alkali-feldspar were seen. Speaking of certain other rocks described as essexites, Washington® says that since they contain neither nephelite nor alkali-feldspar, they are not essexites. Many rocks, clearly not essexites, have been described 1 F, Loewinson-Lessing, ‘‘Kritische Beitrige zur Systematik der Eruptivgesteine, V,” Tscherm. Min. Petr. Mitth., XXI (1902), 322. 2W.C. Brogger, Die Eruptivgesteine des Kristianiagebietes. III: Das Ganggefolge des Laurdalits (Kristiania, 1898), pp. 98-113. 3 Albert Johannsen, ‘‘Suggestions for a Quantitative Mineralogical Classification of Igneous Rocks,” Jour. Geol., XXV (1917), 89. 4 John H. Sears, “‘Elaeolite-Zircon-Syenites and Associated Granitic Rocks in the Vicinity of Salem, Essex County, Massachusetts,” Bull. Essex Inst., XXIII (1801), 146. 5’ Henry S. Washington, “‘The Petrographical Province of Essex County, Massa- chusetts,” Jour. Geol., VII (1899), 53-56. 6 Loc. cit. CLASSIFICATION OF IGNEOUS ROCKS 217 under this name. The original rock is apparently of Class 2 (though some recently described essexites are of Class 3), and - certainly of Order 3. (2325) | Nephelite-(leucite-)gabbro. Rouvillite O’NEILt belongs here. This is a rock from St. Hillaire, Quebec, described by O’Neill.t It may be called a nephelite-gabbro or light-colored theralite. Rosiwal measurements show the rock to consist of nephelite 29.35 per cent, plagioclase (Ab,,.Ango to Ab..Ango) 55.9 per cent, apatite 1.15 per cent, pyroxene 7.50 per cent, and horn- blende 3.64 per cent. (2327) Heronite CoLEMAN. While this is an analcite dike rock, and far from the center point of the family, it is the only rock so far located in this pigeonhole. The name was given by Coleman? to a rock from Heron Lake, north of Lake Superior, consisting of much analcite (53 per cent of the leucocratic minerals), orthoclase (28.24 per cent), labradorite (13 per cent), aegirite (4.04 per cent), ‘p<,lmonite and calcite. 30) Lugarite TYRRELL’ is a porphyritic rock, occurring in - ses and as a sill in the Lugar teschenite-picrite complex in the west of Scotland. It consists of analcite (with some nephelite) 50 per cent, labradorite 10 per cent, apatite 2 per cent, titanaugite 20 per cent, barkevikite 15 per cent, and ilmenite 3 per cent. Tyrrell considers the analcite “original, displacing nephelite.”’ CLASS 2, ORDER 4 (247) Anorthite-granite. (248) Anorthite-adamellite. (240) Anorthite-granogabbro. (2410) Quartz-anorthite-gabbro. (2412) Anorthite-syenite. A syenite whose small percentage of plagioclase is anorthite may be called anorthite-syenite. Here belongs a so-called shonkinite from Elkhorn, Montana, described tJ. J. O'Neill, “St. Hilaire (Beloeil) and Rougemont Mountains, Quebec,” Geol. Surv. Canada, Mem. 53 (Ottawa, 1914), p. 35. 2A, P. Coleman, “‘A New Analcite Rock from Lake Superior,” Jour. Geol., VII (1899), 435- 3G. W. Tyrrell, “The Late Palaeozoic Alkaline Igneous Rocks of the West of _ Scotland,” Geol. Mag., IX (1912), 77-78. 218 ALBERT JOHANNSEN by Barrell.t It is hardly typical of the family, however, since it is near the boundary of Class 3. See note under (2112). (2413) Anorthite-monzonite. See note under (2113). (2414) Anorthite-monzogabbro. See notes under (21714) and (2314). (2415) Anorthite-gabbro. Here belongs an anorthite-augite dike rock from the Carlingford district, Ireland, with the percent- ages 62 and 38 according to a calculated analysis by Roth.? The corresponding extrusive rock, with more than 50 per cent anorthite, 33 per cent augite, and 8 per cent magnetite, was described as anorthite-diabase by Tschermak.3 ; Another rock belonging to this family is kyschtymite Morozx- wicz,4 an anorthite-corundum rock, occurring as an intrusive in granite in Kyschtym in the Urals. Still another is allivalite HARKER,» occurring on Allival, a mountain on the Isle of Rum and consisting of anorthite and olivine.® Finally, there is rougemontite O’NEILL,’ containing anorthite 52.25 per cent, augite 32.51 per cent, olivine 8.35 per cent, hornblende 0.43 per cent, and iron ore 6.52 per cent. CLASS 3, ORDER I (316) Mela-orthogranite. In this family fall two rocks, Prowersose (of the quantitative system of C.I.P.W.), described by Cross,’ a dike rock from Two Buttes, Colorado, and a similar rock t Joseph Barrell, ‘Microscopical Petrography of the Elkhorn Mining District, Jefferson County, Montana,” U.S. Geol. Surv., Ann. Rept., XXII, Part II (1901), p- 510. 2J. Roth, Gesteinsanalysen, lvii. The rock was originally calculated by G. Haughton (Quart. Jour. Geol. Soc., XII [1856], 197) as anorthite 85.84 per cent and augite 14.16 per cent, which was shown to be wrong by Roth. 3 Gustav Tschermak, “Uber secundare Mineralbildungen in den Griinsteingebirge bei Neutitschein,”’ Sitzungsber. d. Wien Akad. d. Wiss., XL (1860), 127. bP) 4J. Morozewicz, ‘‘Kyschtymit—ein Korund-Anorthitgestein,” Tscherm. Min. Peir. Mith., XVIII (2808), 202. ~ 5 Alfred Harker, ‘‘Igneous Rocks from the Ultrabasic Group of the Isle of Rum. Summary of Progress,” Geol. Szirv. (1903), p. 56. 6 J. W. Judd, “On the Tertiary and Older Peridotites of Scotland,” Quart. Jour. Geol. Soc., XLI (1883), 380-90, 305. 7J. J. O'Neill, ‘‘St. Hilaire (Beloeil) and Rougemont Mountains, Quebec,” Geol. Surv. Canada, Mem. 43 (Ottawa, 1914), p. 77. 8 Whitman Cross, “‘Prowersose (Syenitic Lamprophyre) from Two Buttes, Colo- rado,” Jour. Geol., XIV (1906), 165. CLASSIFICATION OF IGNEOUS ROCKS 219 from Knox County, Maine, described by Bastin.‘ They are melanocratic orthogranites. (317) Mela-albite-granite. A poor name, which should be replaced owing to the former use of albite-granite for a different kind of rock. See note under (217). (318) Mela-albite-adamellite. See note under (218). (319) Mela-albite-granodiorite. See note under (219). (3110) Mela-albite-tonalite. See note under (2110). (3111) Mela-orthosyenite. See note under (2111). (3112) Mela-albite-syenite. See note under (2112). (3113) Mela-albite-monzonite. See note under (2113). (3114) Mela-albite-monzodiorite. See note under (2114). (3115) Mela-albite-diorite. See note under (2115). (3116) Orthoshonkinite. Weed and Pirsson? gave the name » shonkinite (from Shonkin, the Indian name for the Highwood Mountains, Montana) to a melanocratic “granular plutonic. rock ‘consisting of essential augite and orthoclase..... It may be with or without olivine, and accessory nepheline, sodalite, etcetra, may be present in small quantities.’”’ In another place* they state that ‘‘a triclinic striated feldspar is also present, but in no considerable amount. ... . It is albite.”” The amount of albite is not mentioned, but it is included in the estimated amount of alkali- feldspar. Based on the definition, however, no albite is necessary, and by analogy with other rocks (see under (111)) the term ortho- shonkinite may be applied to the rocks as defined with less than 5 per cent albite. Where the albite percentage is greater, the rock falls into Family 17 as albite-shonkinite or shonkinite simply. While the original rock contained traces of feldspathoids, by defini- tion none is necessary, and none is shown in the mode of this rock as given by Washington.’ It has, however, affinities with the nephelite rocks; consequently, though it may fall on the feldspar base line of the double triangle, it is to be classed with the « Edson S. Bastin, ‘‘Some Unusual Rocks from Maine,” Jour. Geol., XIV (1906), 173-80. _ 2 Walter H. Weed and Louis V. Pirsson, ‘‘Highwood Mountains of Montana,” Bull. Geol. Soc. Amer.; VI (1895), 415-10. 3 [bid., p. 412. 4 Henry S. Washington, ‘‘The Foyaite-Ijolite Series of Magnet Cove: A Chemical Study in Differentiation,’ Jour. Geol., [IX (1901), 613. 220 ALBERT JOHANNSEN feldspathoid rocks in Family 16. Furthermore, a modal analysis of another rock* shows alkali-feldspar 20 per cent, nephelite 5 per cent, sodalite r per cent, apatite 4 per cent, and mafites 70 per cent (10 of which is olivine). The latter rock, consequently, may be called a nephelite-shonkinite of Family 21. (3117) Shonkinite. This is albite-bearing shonkinite. See note under (3116). (3121) Nephelite-shonkinite. See note under (3116). (3124) Melalitchfieldite. See note under (2124). (3125) Melamariupolite. See note under (2125). (3131) Bekinkinite RosENBUSCH and Missourite WEED and PIRSSON. The melanocratic nephelite plutonic rock of this family is repre- sented by bekinkinite, named by Rosenbusch? from its occurrence on Mount Bekinkina on the peninsula Ambavatoby, as described by Lacroix. While the original rock is said to contain a small amount of anorthoelase, the definition of the type rock of Rosenbusch does not require it. He calls it a plutonic form of nephelite basalt, and says it is related to ijolite as missourite is to fergusite. Missourite was named by Weed and Pirsson‘ from its occurrence on the Missouri River. It is a melanocratic, feldspar-free leucite rock. It contains 16 per cent leucite, 8 per cent analcite and zeolites, 50 per cent augite, 6 per cent biotite, and 5 per cent iron ore. Farrisite BROGGER® is a melanocratic melilite rock of this family. Mela-nephelite-basalt is the extrusive equivalent of bekinite, mela-leucite-basalt of missourite, and mela-melilite- t Louis V. Pirsson, ‘‘Petrography and Geology of the Igneous Rocks of the High- wood Mountains, Montana,” U.S. Geol. Surv., Bull. 237 (1905), p- 104. 2H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (4th ed.; Stuttgart, 1907), p. 441. 3A. Lacroix, “‘Sur quelques roches ijolithiques du Kilima-Ndjaro,” Bull. Soc. Min. France, XXIX (1906), 90. 4 Walter H. Weed and Louis V. Pirsson, ‘‘Missourite, a New Leucite Rock from the Highwood Mountains of Montana,” Amer. Jour. Sci., II (4896), 323;. Louis V. Pirsson, “‘Petrography and Geology of the Igneous Rocks of the Highwood Mountains, Montana,” U.S. Geol. Surv., Bull. 237 (1905), p. 118. sW. C. Brogger, Die Eruptivgesteine des Kristianiagebietes. III: Das Gang- gefolgschaft des Laurdalits (Kristiania, 1898), p. 70. CLASSIFICATION OF IGNEOUS ROCKS 221 basalt or mela-melilitite of farrisite. There are also analcite- basalts, most of which belong in this family, though some fall in Class 2. CLASS 3, ORDER 2 (327) Melagranite. See note under (127). (328) Mela-adamellite. See notes under (127) and (228). (3209) Melagranodiorite. See notes under (127) and (2209). (3210) Melatonalite. See notes under (127) and (2210). (3212) Melasyenite. See notes under (127). Some years ago Weed and Pirsson’ described a rock from the Highwood Mountains as containing about equal amounts of light and dark constituents, and gave toit the name yogoite. ‘Their series of rocks as given was: All orthoclase, no augite = sanidinite. Orthoclase exceeds augite = augite-syenite. Orthoclase equals augite = yogoite. Augite exceeds orthoclase = shonkinite. All augite, no orthoclase = pyroxene and peridotite rocks of various types. Later? they withdrew the name yogoite since they found that it fell into Brégger’s monzonite group on account of the proportions of orthoclase and plagioclase. The original yogoite, computed in the present system, falls in (2212) and is a normal monzonite, but associated with it, on Yogo Peak, and called shonkinite? by Weed and Pirsson, are two other rocks which differ from normal shonkinites in being associated with quartz-bearing instead of with feldspathoid-bearing rocks and in containing andesine instead of albite. They also contain more soda-orthoclase than andesine. These ‘“‘shonkinites,” therefore, may well take upon themselves the discarded name yogoite, since they also occur on Yogo Peak, and fit into the foregoing scheme even better than the original yogoite. (3213) Melamonzonite. See note under (127) and (2213). Here belongs a basic contact monzonite of the Coryell batholith, ™W. H. Weed and L. V. Pirsson, ‘‘Igneous Rocks of Yogo Peak, Montana,” Amer. Jour. Sci., L (1895), 479. 2W. H. Weed and L. V. Pirsson, ‘‘The Bearpaw Mountains of Montana,” Amer. Jour. Sci., I (1896) 357-58. 3 L. V. Pirsson, “‘ Petrography of the Igneous Rocks of the Little Belt Mountains, Montana,” U.S. Geol. Surv., Ann. Rept., XX, Part III (1900), p. 487. 222 ALBERT JOHANNSEN described by Daly,7 as well as a contact rock against granite described by Miller.’ : (3214) Melamonzodiorite. Based on Daly’s’ average analysis of 161 basalts, as named by the original authors, Leith and Mead+ computed the mineral composition of the average ‘“‘basalt” and found that it contained oligoclase 35.4, orthoclase 10.75, augite” 36.90, olivine 7.58, magnetite 5.80, ilmenite 0.73, and titanite 2.84. This rock, according to the computed mode, therefore, is not of the gabbro family at all, but of the monzodiorite. Included in Daly’s average, of course, are all rocks named “‘basalts”’ by the original authors, consequently including many which at the present time would be called andesites. (3215) Meladiorite. Here belong, among dike rocks, many camptonites, kersantites, and spessartites, and some diabases and basalts, though most of the latter rocks belong to Class 2. Among deep-seated rocks there are a few meladiorites. (3217) Oligoclase-shonkinite. Andesine-shonkinite. It would be very desirable if there were terms to express the acid-plagioclases exclusive of albite (Ca Naf, of the present system) and the basic plagioclases exclusive of anorthite (NaCaf). Rosenbusch’ found the same difficulty when in his description of granite he said: “‘Oligoklas steht hier und im Folgenden fiir sauren Plagioklas.” A single term would thus cover the rocks of this family. See note under (3116). ; CLASS 3, ORDER 3 (337) Mela-calcigranite. This term is too awkward. The only modal analysis yet found in this family does not lie near the center point (339) Melagranogabbro. See notes under (127) and (239). ™ Reginald A. Daly, ‘“‘Geology of the North American Cordillera at the Forty- ninth Parallel,’ Geol. Surv. Canada, Mem. 38, Part I (1912), p. 361. 2 William J. Miller, “‘Geology of the North Creek Quadrangle, Warren County, New York,” N.Y. State Museum, Bull. 170 (1914), Pp. 37- 3 Reginald A. Daly, ‘‘ Average Chemical Compositions of Igneous Rock-Types,” Proc. Amer. Acad. Arts and Sci., XLV (1910), 224. 4C.K. Leith and W. J. Mead, Metamorphic Geology (New York, 1915), p. 74. 5 H. Rosenbusch, Elemente der Gesteinslehre (Stuttgart, 1898), p. 76. CLASSIFICATION OF IGNEOUS ROCKS 223 (3310) Mela-quartz-gabbro. Too awkward a term. There are three rocks in this family, but all of them are melanocratic hornblende-quartz-gabbros—an ‘“‘abnormal hornblende-gabbro”’ in the Moyie sill described by Daly,‘ a quartz-bearing gabbro from the Purcell sill, also described by Daly,? and a hornblende-gabbro from Sepainlampi, Hauksuo, Kisko, described by Eskola.s The type of the family should be chosen from an augite rock. (3314) Melamonzogabbro. See note under (2314). Here are included two essexites, a hornblende-gabbro, a hornblende-norite, and a so-called gabbro. (3315) Melagabbro. See note under (2315). Many gabbros fall here, also a few norites, and some diabases and basalts (and “dolerites’’). Melabasalt. The extrusive equivalent of the above. Arapahite WASHINGTON and Larsen? belongs here. (3317) Labradorite-(bytownite-)shonkinite. See note under (3217). (3324) Mela-nephelite-monzogabbro. See note under (2324). (3325) Theralite RoseNBuscH. Theralite, from the Greek Onpav (‘eagerly looked for’’), was applied by. Rosenbusch’ to plagioclase-nephelite rocks first thought to be represented by certain tephrites and basanites described by Wolff.° True theralites, however, were first described by Wolff? some years later. Besides the presence of plagioclase and nephelite, these rocks are character- ized by much predominating dark constituent; consequently they belong to Class 3. Among the lime-soda feldspars of the rocks Reginald A. Daly, “Geology of the North American Cordillera at the Forty- ninth Parallel,’ Geol. Surv. Canada, Mem. 38, Part I (1912), p. 234. 2 [bid., p. 224. 3 Pentti Eskola, ‘“‘On the Petrology of the Orijarvi Region in Southwestern Fin- land,” Bull. d. 1. com. géol. d. Finlande (Helsingsfors), 1914, p. 71. _4Henry S. Washington and E. S. Larsen, “‘Magnetite Basalt from North Park, Colorado,” Jour. Wash. Acad. Sci., III (1913), 452. 5H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (2d ed.; Stuttgart, 1887), p. 248. 6 J. E. Wolff, ‘Notes on the Petrography of the Crazy Mts., and Other Localities in Montana Territory,”’ Northern Transcontinental Survey (1885), pp. 8-13. 7J. E. Wolff, “On the Occurrence of Theralite in Costa Rica, Central America”, Amer. Jour. Sci., 1 (1896), 271. 224 ALBERT JOHANNSEN classed by Rosenbusch’ as theralites “‘ist die Labradoritmischung die herrschende, nach auszen hin aufsteigend bis zum Andesin, in den Kernen sinkend bis an die Grenze zum Bytownit.” The average feldspar, therefore, is basic plagioclase. Olivine may or may not be present. In the rock from Costa Rica described by Wolff and called the true theralite type by Rosenbusch, who named it, the plagioclase is labradorite. The rock contains, however, a little orthoclase, not necessary in the type. Kylite TYRRELL,’ with 31 per cent labradorite, 4 per cent nephelite, 1.3 per cent analcite, 26 per cent titanaugite, 32 per cent olivine, and small amounts of ilmenite, biotite, and apatite, belongs here also. It is a plutonic rock occurring in the Kyle district of Ayrshire, whence its name. Syn.: Olivine-theralite. (3330) No plutonic rock has been located in this family, but a rock described as a pikrit-basalt by Quensel’ from Juan Fernandez occurs among the extrusives. It contains 50 per cent or more of olivine. CLASS 3, ORDER 4 (3414) Ricolettaite. The only rock located here is a dark calcic gabbro from Traversellitthal, north cliff of Ricoletta, Monzoni. It consists of orthoclase 5 to 7 per cent, anorthite 35 to 4o per cent, pyroxene 4o per cent, and a little biotite, olivine, and magnetite. It was described by Doelter* and deserves a new name. (3415) Yamaskite Younc. This anorthite-augite rock from Mount Yamaska, Quebec, was described by Young.’ A similar rock, but carrying olivine, was described by O’Neill.® Olivine-yamaskite. *H. Rosenbusch, of. cit. (4th ed., 1907), p. 413. 2G. W. Tyrrell, ‘‘The Late Palaeozoic Alkaline Igneous Rocks of the West of Scotland,” Geol. Mag., IX (1912), 121. 3P. D. Quensel, ““Der Geologie der Juan Fernandezinseln,’ Bull. Geol. Inst. Upsala, XI (1912), 265. 4C. Doelter, ‘‘Chemische Zuzammensetzung und Genesis der Monzonitgesteine,”’ Tscherm. Min. Petr. Miith., XXI (1902), 102. 5G. A. Young, “Geology and Petrography of Mount Yamaska, Quebec,” Geol. Surv. Canada, Ann. Rept., XVI, Part H (1906), p. 16. 6J. J. O'Neill, ‘St. Hilaire (Beloeil) and Rougemont Mountains, Quebec,” Geol. Surv. Canada, Mem. 43 (Ottawa, 1914), p. 66. CLASSIFICATION OF IGNEOUS ROCKS 225 (3430) Several leucitites, so called, fall here, although they are not true leucitites on account of their feldspar content. No plutonic rocks have been located. CLASS 4, ORDERS I TO 3 No attempt is made in this paper to separate the orders of Class 4, since too few modes have been found in the literature to warrant the separation, and measurements by the author, so far, have been chiefly on the rocks of the first three classes. It is hoped shortly to give the modes of some of the classic types of Class 4. Family 1.—Dunite von Hocustetter. This rock, named by von Hochstetter* from the Dun Mountains, New Zealand, consists of olivine with accessory chromite. The amount of chromite varies greatly. Vogt? says that the normal rocks carry from 2 to 5 per cent. They would thus belong to Order 1. Dunites with between 5 and 95 per cent chromite he calls chromite-dunites. With over 95 per cent chromite the rocks may be classed as chromite ores of Order 4. They are represented by the plagioclase-free varieties of Sjogren’s* Kromit-Olivinit. In other cases the ore is magnetite, as in the magnetite-olivinite series of Sjégren, which includes plagioclase-bearing and plagioclase-free olivine-magnetite rocks. Perhaps good subdivisions would be: Order 1. Dunite. Olivine between too and 95 per cent. Order 2. Chromite-dunite and magnetite-dunite. Olivine between 95 and 50 per cent. Order 3. Olivine-chromitite and olivine-magnetitite. Olivine between 50 and 5 per cent. Order 4. Chromitite and magnetitite. Olivine less than 5 per cent. Families 2 and 5.—These are the families of the mica- (amphibole-)olivine rocks. Among them are mica-peridotite and t Ferdinand von Hochstetter, “Dunit, kérniger Olivinfels vom Dun Mountain bei Nelson, New-Seeland,”’ Zeitschr. d. d. geol. Gesell., XVI (1864), 341. 2 J. H. L. Vogt, “ Beitraige zur genetischen Classification der durch magmatische Differentiationsprocesse und der durch Pneumatolyse entstandenen Erzvorkommen,” Zeitschr. f. prak. Geol. (1894), p. 301. 3A. Sjogren, “Om férekomsten af Tabergs jernmalmsfyndighet i Smaland,” Geol. Foren. i Stockh. Férhandl., 111 (1876), 58. 226 ALBERT JOHANNSEN amphibole-peridotite. Mica-peridotite was named by Diller™ from an occurrence in Kentucky. In the type rock the olivine and its alteration products form over 80 per cent, the ores, magnetite, and ilmenite together 4.2 per cent, while biotite, garnet, etc., are accessory. The rock number is (412). Amphibole-peridotite was first described as amphibole-gabbro by Howitt.2 Later Rosenbusch? compared the same rock with the Schreisheim dike, and still later Verbeek* called it amphibole- peridotite. Hornblende-peridotite is the usual variety. Rocks of this class have been called hornblende-picrites by Bonney® and hudsonites by Cohen,° but since picrite was originally used for an olivine-augite rock and hudsonite for a variety of diallage, Williams? proposed for them the name cortlandtite. A rock whose present mode places it in Family 5 is scyelite Jupp.® It consists of hornblende 58.5 per cent, serpentine after olivine 22 per cent, mica 18.5 per cent, and magnetite I per cent. Judd says that the hornblende is probably secondary after augite; therefore it possibly should be placed in Family 11. Families 3, 6, and ro.—In these pigeonholes fall valbel- lite SCHAFER? and some of the olivinites of SyOGREN’ and tJ. 5S. Diller, ‘‘Mica-Peridotite from Kentucky,” Amer. Jour. Sct., XLIV (1892), 289; also ‘‘Peridotite of Elliott County, Kentucky,” U.S. Geol. Surv., Bull. 38 (1887), p. 11. 2 A.W. Howitt, ‘““The Diorites and Granites of Swift’s Creek and Their Contact Zones, with Notes on the Auriferous Deposits,” Proc. Roy. Soc. Victoria, 1879. 3H. Rosenbusch, review of preceding article, Newes Jahrb., I (1881), 221. 4R. D. M. Verbeek, Topographische en geologische beschrijving van een gedeelte van Sumatra’s Westkust (Batavia, 1883), p. 304; also “‘ Description géologique de tel d’Ambon,” Jaarboek van het Mijnwegen in Nederl. Oost-Indie, XXXIV (10905). 5’T. G. Bonney, ‘On a Boulder of Hornblende-Picrite Near Pen-y-Carnisiog. Anglesey,’ Quart. Jour. Geol. Soc., XXXVII (1881), 137. 6 E. Cohen, ‘“‘Berichtigung beziiglich des ‘Olivin-Diallag-Gesteins’ von Schries- heim im Odenwald,” Neues Jahrb., I (1885), 242. 7G. H. Williams, ‘‘The Peridotites of the ‘Cortlandt Series’ on the Hudson River Near Peekskill, New York,” Amer. Jour. Sci., XXXI (1886), p. 30, note. § John W. Judd, ‘‘On the Tertiary and Older Peridotites of Scotland,” Quart. Jour. Geol. Soc., XLI (1885), 401-7. 9 Raimund William Schafer, ‘‘Der basische Gesteinszug von Ivrea im Gebiet des Mastallone-Thalles,”’ Tscherm. Min. Petr. Mitth., XVII (1807), 512-14. 70 A. Sjégren, ‘‘Om férekomsten af Tabergs jernmalms fyndighet i Smaland,” Foren. t Stockh. Forhandl., III (1876), 58. a CLASSIFICATION OF IGNEOUS ROCKS 227 ErcustAptT.* Valbellite consists of bronzite, olivine, and brown hornblende in variable amounts, with accessory magnetite, green spinel, and pyrrhotite. The magnetite may be very abundant. Olivinite covers a group of rocks of varying composition. Essen- tially they contain olivine with augite and hornblende. Anorthite may be present in some occurrences, but in general it is rare. Another rock belonging here is Saitzew’s? hornblende-diallage- peridotite from the Koswinski-Kamenj, in the Urals, later described by Duparc and Pearce’ under the name koswite. This rock con- sists of much diopside, less olivine and less hornblende in a cement of magnetite, giving a sideronitic texture. Chrome spinel is also present. The main characteristic is the texture, which, the authors say, passes to that of ordinary peridotite by a decrease in the amount of magnetite. The sideronitic texture is not confined to these dikes, but also occurs in other magnetite-rich peridotites. Families 4 and 11.—This group includes the olivine-pyroxene (both orthorhombic and monoclinic) rocks. Among them are lherzolite DE LAMETHERIE,’ named from the original locality of Lherz, in the Pyrenees, and consisting of olivine, enstatite, and diopside, with accessory picotite. In other localities the pyroxene is diopside and bronzite, and chromite may be present in small amounts. Diallage-peridotite KLoos' consists of diallage, olivine, and some chromite. Wehrlite is a name given by von Kobell,® in 1834, to a rock from Wehrle in Hungary, under the impression that it was a mineral. t Fr. Eichstadt, ‘‘Pyroxen och amfibolférande bergarter fran mellersta och 6stra Smaland,” Bihang till Kongl. Svenska Vetenskaps-Akademiens Handlingar, XI, No. 14 (1887), Pp. 95, 123. 2 A. Saitzew, ‘‘Geologische Untersuchungen im Nikolai-Pawdinschen Kreise und Umgebung, im Gebiete des Central-Ural und dessen éstlichen Abhang,” Mem. Com. Geol., XIII, No. 1 (1892), p. or. 3L. Duparc and F. Pearce, “Sur la koswite, une novelle pyroxénite de 1’Oural,” Comptes Rendus, CXXXII (1901), 892-94. 4De Lamétherie, Théorie de la terre, II, 281; also Legons minéralogique, 11, 206. 5 J. H. Kloos, “Uber Uralit und die strukturellen Verschiedenheiten der Horn- blende in einigen Gesteinen des Schwarz- und Odenwaldes,” 58 Vers. deutsch. Naturf. u. Arze (Strassburg, 1885); also ‘Studien im Granitgebiet des siidl. Schwarzwaldes,”’ Neues Jahrb., III (1884), 146. 6 Franz von Kobell, Geschichte der Mineralogie (Miinchen, 1864), p. 660. 228 ALBERT JOHANNSEN The name is now generally applied to peridotites with olivine and much diallage or augite, although the type rock actually contains considerable hornblende and belongs to Family 3, 6, or to. Harzburgite RosENBUSCH™ and saxonite WADSWORTH’ were applied to peridotites composed of olivine and a monoclinic pyrox- ene. Vogt,3 following Brégger, uses saxonite for the iron-poor olivine-enstatite rocks and harzburgite for the iron-rich members. A saxonite from Minnesota, described by Hall,‘ consists of enstatite ~ 60 per cent, olivine 35 per cent, and ores 5 per cent, consequently belongs to (4211). Families 7 and g.—These are the families of the olivine-free amphibole-(or biotite-)pyroxene rocks. Here belong Cromaltite SHAND,® consisting of aegirite-augite 51.9 per cent, melanite 15.6 per cent, biotite, apatite, and ores. Its number is (4209). A hornblende-hypersthenite, named bahiaite by Washington,° with the percentages hypersthene 46, augite 5, hornblende 40.7, and ores 8.3 (rock number 429), also belongg here. Washington says olivine is negligible in bahiaite, but one such rock described by him’ contains 7.5 per cent, which places it in Family ro. Family 8.—This is the family of the amphibolites and horn- blendites. Among these is a hornblendite from Brazil, described by Washington.’ It consists of hornblende 91.6 per cent, olivine 3.6 per cent, and magnetite 5.1 per cent. *H. Rosenbusch, Mikroskopische Physiographie der massigen Gesteine (2d ed.; Stuttgart, 1887), p. 269. 2M. E. Wadsworth, Lithological Studies (Cambridge, Mass., 1884). 3J. H. L. Vogt, “Beitrige zur genetischen Classification der durch magmatische Differentiationsprocesse und der durch Pneumatolyse entstandenen Erzvorkommen,”’ Zeitschr. f. prak. Geol. (1894), p. 384, note. 4C. W. Hall, ‘““The Gneisses, Gabbro-Schists, and Associated Rocks of South- western Minnesota,” U.S. Geol. Surv., Bull. 157 (1809), p. 111. 5S. J. Shand, ‘‘On Borolanite and Its Associates in Assynt,” Trans. Edinburgh eol. Soc., TX Gone), 304. 6 Henry S. Washington, “The Charnockite Series of Igneous Rocks,” Amer. Jour. Sci GIL Goro) .33n-—32. 7 Ibid., “An Occurrence of Pyroxenite and Hornblendite in Bahia, Brazil,” Amer. Jour. Sci., XXXVIII (1914), 86. OTs, We CP i el i aia anil ee CLASSIFICATION OF IGNEOUS ROCKS 220 Another rock belonging here is the biotite-pyroxenite from New Zealand described by Hutton.’ It consists of biotite and horn- blende in about equal proportions. Family 12.—Finally, in Family 12, belong the pure pyroxene rocks, such as diallagite, bronzitite, hypersthenite (together called pyroxenolites by Lacroix).? Websterite Witt1aMs,? named from Webster, North Carolina, contains both orthorhombic and mono- clinic pyroxenes with accessory iron ores. Ilmenite-enstatitite Vocr! contains as much as 60 per cent ilmenite; consequently it ranges from Order 2 to 3. A magnetite- pyroxenite described by Jennings’ and Bastin® also belongs to Order 3. It contains 67.25 per cent magnetite, 8.30 per cent ilme- nite, 16.89 per cent diopside or augite, 6.70 per cent spinel, and .18 per cent apatite. APPENDIX This appendix might be introduced by the Spanish proverb: “El sabio muda consejo, el necio no,” were there no danger of some critic replying with: ‘‘Prudentis est mutare consilium; stul- tus sicut luna mutator.” In the first instalment of this paper (p. 38), the writer spoke of a contemplated change by which seventy-two families were to be omitted, but letters sent to a considerable number of petrog- raphers found no uniformity of opinion. Some were in favor *F. W. Hutton, “On a Hornblende-Biotite Rock from Dusky Sound, New Zealand,” Quart. Jour. Geol. Soc., XLIV (1888), 745-46; also “‘The Eruptive Rocks of New Zealand,” Jour. and Proc. Roy. Soc. New South Wales, XXIII (1889), Part I, p. 154. 2 A. Lacroix, ‘‘Sur les roches basiques constituant des filons minces dans la lherzo- lite des Pyrénées,”’ Comptes Rendus, CXX (1895), 752-55. 3 George H. Williams “‘ The Non-feldspathic Intrusive Rocks of Maryland and the Course of Their Alteration,’’ Amer. Geol., VI (1890), 40-41. 4J. H. L. Vogt, “Bildung von Erzlagerstatten durch Differentiationsprocesse in basischen Eruptivmagmata,” Zeitschr. f. prak. Geol. (1893), p. 8. 5 E. P. Jennings, ‘A Titaniferous Iron-Ore Deposit in Boulder County, Colorado,” Trans. Amer. Inst. Min. Eng., XLIV (1913), 14-25. 6 Edson S. Bastin, ‘‘ Economic Geology of Gilpin County and Adjacent Parts of Clear Creek and Boulder County, Colorado,” U.S. Geol. Surv., Prof. Paper 94 (1917), Pp. 47- 230 ALBERT JOHANNSEN- of dropping all the rocks of the monzonitic series, others of retain- ing all; some favored dropping the quartz-monzonites but not the monzonites, others dropping the monzonites but not the quartz-monzonites. During the past year, while the foregoing paper was awaiting its turn for publication and while going through the press, the writer collected and determined many more modes and found that the monzonitic group is unnecessary except pos- sibly in the granite-granodiorite and syenite-diorite series. Now apparently the only monzonites which anyone desires to retain are those of these two groups. Consequently the writer believes the following-changes will satisfy all sides. All classes and orders are to be subdivided as stated on pages 38 to 40. Families are to be divided and numbered hereafter as shown in Figure 7 instead of as in Figure 4, the divisions Beso [OS 9 IO II 12 13 14 15 16 17 18 19 20 25 Fic. 7.—The new subdivisions for families being at o-5-50-g5-100. In the case of the quartz-monzonites- granodiorites, and syenites-monzonites-monzodiorites (Families 7-8-9, and 12-13-14, Fig. 4), subdivisions may be introduced if desired with the o-5—35-65-95—100 per cent limits as before. In such cases they will be numbered, to conform with the new family numbers, 5, 6’, 6-7, 7,8, 0,10, 10-11, TD, 102) Um@enerone: for example: 226 is normal granite with the Kf-Plag ratio between 95-5 and 50-50. 2210 is normal syenite with the same K{-Plag ratios. eS CLASSIFICATION OF IGNEOUS ROCKS 231 227 1s granodiorite as originally intended by Lindgren (p. 168 above) with the Kf-Plag ratio 50-50 to 5-95. (Family 237 is granogabbro.) 2211 is normal syenodiorite (2311 syenogabbro) with Kf-Plag ratio like the preceding. For those who wish to use the monzonitic subfamilies: 226’ is more limited granite with the Ki-Plag ratio between 95-5 and 65-35. It may be called monzogranite. 2210 is limited syenite with the same Kf-Plag ratios. It may be called sonzosyenite. 226’—7' is quartz-monzonite or adamellite, Kf- Plag ratios 65-35 to 35-65. 22102211’ is monzonite, with same ratios as preceding. 227’ is limited granodiorite with Kf-Plag ratios between 35-65 and 5-95. Better called monzotonalite, leaving granodiorite for 227. 2211’ is monzodiorite, same ratios as preceding. Thus petrographers who wish to use quartz-monzonite and monzonite have subfamilies at their disposal and these subfamilies may be carried through the whole system if desired. Normally only the families would be used. So far as names are concerned, they remain as given in the preceding article except that Families Be Oe een Con eoe hand. .28 drop out entirely and the adjacent families (and their Haines) extend to the center line. In Class 4 the subdivisions are to be made on the same divi- sions, therefore the left face of Figure 5 should be divided as the upper half of Figure 7 and the positions computed in the same manner as the families in the other classes. In Part I, therefore, the following changes are to be made in the rules on pages 43 and 44. Page 43, 3 lines from bottom, omit five. Page 44, lines 7-8, for o-5—35-65-95-100 read o—5—50-9 5-100. Page 44, lines 12-14, for the mineral of one corner, etc., to end of sen- tence, vead olivine to the sum of the biopyriboles, and of biotite plus amphi- bole to pyroxene. Another correction that should be made is on page 42, where zinnwaldite is to be taken away from the auxiliary constituents uae added to the dark micas under mafites. 232 ALBERT JOHANNSEN The principal advantage of the system as it now stands is in the rapidity with which a rock may be classified. It is much easier to determine whether the potash feldspar is greater or less in amount than the plagioclase than it is to estimate whether it falls between the 5—35—65-95 limits. GEOLOGICAL SETTING OF NEW MEXICO CHARLES KEYES Des Moines, Iowa The unique distinction which New Mexico holds in American geology is that it is the meeting-point of four major and diverse geographic provinces. Together these four provinces embraces nine-tenths of the North American continent. Effects of general land depletion under widely different climatic conditions are thus rarely so strongly contrasted. Situated well within the boundaries of the vast southwestern desert, the operations of the epicene geologic processes are rendered the more conspicuous because of the fact that they are so very different from those of the pluvial eastern parts of. the country with which most of us are most familiar. New Mexico is distinctly a mountainous country. Its orogeny, however, is chiefly erosional rather than tectonic. Relief of the area is characteristically that of a land of little rain. Facial expression of the region is clearly not stream-corraded but wind- abraded. Owing to the fact that the wind sweeps up its chips as fast as it cuts them the magnitude of eolic erosion is at first difficult to measure with any great degree of satisfaction. Except under especially favorable conditions definite figures cannot always be given. Only when a desert chances to have, somewhere within its boundaries, remnants of old peneplains highly uplifted may the extent of regional depletion be closely estimated. As do moist lands under the influences of stream activities, so arid regions soon develop strong contrasts of surface relief under wind scour. Belts of weak rocks are soonest worn down, leaving the hard rock masses protruding as mountains In a region of uniform flat-lying strata the relief contrasts are not always marked. When, however, there are rock beds of great thickness, alternating hard and soft members, with close-patterned mountain structures as in the arid lands of western United States, 238 234 CHARLES KEYES differential relief effects attain maximum extremes. In this tract it is that the seeming youthfulness of the lofty desert mountains is at once impressed with amazing vividness upon the mind of the observer fresh from his pluvial homeland. | (— EVEL Fic. 1.—Physiographic regions of New Mexico Once clearly discriminated, wind-graved relief expression is seldom mistaken for any other kind. Its individuality is very strong. Wind-beveled surfaces are smoother than water-formed plains possibly can be. The rock floors which characterize so many desert plains are phenomena as novel as they are unexpected. _ a eS ee ee a ee GEOLOGICAL SETTING OF NEW MEXICO 235 Desert ranges rising abruptly out of the plains about impart char- acteristic form to the enisled landscape. The girdled mountain attests the vigor of natural sand-blast action; and its maximum effectiveness is at the plains line. Plateau plains of the desert manifestly represent former levels of the general plains surface. The notable absence of foothills around the mountain bases appears to be an idiosyncrasy of arid lands. Arid planatation takes place uphill as well as down; anti- gravitational gradation is unknown where streams erode. High gradients of the intermontane plains and strong pitch of valley ‘pretasdous:2:/‘~a — sien ae eeaea 7Sandstones: * bs ie MOR SES 2a Fic. 2.—Sierra de los Caballos: origin of enisled relief axes which are displayed on every hand are not possible in regions where water action is directly the reverse of plains forming. Of minor features attributable to wind abrasion in the lands of little rain, there are a multitude that have been ascribed to normal water corrasion but that are now known never to have been touched by stream. Upon all of these the wind marks when once pointed out are unmistakable. It so happens that the broad arid tract of southwestern United States has within its borders abundant traces of former base-level plains, one of which peneplains is now raised more than two miles above sea-level. The attainment of its present position is regarded as having taken place in late Tertiary times. This great peneplain no doubt once extended over all of this desert region, at a level 236 CHARLES KEYES somewhat above the tops of the present mountains. Inasmuch as desert conditions began to set in about the same time, there is every reason to believe that the magnitude of the general land depletion is represented by the difference between this old pene- plain level and the present plains level—an interval of between 5,000 and 6,000 feet—or something over one mile of thickness, over an area equal to nearly one-quarter of that of the entire United States. There are many considerations supporting the assumption that this area was before recent uplift a vast plain rather than a moun- tainous tract when arid climate was inaugurated. Inappreciable aid from stream corrasion in this prodigious regional depletion is indicated by the very fact of the prevalence of aridity itself. This region is one of the best extant, demonstrating beyond peradven- ture the almost boundless potency of the wind as an epicene power in re-forming the face of earth. For general purposes of earth study no part of our land is so favored as New Mexico. More diverse phenomena are crowded together in limited space than perhaps anywhere else on our globe. Every known category of geologic process appears to be repre- sented. Every known cause of geographic product seems to have been in operation. In great variety and with diagrammatic dis- tinctness of textbook illustration are the larger rock structures displayed. Everything, too, is on such a gigantic scale. Such phenomena as dikes and faults have to be viewed from afar, from distances of miles, in order to get proper perspective of their relations. Orographic features, which are usually assumed to be structural, are found to be mainly erosional. From the lofty cornice of the Sandias a landscape prospect spreads out a distance equal to that from New York to Chicago. Billows of mountain ridge take on an aspect of choppy sea as viewed from the deck of an ocean liner. The silver thread of the Rio Grande glints in the light of the desert sun for 400 miles until finally lost on the verge of the world—so clear is the dry, thin air of the desert. The crystalline framework of the region is both varied and substantial. Crystalline schists of the fundamental complex are _ os Yee a GEOLOGICAL SETTING OF NEW MEXICO 237 abundant and of many colors. If he could have viewed them they would surely have gladdened the heart of Johannes Lehmann. Had the grand old pioneer in the field chanced to dwell in this country instead of Saxony the Erzgebirge might never have become so famous; and Die Untersuchungen uber Entstehung der altkrystallinischen Schtefergesteone might have had the desert range of New Mexico for the central theme of his great epic. The massive plutonic rocks, the granites, diorites, gabbros, and diabases, cut and intrude the pre-Cambrian complex in all directions. ‘Their surface types appear to have utterly disappeared, probably during the long periods of erosion as indicated by the great unconformities. Volcanics belong to all ages from the Cretaceous onward. San Mateo is one of the majestic volcanic piles of the continent. It now stands perched high upon a lofty Cretaceous pedestal. A forest of volcanic necks stretches away to the northward. Modern basalt flows cover hundreds of square miles. Older lava streams constitute the resistant cap of many plateau plains. Intrusive sheets run for scores of miles across the plains like great walls of cyclopean masonry. Laccolithes (Fig. 3) display their tectonic origin rather than formation through simple hydrostatic welling. Cinder cones are numerous (Fig. 4). Beside the fine cone of Capulin that of famed Vesuvius sinks into utter insignificance. The phenomena of classic Auvergne and the Phlegrean fields are reproduced again and again but on grander scale. Stratigraphical succession in New Mexico is instructive to an extraordinary degree. It is, perhaps, the most complete to be found in any state in the Union. Curiously enough, above the pre-Cambrian crystalline complex highly resistant rocks compose the lower half of the geological column; while in strong contrast weak friable beds are segregated in the upper half. With a close- patterned orogeny this disposition has telling effect upon the final relief expression. The geological column attains an enormous thickness. Arch- eozoic, Proterozoic, Paleozoic, Mesozoic, and Cenozoic sections are each about 10,000 feet in vertical measurement. Altogether there are 50,000 feet of strata reposing upon the non-clastic Azoics. It is 238 CHARLES KEYES * one of the great critical sections of the country. It is in reality the standard succession of the Western continent. It is a ‘rock sequence such as James Hall vainly sought in his lifelong endeavor Fic. 3.—Uncovered laccolith of Multiplex Ortiz Fic. 4.—Zuni Salt Lake: birth of a volcano. Two cinder cones on floor of explosive crater, 600 feet deep. to establish a standard stratigraphy for the world. It thrice transcends the united sections of Murchison, Sedgwick, and Lons- dale. With the depositional equivalents of its numerous hiatuses GEOLOGICAL SETTING OF NEW MEXICO 239 it constitutes the longest and best sedimentative record of which we know. Huge as is the sedimentative prospect, erosive depletion looms up in even vaster proportions. Thirty major unconformities stand for a very much longer interval of time than that which deposition occupied. Nowhere else on the face of the earth does it seem that the stratigraphic record is so clearly defined for a perfectly independent classification of geologic terranes according to diastrophic movement. It is the one place of all where orotaxial principles should sustain themselves under severest test. Any world-scheme of formational arrangement must stand or fall when fitted to this titan among rock sections. That the pre-Cambrian rocks beyond the southern Rocky Mountains have never received the attention which they really merit recent disclosures amply attest. It seems possible that some day ere long they will divulge as clear a succession as did the transi- tion rocks to English geologists a century ago. At least this is certainly the most promising field for new and large results that the American continent today offers in stratigraphy. Three grand successions are presented. There is first at the top a thick section composed of relatively slightly metamorphosed and mildly deformed rock masses; then, in the middle, separated above and below, by a great erosional unconformity, a sequence of terranes highly altered, closely flexed, and repeatedly broken through by eruptives of various types; and third, at the bottom, an intensely metamorphosed and sheared complex in which no signs of classic origin are discernible. ‘These grand successions are respectively Proterozic, Archeozoic, and Azoic. To the Azoic basement are referred all of those lowest pre- Cambrian masses which, intensely altered and profoundly deformed, present no evidence of sedimentary origin. If any such classic character ever existed all trace is now completely obliterated. | This intensely metamorphosed complex lying at the very bottom of the exposed rock column is a new find. Its discovery is yet too recent to enable its full significance to be properly evaluated. Composing this fundamental complex are mainly thinly foliated gneisses, micaceous schists, squeezed granites, and other sheared 240 CHARLES KEYES eruptives. The slightly altered granites, diorites, and diabases which cut the mass are manifestly relatively late intrusives. Their evident enormous extent, their intensely metamor- phosed condition, their extensive deformation, their sharp lithologic contrast with the superior metamorphics, and the marked eros‘onal unconformity dividing the two successions all attest the supreme antiquity of the complex. The depositional equivalent of the summital unconformity may itself surpass in duration the strati- graphic record of the entire Paleozoic section. The best develop- ments of the typical non-clastic Azoics are in the southwestern portions of the state. The Archeozoic platform is bounded both above and below by marked erosional planes of unconformity. The rocks are all highly metamorphosed. The presence of quartzites, slates, and marbles indicates the clastic origin of a large part of the mass. The section is very thick, possibly not less than two miles. In marked contrast with the inferior Azoic rocks are the evi- dences of a clastic origin of the major portion of the section, and a distinct lithologic sequence is plainly discernible. Unconform- ities which are associated may correspond to those shown in the Grand Canyon, but it is believed that some of the latter are super- posed in New Mexico. It is really the Archeozoic rocks chiefly which heretofore have been called Archean, the Azoic masses not being recognized and the Proterozoic segregation not being dif- ferentiated. A thick sequence of pre-Cambrian clastics which are only slightly altered is displayed in the Tijeras Canyon, in the Sandia Range east of Albuquerque, where in a sharp fold a mile of strata outcrops in continuous horizontal section. In discontinuous exposure at least another, mile of beds is evidently present. The strata are chiefly shales, locally more or less indurated with some quartzite beds and intrusive granites. The quartzite beds which stand at high angles are commonly mistaken for immense quartz reefs, and under such misconception they are extensively prospected for gold. Microscopical examina- tion in thin slices demonstrates conclusively that the rock has a clastic origin, and that it is an old sandstone indurated by the ee a eT GEOLOGICAL SETTING OF NEW MEXICO 241 interstitial deposition of silica disposed in optical continuity with the separate sand grains. ‘The enormous thickness of the shales is especially noteworthy. It is probable that eventually they will yield an extensive fauna, or a succession of faunas. Should the rock section prove fossiliferous the opportunity for determining faunal sequence would certainly be as favorable as among any Paleozoics of the continent. Where the crest of the great Tijeras fold of the Proterozoics is deeply beveled off flat-lying limestones of latest Paleozoic age recline directly upon it. This notable unconformity plane repre- sents a period of time of vast duration, one almost coextensive with the Paleozoic era. No less than ten great erosion intervals are superposed one upon another. Although the early Paleozoic strata are entirely absent over the northern half of New Mexico they are extensively developed in the southern part of the state. There some of the major terranes were identified and correlated with the European sections almost as soon as they were in the Upper Mississippi Valley, or within a decade of the first appearance of Murchison’s and Sedgwick’s classic works. When in 1874 undoubted Cambrian beds were first recognized by Jenney in New Mexico only unimportant sandstones were dis- closed in the Franklin Mountains north of El Paso. Since that date the extent of the section has been greatly expanded, until now over 1,000 feet of sandstones, quartzites, and limestones are known. Westward these formations connect with the great Cambrian sections of Arizona. As indicated by the contained fossils both mid- and late- Cambrian sections are fairly well represented, the former by about 700 feet of strata and the latter by 400 feet. ‘There are no evidences of the presence of early Cambrian beds within the borders of the state, and it does not seem likely that any ever will be found. Ordovician strata were the first Paleozoic rocks recognized within the limits of New Mexico. As early as 1848 Wislizenus recorded the finding of Ordovician (lower Silurian) fossils in rocks west of El Paso. A few years later both Shumard and Antisell 242 CHARLES KEYES discovered similar organic remains in the mountains lying to the north of the same point. The section developed rapidly until it reached a thickness of more than 500 feet, extended entirely across the southern part of the state, and comprised three important series corresponding to the three subdivisions of the period. These limestones are abundantly fossiliferous. The forms indicate the same sub-periodic divisions that are commonly recog- nized on the eastern side of the continent. Much of the bottom and top of the general section for the continent is missing in New Mexico. Between the Ordovician beds and the Cambrian below and the Silurian above marked unconformities prevail so that the first-mentioned unit is sharply defined. Silurian sediments are poorly developed. The rocks out-crop in a thin broken line across the southwestern corner of the state. That deposition during the period was extensive in this region is quite manifest; but it is also evident that during Devonian and Mississippian times the deposits were largely removed through profound erosion. ‘The contained fossils indicate only the Niggoste horizons of the standard eastern section. From the character of the organic remains the presence of Devonian rocks in New Mexico receives early announcement. Both Antisell and Hall in 1856 call attention to the fossil evidence. Dutton’s statement that Devonian strata were generally wanting in the eastern part of the Colorado High Plateau region is some- what misleading. These rocks are really very much better repre- sented in southwestern New Mexico than has been commonly supposed. In the vicinity of Santa Rita, in Grant County, are 4oo feet of light-colored, fine-grained limestones and shales which carry abundant organic remains. Two parts of Devonian time appear to be represented. The basal shales seem to belong to the mid-Devonian section; while the limestones are late Devonian in age. A surprising feature is that the fauna is the typical Lime Creek phase of northern Iowa. This horizon corresponds to the uppermost part of the section represented in the Upper Mississippi Valley, which is, according to Tschernyschew, Williams, and others, who have given the subject most attention, mid-Devonian in age. These authorities also GEOLOGICAL SETTING OF NEW MEXICO 243 agree on the appearance of the fauna in the West, or Mississippi Valley region, much earlier than in the East, or New York province; so that, although its age in the first-named locality is strictly mid- Devonian, in the second locality it is late Devonian. This being the case, the same migratory fauna may have put in an appear- ance in the southwestern or New Mexican province, considerably earlier even than it did in the Upper Mississippi Valley region— possibly in the middle or latter part of early Devonian time. The Mississippian succession in New Mexico comprises only a limited portion of the middle division of the Mississippi Valley section. Both the upper, or Tennesseean, and the lower, or . Waverleyan divisions seem to be entirely missing. Because of its restricted representation and its other notable peculiarities the southwestern sequence is designated the Socorran series. The maximum thickness of these rocks is about 400 feet. Both the Chouteau and Burlington faunas of the Mississippi Valley are well represented. The list of organic remains is almost as extensive as that of the type localities. Fossils from the Modoc limestone, which extends eastward from Arizona, indicate higher horizons, probably so high as the Keokuk level. Curiously, the huge gastropods which are so rare in the East are very abundant | in the Southwest. The especial importance of the New Mexican section of the _Mississippian rocks is that it directly connects the eastern or Ozark sequence with that of the Far West. Summing up the evi- dence regarding these strata as they are represented around the border of the Colorado High Plateau it appears that there is in the Southeast and South, in southwestern New Mexico and eastern Arizona, the Socorran series; in the West, in northwestern Arizona, the Lower Red Wall formation; in southwestern Colorado the Ouray (upper part) limestone; and east of the southern Rocky Mountains the Millsap limestone. In New Mexico the succession _is probably most completely represented and most varied, although perhaps not including all that is represented in Colorado. In strong contrast with the Pennsylvanian sequence of the middle and eastern portions of the continent the southwestern succession is almost unbrokenly calcareous. A single plate of 244 CHARLES KEVES limestone, which extends over the entire area of the state, reaches far into Texas on the one hand, and on the other hand completely over the Colorado dome to the Grand Canyon, where it appears as the Aubrey limestone, has in New Mexico a thickness of 5,000 feet. These unbroken limestones are the open-sea analogues of the coastal coal measures of the Mississippi Valley. Where in lowa they are mainly represented by a great erosional interval, and in Arkansas by 20,000 feet of shore deposits, in the New Mexican field coal-bearing measures are all but completely vanquished. Something of the enormous Arkansan series seems to find expres- sion in the diminutive Ladronesian series with its bare 200 feet of Alamito shales and a scant foot of coal. The last-mentioned for- mation, which no doubt was once one of very considerable magni- tude, is almost effaced through erosion which took place before the laying down of the limestones, which in marked unconformity rest directly upon these shales. Pennsylvanian, or upper Carboniferous, limestones attain a thickness of 2,000 feet in northern New Mexico. There they recline directly upon the surface of the old pre-Cambrian basement, all else of the Paleozoics being absent. In the South, where they reach a measurement of upward of 5,000 feet and are known as the Hueco succession of limestones, they repose successively upon Mississippian, Devonian, Silurian, Ordovician, and Cambrian formations. In the North the great limestone plate is best known in the lofty Sandia and Manzano ranges, east of Albuquerque. Through- out this district it appears to be broadly separable into two strongly contrasted formations—a lower shaly and black limestone, and an upper member comprising chiefly massive blue and gray limestones. To the inferior member, 1,000 feet thick, the title “Lunasian series’ is given. To the upper sequence, also about 1,000 feet in thickness, the term ‘“‘Maderan series’? seems most appropriate. On the basis of the determined faunal characteristics the stratigraphical position of the Lunasian series appears to corre- spond nearly with that of the Missourian series of the Mississippi Valley. Upon the same grounds the Maderan series is paralleled re a ee GEOLOGICAL SETTING OF NEW MEXICO 245 with the Oklahoman series of the eastern field, but which some- Kansas workers are prone to correlate with the far-off Permian rocks of the Russian Urals (Fig. 5). feo Se aereeee canes aoe ST ar weer RET TA RMR a i NEW MEXICO Fic. 5.—Areal range of periodic formations As Pennsylvanian time was ushered in with the deposition of sands and clays, so also did it close. Succeeding the huge lime- stone plate is the Bernalillan series of “Red Beds.” As first pointed out in 1904! the “‘Red Beds”’ problem which so long had baffled t Report of the Governor of New Mexico to the Secretary of the Interior for 1903, p. 339 (Washington, 1904). 246 CHARLES KEYES geologists found a ready solution in the discovery that these deposits were not accumulations of a single geologic age, but, in their different parts, of three distinct ages. Beds of similar litho- logic aspect of Pennsylvanian, Permian, and Triassic ages were directly superposed, the normally intervening formations being absent. This was, then, the real explanation why different inves- tigators in different localities had determined such diverse dates for their several sections. Even the unconformity planes were overlooked. The recognition of “‘Red Beds” of Pennsylvanian age was a distinct advance in the stratigraphy of the Southwest. It is a curious travesty on the fates that despite the acrimonious controversy which waged for more than a full generation over the possible presence of Permian beds in America, the one section which would have most speedily ended it remained unnoticed, albeit its fossils had long been fully made known. As early as 1860 Shumard described what he distinctly designated a Permian fauna from the Sierra Guadalupe on the New Mexico-Texas boundary; but its true significance remained in complete obscurity for half a century, when Girty accidentally pointed out its global relation- ships. Although the Guadalupan succession of sandstones and lime- stones is nearly 4,000 feet thick at the typical locality the forma- tion rapidly diminishes in force to the northward. Before the Sandias are reached the Bernalillan “Red Beds” and the Cimar- ronian “‘Red Beds” are brought together to form an uninterrupted “Red Beds” section. Cimarronian “Red Beds” clothe the backslope of the Guadalupe _ Mountains and extend northeastward far into Kansas. ‘There, and through the Panhandle of Texas, they are in turn overlaid by Triassic ‘‘Red Beds.” The fact that a marked erosional uncon- formity separates the two terranes is a recent observation. To the northwest similar merging of ‘Red Beds” in continuous sequence obtains; and the Triassic Doloresian formation immediately succeeds Cimarronian deposits. In New Mexico the strata of Triassic age are predominantly typical ‘“‘Red Beds.” There are two series of red shales and sand- GEOLOGICAL SETTING OF NEW MEXICO 247 stones separated by an erosion interval. The earlier of the two series is confined to the east side of the Rockies, while the later one is represented chiefly on the west side of the Cordillera. Their total thickness is nearly 2,000 feet. Like the Triassic sediments the Jurassic succession comprises an earlier and a later series, set off from each other by a great erosional unconformity. Together they represent a column 1,200 feet thick. The Morrisonian series is chiefly composed of argillaceous deposits. Of these the Chaquaqua (Chicago) shales occupy one-half of the entire section. They are underlaid by a basal sandstone. A notable erosion unconformity separates them from the succeeding Comanchan deposits. When in the early fifties of the last century Jules Marcou, colleague and exiled compatriot of Louis Agassiz, fresh from the Jura Mountains, introduced into this country the European title “Jurassic” it was the Tucumcari section in eastern New Mexico that was involved. Around this Cerro Tucumcari raged a storm center for a full generation. After being overborne by sheer weight of numbers and after being submerged for seventy-five years Marcou finally comes into his own. Cerro Tucumcari proves to contain the full Jurassic section of the region. Marcou erred only in including some of the Dakotan sandstone layers. The important Comanchan succession of Texas and the Gulf region finds but feeble representation in New Mexico, and then only as an attenuated border which soon vanishes completely against the Cordilleran uplift. The basal limestones of Texas are traceable around the Llano Estacado. The shales extend farther. Erosional unconformity separates its two series, and like sedi- “mental breaks mark both top and bottom of the sequence. Because of the fact that the thin shale beds are immediately under the mas- sive Dakotan sandstone they usually escape notice. The basal Dakotan sandstone of the Cretaceous succession is especially noteworthy by reason of its very wide geographical dis- tribution. In the Rocky Mountain region it is further remark- able because of its disposition on an old peneplain surface, which, although much deformed, now coincides with the tops of the highest peaks. Where the southern Rockies plunge in a triple fold 248 LasVecas Mrs. YLVAN evoaasv sy) x . SS s oe Santa Fe Mrs. re) rs = > if) <= 5 fE ° ms 11000 7 15.0.00' a ” ‘ ‘ oloradanSh. ARCHEOZOIC METAMORPHICS Fic. 6.—Ceja Glorietta: end of the Rocky Mountains CHARLES KEYES beneath the general plains’ surface of the Mexican table-land. the Dakotan sand- stone forms a magnificent inward-facing escarpment. Along the steep interior face | the Atchison, Topeka, & Santa Fe Rail- road skirts for a distance of fifty miles. There, as elsewhere on the continent, the Dakotan sandstone is one of the geologi- cal landmarks of the region. With a thickness of 300 feet it is, perhaps, the most extensive basal sandstone of which we know (Fig. 6). Immediately following the Dakotan sandstone is a fine succession of marine shales and limestones, which in New Mexico have a thickness of upward of 2,000 feet. This is the Coloradan series. A quite unusual feature is that the series preserves its subdivisional integrity through a distance of 1,000 miles, to Iowa and Minnesota, on the far side of the Continental Interior basin. A succeeding 2,000 feet of strata com- prising sandstones and shales constitute the Montanan series. It is an impor- tant coal-bearing formation, containing a larger fuel supply than do the coal meas- ures of the entire Mississippi Valley. In early accounts of the region these coals were all regarded as Laramian in age. Locally the coal seams overlying the Ortiz laccolith are changed into high-grade anthracite, the quality of which compares favorably with the best hard coals of the Appalachians. East of the Rockies the Montanan beds merge into the marine Pierre shales. GEOLOGICAL SETTING OF NEW MEXICO 249 As the uppermost member of the Cretaceous section the Laramian series embraces only a small part of the beds formerly paraded under this title. By divesting the old Laramie section of that portion which is really Tertiary in age the long drawn-out controversy concerning the true age of these beds comes to an abrupt end. With the principal coal beds referred to the Montanan series below and the Ratonan series above the so-called Laramie formation of New Mexico becomes almost as barren as the marine Coloradan series. Still, in the northwestern part of the state, the Pictured Cliffs sanstones and the Navajo shales attain a thickness of nearly 1,200 feet. Tertiary sedimentation in the New Mexican area begins with beds the deposition of which seem long to antedate the earliest Eocene formations of other parts of the world. When in the course of coal investigations in 1906 a marked erosional uncon- formity and basal conglomerate were found in the lofty Raton Mesa about 1,500 feet below its massive lava cap, it was surmised that the solution of the Laramie problem had been stumbled upon and that above the erosion plane the beds were really Tertiary in age, while those beneath that line were Cretaceous. These con- clusions were fully substantiated a decade later by Lee in an elaborate monograph on the geology of the region. The Ratonan series is, then, older than the oldest Eocene deposits of the state, the Puerco beds, which, since the days of Cope in the region, were believed to be the earliest Tertiaries extant. It is an important coal-bearing section, which fact prob- ably largely accounts for its early confusion with the original Laramie coal formations of the more northern Rocky Mountains districts. The erosion plane upon which the Ratonan sediments rest is a peneplain of wide proportions evidently worn down on the entire southern Rocky Mountains province. In its production no less than a mile of rock was removed. The Aztecan-series is represented by a basal conglomerate of very considerable thickness. Beyond the borders of the state the Arculeta conglomerate is succeeded by shales. No sediments are correlated with the erosional interval below. It is possible that 250 5 CHARLES KEYES v the basal unconformity is coextensive with the one beneath the Ratonan series on the opposite side of the Rockies, in which case the Arculeta and Maya formations are homotaxial. The term ‘“‘ Animas,”’ which is sometimes applied to the beds of the San Juan basin is preoccupied. In the Nacimientan series are ‘acluded the basal beds of the old Wasatch group. It is regarded as earliest Eocene. It is sub- divided into the Puerco clays and.the Torrejon marls and sand- stones. The deposits are doubtless entirely epeirotic in character and perhaps eolian. There is a large and varied vertebrate fauna. The outcrop constituted one of Cope’s favorite collecting grounds. The beds have a maximum vertical measurement of 1,000 feet. Marked erosional unconformities separate all of the Tertiary ter- ranes of the San Juan basin. The Chaman series comprises the main body of clays, sands, and shales of the San Juan Wasatch succession. Canyon Largo sand- stone is Newberry’s early designation. Chaco terrane covers the principal clay deposits. ‘Together they are nearly 2,000 feet thick. Of manifest later date are the Tertiaries of the Rio Grande basin. ‘These consist of the Galesteo sands and the Santa Fe marls, 1,300 feet in thickness. Over the Llano Estacado the latest Tertiary sands, 300 feet thick, are assigned to the Pecosian series. The terranal names applied to the New Mexican Tertiaries cover only the main bodies of deposits. No doubt other titles will eventually be attached to the numerous minor members. Quaternary deposits of New Mexico are chiefly desert con- centrates. They are not mainly deflated materials, but accumu- lations left behind after the main bulk of fine rock débris has been sorted out and exported. The deflated dusts come to rest far outside of the arid region. Gravel and bowlder trains are prin- cipally the results of arroyo wash. Till-like materials are brought down from the highlands. Some glacial till yet remains on the sides of the highest mountains where ice fields were once feebly represented. Adobe soil is deflated dust temporarily at rest. Some fluviatile deposits are present. Lacustrine beds are rare, very limited, and.quite ephemeral. In the accompanying chart of New Mexican terranes the latter are fitted to the Chamberlin and Salisbury classification. | GEOLOGICAL SETTING OF NEW MEXICO Zsa GENERAL GEOLOGICAL SECTION OF NEW MEXICO Thickness Eras Periods Series Terranes in Feet Rocks Present ormadaneee fees ose Waiver, at, Ch _ Press, 5750 Ellis Avenue, Chicago, Illinois, semi-quarterly, on or about the following dates: Febru “March 15, May 1, June 15, August 1, September 15, November 1, December 15. — q The ane price is $4.00 per year; the price of single copies is 65 cents. Orders for service of less than a year will be charged at the single-copy rate. {| Postage is prepaid by the publishers on all orden : the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of ‘Panama, Ha Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. {] Postage is charged extra as fol For Canada, 30 cents on annual subscriptions (total $4.30), on single copies 4 cents (total 69 cents for all other countries in the Postal Union, 53 cents on annual subscriptions (total $4.53), on copies 11 cents (total 76 cents). {[ Patrons are requested to make all remittances peek to The versity of Chicago Press in postal or express money orders or bank drafts. 2 The following are authorized to quote the prices indicated: A, PEN For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. 4. subscriptions, including postage, 21s. each; single copies, including postage, 3s. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Téri. Sanch Tokyo, Japan. Yearly subscriptions, including postage, Yen 9.00 each; single copies, = . postage, Yen 1.50 each. For China: The Mission Book ‘Company, 13 North Szechuen Road, Shanghai. Teal scriptions $4.00; single copies 65 cents, or their equivalents in Chinese money. Postage ex mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies 11 cents. Claims for missing numbers should be made within the month following the regular month of pl cation. The publishers expect to supply missing rick wee only when losses have been sustained transit, and when the reserve stock will permit. Business correspondence should be addressed to The University of Chicago Press, Chicago, Communications for the editors and manuscripts should be addressed to the Editors: of ' JouRNAL or GEoLocy, the University of Chicago, Chicago, Ill. - Contributors are requested to write scientific and proper names with particular care and i in ctati to follow the form shown in the pages of the Journal. Separates, if desired, must be ordered in advance of publication. Separates of eine articles printed as ordered, of which 50 (without covers) will be furnished gratis, the actual cost of the rema. (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend the amount of work in remaking the pages into forms, presswork, paper, binding, etc. Separates cc taining half-tones may be expected to cost somewhat more, the increase depending ee shes n mb of cuts and the amount of work required upon them. es Entered as second-class matter, March 20, 1893, at the Post-office at Chicago, Ill., under the Act of March 3, 1871 Acceptance for mailing at special rate of postage provided for in Section rro3, Act of October, 3 1917, July ts, 2928. an Among articles to appear in early numbers of the Journal of Geology are the following: Diastrophism and the Formative Processes. XII. The Physical States of the Planetary Nuclei'during their Formative Stages. XIII. The Planetesimals and their Infall. By T. C. CHAMBERLIN. Factors Producing Columnar Structure in Lavas and Its Occurrence Near Melbourne, Australia. By ALBERT V. G. JAMES. Pleistocene Mollusca from Indiana and Ohio. By FRANK COLLINS BAKER. The Heart Mountain Overthrust. By D. F. HEwEtTt. Recent Studies of the Upper Cretaceous of Tennessee. By BRUCE ~ WADE. Concerning the Process of Thrust Faulting. By TERENCE T. QUIRKE. Notes on the Mechanics of Geologic Structures. By WARREN J. MEAD. Geologic Reconnaissance of the Southern Part of the Taos Range, New Mexico. By JoHN W. GRUNER. The Geographic Society of Chicago | BULLETINS . . The Geography of Chicago and ‘Its Environs. Illustrated. | By Rotirn D. SALISBURY and WILLIAM C. ALDEN. 64 pages, 8vo, paper, postpaid, 38 cents; met 35 cents. . The Plant Societies of Chicago and Vicinity. JIlustrated. By Henry C. Cowes. 76 pages, large 8vo, paper, postpaid, 60 cents; met 50 cents. . Lantern Slide Illustrations for the Teaching of Meteorology. : t val ty | : ‘a \ Edited by Henry J. Cox and J. PAauL GOODE. 130 pages, i] 8vo, paper, postpaid, 85 cents; met 75 cents. and JoHn H. ARrMINGTON. xxvi+376 pages, 8vo, cloth, postpaid, $3.25; met $3.00. . Animal Communities in Temperate America. A Study in _ The Weather and Climate of Chicago. By Henry J. Cox — Animal Ecology. By Victor E. SHELFORD. xiv+362 pages, — 8vo, cloth, postpaid, $3.20; met $3.00. . Starved Rock State Park and Its Environs. By Cart O. SAUER, GILBERT H. Capy, and HENry C. COWLES. x+148 pages, royal 8vo, cloth, illustrated, postpaid, $2.15; met $2.00. O. SAUER. xvili++246 pages, 8vo, cloth, illustrated, postpaid, $3.15; net $3.00. EXCURSION BULLETINS . Excursion through the Rivers and Harbors of Chicago. Pre- pared by R. E. Brount and C. 5. JEWELL.. 20 pages, $vo, paper, postpaid, 12 cents; met to cents. . Excursion on the Rock River of Illinois between Rockford and Dixon. Prepared by R. E. Blount. 4 pages, royal 8vo, paper, postpaid, 11 cents; met 10 cents. . Stony Island. By Zonta BABER. 16 pages, 8vo, paper, post- paid, 11 cents; met 10 cents. . The Geography of the Ozark Highland of Missouri. By Car THE UNIVERSITY OF CHICAGO PRESS CHICAGO” - - ILLINOIS CONTRIBUTIONS FROM WALKER MUSEUM A NEW SERIES—VOLUME Ii INO: 1. With plates. No. 2. from Texas. HINO: 3. By Samuel W. Willistcn. The Structure and Relationships of Diplocaulus. By Herman Douthitt. 28 cents postpaid. Labidosaurus Cope, a Lower Permian Cotylosaur By Samuel W. Williston. The Phylogeny and Classification of Reptiles. 28 cents postpaid. 28 cents postpaid. No. 4. (1) The Evolvtion of Vertebrae; (2) The Osteology of Some-American Permian Vertebrates. 52 cents postpaid. W. Williston. III. By Samuel TITLES IN VOLUME | No. x. The Vertebrates from the Permian Bone Bed of Vermilion Courty, fillincis. With plates. By E. C. Case. 53 cents postpaid. No. 2. Prodromites, A New Ammonite Genus from the Lower Carboniferous. Withplates. By James Perrin Smith and Stuart Weller. 27 cents postpaid. No. 3. Palaeontolegical Notes. With plates. By E. C. Case. 27 cents postpaid. No. 4. New or Little-known Vertebrates from the Permian of Texas. With plates. By E. C Case. 27 cents postpaid. No. 5. The Stokes Collection of Antarctic Fos- sils. With plates. By Stuart Weller 27 cents postpaid. No. 6. (1) The Osteology of the Skull of the Pelycosaurian Genus, Dimetrodon. (2) On the Structure of the Fore Foot of Dimetrodon. By E. C. Case. 27 cents postpaid. No. 7. Notes on Some Carboniferous Cochlic- donts. With descriptions of seven new species. With plates. By E. B. Branson. 27 cents post- paid. No. 8. The Gsteology ef Some American Per- mian Vertebrates. By Samuel W. Williston. 26 cents postpaid. No. @. (1) The Osteology of Seme Americar Permian Vertebrates. II. (2) Synopsis of the American Permo-Carbonifereus Tetrapoda. By Samuel W. Williston. 29 cents postpaid. No. 10. (xz) Atactocrinus, a New Crinoid Genus frem the Richmond of Illinois. (2) Description ofa Ste. Genevieve Limestone Fauna from Monroe County, Illinois. By Stuart Weller. 28 cents postpaid. Vol. I, 274 pages, cloth; $2.50 (postpaid $2.70) THE Dee CHICAGO - OF Ee PRESS - ILLINOIS “N i : q ¥ POSITIONS OF ALL KINDS] || Outlines of Geologic History with Never was the demand so great for qualified teachers and specialists. For ten years we have Especial Reference to North America, given our time and energy to this work. Write for our free literature. State qualifications briefly. Co-operative Instructors’ Association Marion . . Indiana Symposium Organized by BAILEY WILLIS Compilation Edited by ROLLIN D. SALISBURY (Sie and all readers of geologic | A Series of Base Maps and Graphs Prepared by J. PAUL GOODE Professor of Geography in the University of Chicago eS maps are accurate in detail, of highest quality in drawing, being superior to most maps used in books and magazines. They are printed on a fine quality of paper and give excellent results with colored pencil, water color, or ink. Maps come in five sizes from one to forty cents each. literature will welcome the publication, | in book form, of an important series of essays and discussions on the subject of geologic correlation under the title, Outlines of Geologic | History with Especial Reference to North America, | They embody the present state of knowledge | and opinion concerning many of the funda- | mental problems of North American geology, ’ : and form an admirable supplement. to earlier f treatises and manuals. The value of the book is greatly enhanced | by the fifteen paleogeographic maps by Pail i Willis which accompany the papers. 316 pages, 8vo, cloth; price $1.50, postpaid $1.65) Write for complete list with prices The University of Chicago Presi | The University ui Chicago Press Chicago ¥ : Ilinois | Chicago . - - - Illinois THE ORIGIN OF THE EARTH By THOMAS C. CHAMBERLIN Head of the Department of Geology, The University of Chicago One of “The University of Chicago Science Series” xii+272 pages, 8vo, cloth; $1.50 net, postpaid $1.65 FROM THE PREFACE “Tn telling the story of this search for the mode by which the earth came into being, we have let the incidents that led the inquiry on from one stage to another fall in with the steps of the inquiry itself. It is in keeping with the purposes of this series of booklets that the motives which set researches going should have their place with the quests that arose from them... .. The final story of the birth of the earth will come only after a time, when the vestiges of _ creation have been more keenly discerned and more faithfully rendered than is possible now.’ THE UNIVERSITY OF CHICAGO PRESS CHICAGO : : ILLINOIS VOLUME XXVIII NUMBER 4 MoURNAL OF GEOLOGY MAY-FUNE 1920 THE CHESTER SERIES IN ILLINOIS! STUART WELLER University of Chicago PART I The Mississippian rocks in Illinois occupy three distinct areas along the western and southern borders of the state. The northern- most of these areas is the larger, and extends from southern Mercer County on the north to northern Madison County on the south. Throughout this entire distance, except for an intervalin Pike, Cal- houn, and Jersey counties, where older rocks are exposed, the rock formations of the Mississippian system constitute the Mississippi River bluffs. This area also includes the Mississippian strata which are exposed in the valley of the Illinois River as far north as Scott, Brown, and Schuyler counties. Nowhere in this area do the higher formations of the system occur, the youngest formation exposed being the Ste. Genevieve limestone, in the summit of the bluffs above Alton. The second of the three areas occupies portions of St. Clair, Monroe, Randolph, and Jackson counties. This area includes about 85 miles of the Mississippi River bluffs from a short distance below East St. Louis to the gap formed by the valley of the Big Muddy River, and at only one locality in this entire distance, at t Published by permission of the Directors of the Geological Surveys of Illinois and Missouri. 281 282 STUART WELLER Valmeyer in Monroe County, are any formations other than the Mississippian exposed. The greatest width of this area is in Monroe County, where the Mississippian formations form the sur- face rocks for a distance of about fifteen miles back from the river bluffs, where they pass beneath the Pennsylvanian strata. Within this area both the lower and upper series of Mississippian forma- tions are present, and it includes the typical area of the Chester series, as these rocks were described by Hall and Worthen more than half a century ago. The third area of Mississippian rocks in L[linois is in the extreme southern portion of the state, where these formations constitute the surface rocks throughout a belt ranging from fifteen to thirty or more miles in width, across Union, Johnson, Pope, and Hardin counties. The greater portion of this area is occupied by the upper Mississippian formations of Chester age, although the lower forma- ; tions do occupy considerable areas in Union and Hardin counties. The northwestern corner of this southern belt is separated from the central area by the valley of the Big Muddy River in Jackson County. The main portion of this paper will be devoted to a discussion of the Upper Mississippian or Chester series, although the Lower Mississippian, or Iowa series as it may be called for want of any comprehensive name already in use, will be given some considera- tion in the discussion of the geological history. The Iowa series was subdivided into a number of well-recognized formations more than half a century ago, mainly through the work of James Hall' in Iowa, although some attempt at subdivision had been made before Hall’s time, and the subdivisions and classifica- tion has been somewhat elaborated in later years. In the main, however, the divisions established by Hall constitute the formations that are recognized at this time. Not so with the upper Missis- sippian. The early workers generally recognized in this series a more or less confused succession of limestones, shales, and sand- stones, and but little attempt was made to subdivide the series. Hall gave the name Kaskaskia limestone to the whole of the suc- cession above a conspicuous sandstone formation in the Mississippi 1 Report on the Geological Survey of Iowa (1858). THE CHESTER SERIES IN ILLINOIS 283 River bluffs of Randolph County which was commonly called the “*Ferruginous Sandstone.”’ Worthen used the name Chester lime- stone for the same beds which Hall called Kaskaskia, but included this Chester limestone with the underlying sandstone in what he called the ‘‘ Chester Group.” While both Hall and Worthen based their descriptions of the upper Mississippian rocks upon observations made for the most part in the second of the areas which have been mentioned, Henry Engelmann carried on field studies in the more southern counties of the state, under the direction of the Illinois Geological Survey. In Johnson, and in the counties to the east and west, Engelmann recognized an alternating ‘succession of limestone and sandstone members of the Chester Group, ten in all, which he designated by the numbers 1 to 10, beginning the numbering at the top. The sandstones in the series received the even numbers and the lime- stone and shale members the odd numbers. ‘The only one of these members to which a distinct name was given was No. 8, which was called the Cypress sandstone’ from the good exposures in the bluffs ° of Cypress Creek, but even this name was abandoned in the later reports by Engelmann and was never used by Worthen. The real importance of the Chester series in the Mississippian as a whole is well shown by its comparative thickness. The whole of the lower Mississippian or Iowa series has a thickness of approxi- mately 1,000 feet, which was subdivided at an early time, as has been stated, into a succession of well-defined formations, but the Chester series, with a maximum thickness of more than 1,200 feet, commonly has been treated as a single formation by all geologists up to a very recent date. The first serious attempt to subdivide the Chester was made by Ulrich? in 1905. He recognized four formations as follows: 4. Birds- ville formation; 3. Tribune limestone; 2. Cypress sandstone; r. Ste. Genevieve limestone. The observations which led to this division of the Chester into definite formations were inadequate for the proper understanding of the whole series, and mistakes of so serious a character were t Trans. St. Louis Acad. Sci., Vol. I, Part 1 (1863), p. 180. 2 Prof. Paper, U.S. Geol. Survey, No. 36. 284 STUART WELLER made that it has not been possible to adapt any part of the scheme to the more recent work on the series. In the first place the Ste. Genevieve limestone was mistakenly included in the Chester Group because of the failure to recognize that the upper member of this limestone as defined, the Ohara, was really made up of two very distinct parts, only the upper one of which is really Chester, and this ‘‘Upper Ohara”’ really has no place whatsoever in the Ste. Genevieve limestone when that formation is properly limited in accordance with its typical exposures in Ste. Genevieve County, Missouri. In the second place the sandstone designated as Cypress by Ulrich was not the Cypress of Engelmann, but the bed that was properly sandstone No. 10 of that author. In the third place, beds which really belong in three totally different positions in the Chester series were designated as Tribune limestone. The limestone at Tribune, Kentucky, which gave origin to the name, has more recently been shown to occupy a position far above that designated for the formation, and is in fact representative of a limestone ‘member far up in the Birdsville formation as defined by Ulrich. At another locality the so-called Tribune is a limestone beneath the sandstone that was mistakenly called Cypress, while elsewhere it does occupy the position assigned to it in the definition of the formation, above the miscalled Cypress sandstone. In the fourth place the Birdsville formation of Ulrich comprises a succession of limestones, sandstones, and shales, and is as lacking in utility as a formation as was the older name, Chester formation. The work upon which the present paper has been based has been carried on continuously under the auspices of the Illinois State Geological Survey, from ror1r to the present time, and was preceded by more general observations upon the Chester series since 1906. From 1orr to the present time the work of mapping in detail the Chester series in Illinois has been in progress, and it has now covered the counties of St. Clair, Monroe, Randolph, Jackson, Johnson, Pope, and Hardin. The only portion of the Chester belt across the state that has not been studied and mapped in detail at the present time is in Union County and a corner of Jackson, and reconnaissance observations in Union County have se THE CHESTER SERIES IN ILLINOIS 285 shown that very little if anything new in the section can be looked for there. In the course of these studies it has been found to be necessary and perfectly practicable to subdivide the Chester series into six- teen distinct formational units, which can be distinguished and mapped with ease. The limestones of the series, with one possible exception, are all continuous across the state, from their first ~ appearance from beneath the Pennsylvanian strata in St. Clair, Monroe, or Randolph counties, on the northwest, to Hardin County at the extreme southeastern part of the belt. The sandstone for- mations, however, are not all continuous across the state; one has its greatest development in the west, thins out, and disappears to the east. Several of them have their great development in the east and become much thinner or disappear entirely in the more western portion of the state. The two uppermost sandstones of the series, however, are present uniformly across the state. This entire series of Chester formations in Illinois may be arranged in three larger groups that possess rather distinct faunal characteristics, and these three divisions may be designated as lower, middle, and upper Chester. The names of these Chester formations, with their arrangement in the larger divisions are as follows: Upper Chester Group: Middle Chester, or Okaw Group: 16. Kinkaid limestone 8. Glen Dean limestone 15. Degonia sandstone 7. Hardinsburg sandstone 14. Clore limestone 6. Golconda limestone 13. Palestine sandstone 5. Cypress sandstone 12. Menard limestone 11. Waltersburg sandstone ro. Vienna limestone g. Tar Springs sandstone Lower Chester Group: 4. Paint Creek limestone 3. Yankeetown formation, and Bethel sandstone 2. Renault limestone 1. Aux Vases sandstone In the naming of these units, those formations are designated as limestones which include notable limestone beds. In all cases such formations include a considerable amount of shale, in some 286 STUART WELLER ~ cases, locally at least, more shale than limestone, and some of them do include minor arenaceous layers. They are called limestones, however, because they are primarily calcareous as distinguished from the alternating sandstone formations. Each of these forma- tions will be considered briefly, their leading lithologic and faunal characteristics will be pointed out, as well as their geographic dis- tribution in the state, and in some cases their distribution beyond the limits of Illinois, in part at least. This will be followed by some statements concerning the geological history of the Illinois basin in Chester time, and its relations to the history of the pre- ceding Lowa time. ; LOWER CHESTER GROUP Aux Vases sandstone.—The Aux Vases sandstone is typically exposed in the Mississippi River bluffs of Randolph County, Illinois, and Ste. Genevieve County, Missouri. It is the formation that was called ‘‘Ferruginous sandstone” by the early Mississippi valley geologists, the name Aux Vases being first used by Keyes in 1892,. from the exposures in Ste. Genevieve County, Missouri, near the mouth of River Aux Vases. It was the belief of Engel- mann and also of Worthen that this basal sandstone in the Missis- sippi River section was the exact equivalent of sandstone No. 8, or Cypress sandstone of Engelmann’s Johnson County section. With such a correlation accepted, Keyes name would be synony- mous with the earlier Cypress. In the assumption that the Aux Vases—Cypress correlation was correct, the name Aux Vases was abandoned in our earlier work in Illinois. It was early recognized, however, that there was a stratigraphic break within the arena- ceous beds of the basal portion of the Chester series in Monroe and Randolph counties, and with the belief that the name Cypress covered all of these beds and that the Aux Vases was the exact equivalent of the Cypress, the name Brewerville? was used by the writer for that portion of the sandstone which lies beneath the break. Later, when studies in the more southern counties of Illinois established the fact that the Cypress and the old “Fer- ruginous sandstone’ were not equivalent, and when studies t Bull. Geol. Soc. Amer., Vol. III, p. 206. 2 Trans. Ill. Acad. Sci., Vol. VI (1913), p. £21. THE CHESTER SERIES IN ILLINOIS 287 across the Mississippi River, in Missouri, showed that the typical section of the Aux Vases was the exact equivalent of the beds for which the name Brewerville had been used, the latter name was abandoned and the name Aux Vases adopted for the lowest sand- stone formation of the Chester series in the Mississippi River section. In its surface outcrop this formation is restricted to a belt through Monroe and Randolph counties, Illinois, continuing into Ste. Genevieve County, Missouri. The formation is a very mas- sive, fine- or medium-textured sandstone in thick beds, in most places more or less conspicuously cross-bedded. Its color on freshly broken surfaces is a soft brown tint, in some localities becoming nearly white. Not infrequently it is mottled with small, dark-brown specks. On long-exposed weathered surfaces, the color in most localities is a darker brown than that of freshly broken sur- faces. The massiveness of the formation is well shown in the Mississippi River bluffs between Prairie du Rocher and Modoc and in some of the picturesque gorges which have been eroded in the formation where it is crossed by stream valleys. No fossils of any sort have been found in the Aux Vases sandstone in Monroe or Randolph counties. - The unconformable relations of the Aux Vases sandstone upon the underlying Ste. Genevieve limestone are well shown in a num- ber of places in Illinois. The uneven line separating the two for- mations can be clearly seen in the Mississippi River bluffs above Modoc. Elsewhere there is an important basal conglomerate in the Aux Vases, such conglomerates being well exposed two miles southeast of New Design, in S.E. 4, $.W. 4, Sec. 28, T. 3 S., R. g W., and again five miles southeast of Waterloo in the bluffs of Rock House Creek, in S.W. 4, Sec. 4, T. 3 S., R.g W. Still another excellent exposure of the basal conglomerate, apparently resting upon the St. Louis limestone rather than the Ste. Genevieve, is about 6 miles west of Red Bud, in S.E. 3, N.E. 4, Sec. 4, T. 4 S., R.gW. A very excellent exposure of this same basal conglomerate is exposed in the Mississippi River bluffs just below McBride, Perry County, Missouri. The pebbles in these conglomerates are practically all chert, they are more or less angular for the most part, 288 STUART WELLER and were clearly derived from the underlying Ste. Genevieve or St. Louis limestones. The presence of these conglomerate beds establishes the fact that subsequent to the deposition of the Ste. Genevieve limestone, the calcareous sediments hardened into limestones, the cherts which are clearly secondary in origin were formed, and were in essentially the same condition in which they are found today. Then an erosion period set in and in places the entire thickness of the Ste. Genevieve limestone was removed, along with a part of the St. Louis limestone. ‘The subsequent sedimentation laid down the sands of the Aux Vases formation. ‘This interruption in the deposition of the sediments of the Mississippi Valley section must have represented a considerable length of time, and it must be reckoned as an important break in Mississippian history. Other phenomena connected with this sedimentary break will be dis- cussed later, in connection with the geological history. Beyond Monroe and Randolph counties, to the south, the Aux Vases sandstone has not been certainly identified. There is, how- ever, a flaggy sandstone, about 20 feet thick, in the base of the Chester section east of Anna, in Union County, Illinois, which may be an extension of the Aux Vases, but in view of the fact that this sandstone contains numerous fossils in some beds, while the Aux Vases is quite barren of fossils, and further that sandstone beds are commonly present in the Renault formation of Monroe and Randolph counties, it is possible that this Union county sand- stone may be younger than any part of the Aux Vases, and is per- haps referable to the Renault. The maximum thickness of the Aux Vases sandstone is about 75 or 8o feet, and it varies from this amount to nothing at all, for in places the overlying Renault formation overlaps the Aux Vases and rests upon the underlying Ste. Genevieve limestone. In the more southern counties of Illinois, east from Union County, the position of the Aux Vases sandstone in the section is represented by an unconformity in the midst of the so-called “Ohara limestone member’’ of the Ste. Genevieve limestone, as described by Ulrich. Renault limestone.—It would perhaps be better to call this unit the Renault formation, for in addition to its limestone content it THE CHESTER SERIES IN ILLINOIS 289 includes much shale and sandstone. It is, however, the first epoch of calcareous sedimentation in Chester time, and while locally there were considerable accumulations of clastic materials near the shore lines of the period, at a distance from the shore the material deposited was wholly limestone and calcareous shale. The name of the formation has been derived from Renault township, the southernmost township in Monroe County. ‘The belt of outcrop of the formation crosses the whole of Monroe County in a north and south direction, and extends northward across the southwestern portion of St. Clair County and southward across the northwestern corner of Randolph County. The outcrops of this formation along Hickman Creek, in St. Clair County, are the most northerly exposures of any Chester formation. In a southerly direction the formation is exposed west of the Mississippi River across the southeastern corner of Ste. Genevieve County, Missouri, and con- tinues for a short distance into Perry County. Throughout the area of outcrop of the Renault in these Missis- sippi River counties, the formation is constituted of a very great variety of sediments, limestone, sandstone, and shale being repre- sented, with each type of rock exhibiting great variation in its lithologic characters. In fact, one of the characteristics of the formation in this typical region, is its notable heterogeneity. This great variety in sedimentation is doubtless due to the beds having been laid down in proximity to the shore line of that time. Beyond the Mississippi River counties, the Renault is known in Union County and from here it outcrops in a continuous belt, except where it is interrupted by faulting, across Illinois to Hardin County, and is also known across the Ohio River in Kentucky. In Union County the formation contains a considerable amount of clastic material in its lower part, perhaps including the flaggy sandstone east of Anna, which has already been mentioned as possibly representing the Aux Vases. Besides this sandstone and some overlying, variegated shales there is nearly or quite 100 feet of limestone referable to the Renault in the Union County section, _ and the limestone continues across the state, but not everywhere with this thickness. In the southeastern part of the state, espe- cially in Hardin County, and also in Crittenden County, Kentucky, 290 STUART WELLER there are some shaly beds at the base of or just beneath the Renault which have been called the Shetlerville formation, from Shetler- ville, Hardin County, Illinois. These beds might perhaps be con- sidered as a member of the Renault rather than as a distinct for- mation, but they are characterized by certain faunal elements that are somewhat different from the overlying Renault. There is some reason to believe that the Shetlerville beds are represented in the lower portion of what has been called Renault in Union County, but further detailed field work is necessary to establish such a conclusion. East of Union County all of the beds of the Renault or Renault-Shetlerville interval are limestones and more or less calcareous shales. In the region of its typical development in Monroe County, Illinois, the Renault exhibits a maximum thickness of about 100 feet, but it varies from this maximum to a minimum of less than 20 feet, and doubtless actually thins out to nothing at all. The exposures of the formation in Ste. Genevieve County, Missouri, vary in thickness from about 46 feet to 75 feet or more, and there may be a maximum thickness of roo feet in the county. In Union County there is 100 feet or more of Renault, but to the east of this county the formation in combination with the Shetlerville, is some- what less than this, varying from 60 to 80 feet in most sections. The Renault formation rests unconformably upon whatever lies beneath it, wherever it has been observed in Illinois and Missouri. In the Mississippi Valley counties it overlaps the Aux Vases sandstone and in many places rests upon the older Ste: Genevieve or even on the St. Louis limestone in places. The sub-Renault unconformity is well indicated by the presence of a basal conglomerate at a number of widely separated localities. The best exhibitions of this con- glomerate are in St. Clair County, Illinois, on a tributary of Hick- man Creek three miles northwest of Millstadt, and in Ste. Géne- vieve County, Missouri, about halfway between the mouth of Saline Creek and St. Marys. In both of these localities the under- lying formation is the Aux Vases sandstone. The conglomerate is constituted of rounded pebbles of chert with an occasional pebble of igneous rock, ranging in size from two inches in diameter to a fraction of an inch. All through the southern counties of Illinois Le ee eS Eee THE CHESTER SERIES IN ILLINOIS 291 the Renault-Shetlerville rests unconformably upon the Ste. Gene- vieve limestone, and this unconformity must represent a time interval not only equivalent to that between the Renault and Aux Vases in Monroe and Randolph counties, but a very much greater time during which the Aux Vases sandstone was deposited and also the time interval preceding the Aux Vases during which the under- lying Ste. Genevieve and St. Louis limestones were solidified and their secondary chert formed, following which the whole of the Ste. Genevieve and a part of the St. Louis limestones were removed by erosion.in some parts of the region. The unconformity repre- sented by all of these events in Mississippian history must be con- sidered as being of great importance in the classification of the Mississippian as a whole. The limestones of the Renault are all more or less fossiliferous wherever they occur, and in some localities faunas of considerable magnitude can be secured. One of the forms which can be found with careful search, wherever good exposures of the Renault are present, is the crinoid Talarocrimus. ‘This crinoid genus is repre- sented by several species whose geographic distribution is some- what different, but the same species is known to occur in localities as far apart as Monroe and Hardin counties. A peculiar feature of the genus is its two basal plates, and nearly all of the Renault species have the suture between the two plates somewhat impressed, giving to the base a distinctly bilobed form. These bases and the separated radial plates are the portions most commonly met with, and from these fragments the species cannot be certainly deter- mined, but these bases alone seem to be sufficiently characteristic to be distinctive of the Lower Chester faunas, and they are much more commonly met with in the Renault than in the Paint Creek, the higher limestone unit of the Lower Chester. Another fossil form which is very characteristic of the Lower Chester beds, is the bryozoan Cystodictya labiosa, which occurs in both the Renault and the Paint Creek, but has nowhere been observed in any higher formation. The Renault fauna can be differentiated from that of the higher Paint Creek limestone, among other ways, by reason of the much less number of Archimedes and Pentremites, representa- tives of both of these genera being very conspicuous in the Paint 292 STUART WELLER ry Creek while Archimedes especially, which is such an abundant form in most of the Chester faunas, is inconspicuous in the Renault in most localities, and in very many collections does not occur at all. The basis for correlating the Renault across the entire state of Illinois, from St. Clair County to Hardin County, is not only the position of the formation in the stratigraphic column, but also the uniformity of the fossil faunas which occur in the formation. Every species which has been recognized in the Shetlerville-Renault faunas of the southern counties, with the exception of four which are wholly restricted, so far as known, to the Shetlerville beds of Pope and Hardin counties, are known to be present in the typical Renault of Monroe County, except one form which is known in the Paint Creek. Furthermore, the especial index fossils of the Ste. Gene- vieve limestone have nowhere been found in association with the Renault-Shetlerville fauna. While it is not possible in this place to enter into a discussion of the details of the faunal characters of the horizon, it can be said that most detailed studies of these Lower Chester faunas seem to establish without any doubt the paleon- tological correlation of the Renault horizon across the entire state. The sandstone beds of the Renault are commonly less massive than those of the Aux Vases, and they not infrequently contain the fossil trunks of a species of Lepidodendron, while no fossils at all have been observed in the Aux Vases. Vankeetown chert and Bethel sandstone.2—Succeeding the Re- nault formation in the Monroe-Randolph County area in Illinois, there is a thin, but very peculiar and persistent bed, which has been called the Yankeetown chert. This formation is siliceous through- out, much of it is a true chert, but in many localities it is seen to include numerous sand grains and locally it is a quartzite. . The bedding of the formation is exceedingly irregular and knotty in many places, but locally at least it is quite even. In many places the rock exhibits a distinct, horizontally banded appearance, the separate bands being slightly different in color and only a small fraction of an inch in width. As ordinarily seen in surface outcrops * Weller, Trans. Ill. Acad. Sci., Vol. VI (1914), p. 124; also Ill. State Geol. Surv., Monog. I (1914), p. 25. 2 Butts, Mississippian Formations of Western Kentucky (1917), Pp. 63 - THE CHESTER SERIES IN ILLINOIS 293 the Yankeetown is rather light colored, and it may be detected in many places by the presence of the fragments of nearly white chert scattered through the surficial deposits. The thickness of the Yankeetown in the Mississippi River counties nowhere exceeds 20 feet, and in places it is perhaps less than to feet thick. In spite of its thinness, however, the Yankee- town is very persistent, and is uniform in its characters from a point in St. Clair County not more than eight or nine miles south of East St. Louis, to near Lithium in the northern part of Perry County, Missouri. Where the Lower Chester formations reappear in Union County, Illinois, the horizon of the Yankeetown is occupied by a sandstone formation quite different in character from the Yankeetown, which has been named the Bethel sandstone by Butts from outcrops in Kentucky. ‘This sandstone holds its position in the Chester sec- tion from Union County to Hardin County, except where the out- cropping belt is interrupted by faulting, although in southern John- son County there is a short interval where the formation is entirely lacking. In the first section in Union County, east of Anna, where the Bethel sandstone has been observed, its thickness is compa- rable to that of the Yankeetown in Monroe and Randolph counties. It is certainly not greater than 20 feet, and perhaps does not exceed 1o feet. Traced to the eastward across the southern counties to the eastern edge of Johnson County, the Bethel nowhere exhibits a thickness greater than 25 or 30 feet, and at one locality at least, in Johnson County, it is lacking altogether. In western Pope County the formation is interrupted by a great, down-dropped » fault block, and where it is exposed to the east of this fault block it is considerably thicker, and continues to increase to the east, attaining a thickness of at least 100 feet in southwestern Hardin County. This sandstone continues southward across the Ohio River into Kentucky, and it is this formation which Ulrich mistakenly con- sidered to be the equivalent of the Cypress sandstone of Engel- mann, an error which he has acknowledged and corrected in his latest contribution to the subject.* t Formations of Chester Series in Western Kentucky (1917), p. 8. RAAT eine STUART WELLER Wherever the contact of the Bethel sandstone with the under- , lying Renault is exposed, there is evidence of unconformity between the two formations. In the Ohio River bluffs in southeastern Hardin County this contact is well exposed, the lower layer of the sandstone, 6 to 18 inches in thickness, is composed of fragmental material consisting of flat pebbles, slabs more or less irregularly disposed, much lime sand, quartz sand of large, rounded grains, with many fragments of fossils, some of which are worn and rounded. The actual line of contact between the two formations is uneven and undulating. At Indian Point, in southern Johnson County, the basal layer of the Bethel sandstone is a lime conglomerate with more or less flattened pebbles up to two or three inches in maxi- mum dimension. The unconformity of the Yankeetown upon the underlying Renault in the Mississippi River counties, is suggested by the varying thickness of the Renault, and by the uniform charac- ter of the Yankeetown, resting in different places upon limestone, shale, and sandstone layers of the Renault. The correlation of the Yankeetown-Bethel horizon entirely across the state must be based upon the correlation of the under- lying and overlying formations, both of which are abundantly fos- siliferous. No determinable fossils have anywhere been collected from the Yankeetown, and the invertebrates that have been found in the Bethel are a few very imperfect examples of common Chester types of brachiopods and bryozoans. This sandstone does contain, in places, numerous fragmentary plant remains, mostly tree trunks, of which the only form that can be identified is Lepzdodendron, ' probably of the same type that was present in the sandstone layers of the Renault, and which is present in most of the Chester sand- stones. Paint Creek limestone.—Overlying the Yankeetown and Bethel formations is the Paint Creek limestone and shale. In the Missis- sippi River counties, extending from St. Clair County, Illinois, to Perry County, Missouri, there is present in the lower part of this formation, a persistent bed of deep-red, non-laminated clay or shale, 12 to 15 feet in thickness. Between this red clay and the 1 Weller, Trans. Ill. Acad. Sci., Vol. VI (1914), p. 125; also Ill. State Geol. Surv., Monog. I (1914), p. 26. THE CHESTER SERIES IN ILLINOIS 295 Yankeetown for a thickness of about 1o feet there is a series of bluish, calcareous shales with platy limestone layers, and above the red bed there are other calcareous shales which pass up into lime- stones, thinly bedded and shaly below, becoming more massive above, these beds being succeeded by more shale beds some of which are variegated red and blue in color. Although there are other reddish or at least variegated shale beds elsewhere in the Chester section, there is no bed anywhere in the series in Illinois that can be mistaken for the deep-red clay bed of the lower part of the Paint Creek formation. Not only is this bed recognizable in surface outcrops, but it can be easily detected in many well — records. The red shale bed of the Paint Creek formation outcrops at intervals throughout the Chester belt from St. Clair to Randolph - counties, the northernmost exposure being about one mile northwest of Millstadt. The formation continues across the Mississippi River into Missouri, and the southernmost exposure is in northern Perry County of that state. Between these two localities the same red shale bed is exposed at many localities. It is exceedingly uniform in its characteristics, and where it is met with it is absolutely impos- sible to mistake it for any other bed jin the Chester series. The limestones of the Paint Creek formation are similar in lithologic character to many other limestones of the Chester series. The several beds are separated by shale layers varying in thickness from an inch or so to several feet, and the limestone beds them- selves vary in thickness from less than one foot to three or four feet. Most of the shale beds are more or less calcareous, but above the main mass of limestone there is a considerable body of shale in many sections that is little or not at all calcareous, and is varie- gated red and blue or purple. Most of the limestone beds are crystalline, some are quite pure and white, others are more impure and much darker in color. In the southern counties of the state, from Union to Hardin, the Paint Creek is represented mostly by shales, with only subor- dinate limestone layers, commonly very thin and exhibiting con- siderable variation in the entire amount that is present. The deep-red shale bed is wanting in the section in these southern 206 STUART WELLER counties, but the shales that are present commonly weather into a red, residual clay which somewhat resembles the material in the red bed of the Mississippi River section. When not weathered, the Paint Creek shale of the southern counties is very fissile, breaking into thin, brittle flakes which are slightly olive green in color when dry, but appear quite black where the exposures are in situations where the rocks are kept constantly wet. In one locality in Johnson County a bed of somewhat variegated red and blue shale has been observed similar in character to some of the beds in Monroe County. The limestone layers included in the Paint Creek formation in ~ the southern counties vary greatly in character. In places some of these layers are very siliceous, some of them being little more than layers of sand firmly cemented with calcium carbonate, other layers are hard, dense, and compact with few or no sand grains, still other beds are quite free from silica, and some of them, at least, are more or less coarsely crystalline, dark limestone, quite like some of the beds in the more typical exposures of the forma- tion in Monroe County. The fauna of the Paint Creek is uniform in its essential features, through the full extent of the formation in Illinois. It has much in common with the faunas of the Renault, and the two formations together constitute the two fossiliferous horizons of the Lower Chester. The bryozoan Cystodictya labiosa is common in both horizons, as are the species. of Talarocrinus with bilobed bases, but the Paint Creek fauna includes a much greater number, both of individuals and species, of Peniremites, and the bryozoan genus Archimedes is far more abundant than in the Renault. The same species of Pentremites are present in the fauna from St. Clair to Hardin counties. The Paint Creek occupies the position in the section which was originally assigned to the Tribune limestone by Ulrich, though it is by no means the equivalent of the formation so named, at Tribune, Kentucky. More recently Buttst has proposed to substitute the name Gasper for Tribune, because of the unfortunate choice of «Butts, Mississippian Formations of Western Kentucky (1917), p. 64. THE CHESTER SERIES IN ILLINOIS 297 that name for the formation by Ulrich. The Paint Creek is the equivalent of the higher portion, at least, of the Gasper limestone of Kentucky. MIDDLE CHESTER GROUP In passing from the Lower to the Middle Chester formations, the region of typical and more complete development is found to be in the more southern counties of Illinois, rather than in the Mississippi River counties. As in the case of the Lower Chester, the Middle Chester is constituted of four formations, two siliceous and two calcareous. Cypress sandstone.—This is the formation for which Engelmann chose the name Cypress sandstone, from the exposures in the blufis of Cypress Creek, Union County, but it is not the sandstone for which Ulrich used the same name in the report on “The Lead Zinc and Fluorspar Deposits of Western Kentucky.’* The for- mation is continuously present in the Chester section from Hardin County at the east to Union County at the western extremity of the southern belt of outcrop of the formations. It is a very mas- sive, clifi-forming sandstone, and except where it is interrupted by faulting in Hardin, Pope, and Johnson counties, it forms the upper portion of a nearly continuous escarpment across the state which is a conspicuous topographical feature. The formation is more uni- form in its character throughout its extent in these counties than any other sandstone formation in the Chester series. Some other sandstones are just as massive and make just as conspicuous cliffs in places, but they do not retain such a character throughout for the reason that the massive portions of the other sandstones are much more interrupted, both vertically and horizontally, by thinly bedded and less resistent layers. The lithologic character of the Cypress is similar to other sand- stones of the series, or at least to certain portions of most of the other sandstones. It is rather fine in texture, yellowish brown in color, with more or less cross-bedding, although certain portions of the formation are conspicuously even-bedded, and in places, especially in the upper portion of the formation, the even beds t Prof. Paper, U.S. Geol. Surv., No. 36 (1905). 208 STUART WELLER suggest the regular courses in a well-built masonry wall. The weathered surfaces of the cliffs become darker colored than the freshly broken rock, and in places more or less iron stained. The fos- sils of the Cypress sandstone consist of more or less fragmentary plant remains, the only recognizable form being Lepidodendron trunks. | - It has not been possible to measure the exact thickness of the Cypress sandstone in any section in the southern counties of the state. The base of the formation, resting upon the Paint Creek shale, can be approximately determined in many places, but the top of the section in these same sections is in all cases missing and the upper portion has been more or less reduced by weathering. The greatest actual thickness that has been observed in a cliff is about 70 feet, but the thickness has been estimated as t10 feet in at least one section, and the average thickness across these counties is about too feet. In tracing the stratigraphic position of the Cypress sandstone into the section of the Chester series of the Mississippi River counties, the sandstone is found to be much reduced in thickness and much less massive in character. In this section, as originally described by the writer,’ a sand and shale formation overlying the Paint Creek limestone was named the Ruma formation. The later study of the section in the more southern counties has shown that the sandstone of the Ruma should be considered as the thinned- out margin of the Cypress sandstone, and that the shales below should more properly be considered as being a part of the Paint Creek. With this interpretation the name Ruma becomes super- fluous, and Cypress may be extended to include these sandstone beds of the Ruma in Monroe and Randolph counties. In follow- ing the section still farther, into Missouri, it is found that the Cypress sandstone disappears entirely, and the super-Cypress limestones rest directly upon the Paint Creek. In a recent contribution Ulrich? has proposed the correlation of the Cypress sandstone of the southern counties with the Lower 1 Weller, Trans. Ill. Acad. Sci., Vol. VI (1914), p. 126; also Ill. State Geol. Surv., Monog. I (1914), p. 26. 2‘“The Formations of the Chester Series in Western Kentucky, and Their Corre- lates Elsewhere,” Ky. Geol. Surv., Plate D, opposite p. 47. THE CHESTER SERIES IN ILLINOIS 299 Okaw limestone of Monroe and Randolph counties. Such a correlation, however, is not supported by the evidence of the fossils and the faunal studies of the Chester have established beyond question the exact equivalence of the Golconda limestone of the southern counties with the Lower Okaw in Randolph County. Golconda limestone-—When the Chester section in the Missis- sippi River counties was first elaborated by the writer, the name Okaw limestone was given to a thick series of limestones with shale partings overlying the so-called “‘Ruma”’ formation. It was recog- nized that this was probably a composite formation, and an attempt was made to map the higher beds as a separate unit from the lower ones, but this was finally abandoned because the heavy covering of drift seemed to make such a procedure impracticable. When the _ studies were carried into the more southern counties, it developed that the limestone beds equivalent to the Okaw were divided into two distinct units separated by an important sandstone formation. The lower of these two units has been named Golconda limestone from the excellent exposures in the Ohio River bluffs just above Golconda, in Pope County. The Golconda limestone is constituted of a succession of lime- stone and shale beds, the details of which are commonly obscured _by surficial material, and it is not known whether the details of the succession of beds are uniform throughout the areal extent of the formation. ‘The limestone beds vary considerably in character, but in general they are of a light- or dark-gray color, and more or less crystalline in texture, with some layers odlitic. The shales are fully as variable and perhaps more variable than the limestones. Some of them are highly calcareous, while others are quite purely argillaceous; many of the beds are gray or buff, but others are dark and even black, and at a number of localities a layer of reddish shale has been observed. In the basal part of the forma- tion there are shale beds with a considerable content of sand, and even some thin sandstone layers, but beds of this character are not present higher up in the formation. In tracing the Golconda limestone into the Mississippi River counties, where it is represented by the lower and main portion of 300 STUART WELLER the Okaw limestone, the characteristics of the formation remain much the same, although the local details are different. As in the southern counties there is a succession of limestone and shale members, but there is a larger content of limestone in the more western region. ‘The limestone beds themselves are crystalline in texture, like those in the south, they vary in color from essentially white to dark gray, the lighter colors on the whole being more dominant in Monroe and Randolph counties, and the odlitic beds being much more conspicuous. ‘The shale beds are similar in the two regions. The establishment of the continuity and equivalence of the Golconda and the lower portion of the Okaw is based not alone upon their occupying an equivalent position in the section, but upon the paleontological characters as well. One of the notable horizon markers of this lowest limestone formation of the Middle Chester is the little brachiopod Camarophoria explanata. This species 1s unknown in the Lower Chester faunas, but is a common member of all the Middle Chester faunas, and is present, abun- dantly in places, in some of the Upper Chester formations. ‘The horizon where it is first introduced in the section can be considered as being well toward the base of the Golconda limestone. In the southeastern counties of the state one of the most reliable guide fossils for the lower Golconda is the Crinoid Pterotocrinus capitals, which is commonly represented by the “‘wing-plates” alone. This species has not been recognized in Randolph or the adjoining counties, but in this region the near basal beds of the lower Okaw are characterized by the presence of a peculiar and very unusual fauna, for the Chester series at least, composed very largely of small pelecypods and gastropods, including many Bellerophontids. Many of the species of this fauna are undescribed, and some of them are peculiar and extraordinary. Insouthern Johnson County, at one locality, a fauna has been collected from near the base of the — Golconda, in which most of these peculiar basal Okaw species are present, and associated with them are many examples of the char- acteristic Pterotocrinus capitalis. This mingling of forms, so pecu- liar in character, is assumed to be sufficient evidence to establish the equivalent of the Golconda with the lower Okaw, and the THE CHESTER SERIES IN ILLINOIS 301 name Golconda may be extended to include the equivalent beds in Randolph and Monroe counties. The lithologic character of the Golconda limestone is of such a nature that its contacts with the underlying and overlying forma- tions are not commonly exhibited, and at no locality have both of the contacts been observed in the same section. This condition makes the determination of the thickness of the formation a matter ofestimate. In the neighborhood of Golconda the interval between the top of the Cypress sandstone and the base of the Hardinsburg is about 150 feet, and as this is the interval occupied by the Gol- conda, an approximate thickness of 1s0 feet may be assumed for the formation. The thickness of the whole of the Okaw limestone in the Mississippi River counties is something over 200 feet, and of this total thickness the lower Okaw, which is the equivalent of the Golconda, includes approximately 150 feet, being about equal to the Golconda in its typical exposures. Hardinsburg sandstone.—Overlying the Golconda limestone and resting upon it unconformably is an important sandstone forma- tion which in many places is scarcely less massive than the Cypress. Butts has given the name Hardinsburg" to this formation from a Kentucky locality. In general appearance, texture, color, etc., the Hardinsburg closely resembles the Cypress, and in isolated out- crops not seen in relation to an underlying or overlying limestone, it would not be possible in many places to differentiate the two formations. The Hardinsburg, however, is somewhat less massive on the whole, and includes considerable amounts of more thinly bedded sandstones in places. In general the Hardinsburg is some- what thinner than the Cypress, although it does have a maximum thickness of at least 100 feet. There are places, however, in the southern counties where the thickness does not exceed 30 feet, and the average thickness is probably about 60 or 70 feet. In the Mississippi River counties there is no conspicuous sand- stone formation which corresponds in position with the Hardins- burg in the southern counties. There is present, however, within the formation to which the name Okaw was originally given, a horizon marked by a discontinuous sandstone layer which in * Miss. Form. W. Ky. (1917), p. 96. 302 STUART WELLER places is as much as Io feet thick, elsewhere being wanting alto- gether. This layer is best exhibited in the vicinity of Chester, in the outside prison quarry at Menard, between Menard and Ches- ter, and just below Cole’s mill in Chester. This sandy layer in the Okaw is undoubtedly the attenuated margin of the Hardinsburg sandstone which has its greatest thickness in the southeastern part of the state, for the limestone beds above it possess many faunal characters which unite them with the limestone formation overly- ing the Hardinsburg in the southern counties. Glen Dean limestone-—The Glen Dean limestone is another formation that has been named by Butts from exposures in Ken- tucky.t In the southern counties of Illinois the formation resembles the Golconda in general character, being composed of interbedded limestone and shale layers, but in most localities the proportional amount of shale is much greater in the Glen Dean, in places nearly the whole of the formation being shale. Many of the limestone beds in the formation are similar lithologically to those of the Golconda, being gray in color and crystalline in texture for the most part, but locally certain of the layers are somewhat more dense and compact. ; The Glen Dean has certain faunal characters that differentiate - it rather sharply from the Golconda. One of the best index fossils is a species of bryozoan, Prismopora serrulata. Examples of this species are triangular in cross-section, with three faces bearing zooecia, these prismatic zoaria dividing at intervals. This bryo- zoan is not entirely confined to the Glen Dean, for it has been observed rarely in the Golconda, and is not uncommon in the Vienna limestone, still higher than the Glen Dean, but it is far more common in the Glen Dean than elsewhere, and in places some of the limestone ledges of this formation are veritable Pris- mopora gardens. Pentremites spicatus is another characteristic form, which has not been observed outside of this formation, but it is far less common than the Prismopora. A number of other bryozoans and some other fossil forms are more or less conspicuous in this formation, which are nearly everywhere or entirely un- known from other Chester horizons. t Miss. Form. W. Ky. (1917), p. 97. THE CHESTER SERIES IN ILLINOIS 303 ‘In the Mississippi River counties the whole assemblage of fossil forms which characterize the Glen Dean formation in the southern - counties has been found to be present in those beds of the Okaw limestone which overlie the interrupted sandstone. horizon in the midst of that formation, and these upper Okaw beds may be corre- lated directly with the Glen Dean and this name may be extended to include these beds in the Randolph-Monroe County section. The thickness of the Glen Dean in the southern counties exhibits some variation from a minimum of 4o feet to a maximum of perhaps 75 feet. In the thinner sections it is apparently the higher beds that are missing, due perhaps, to the erosion of the upper surface of the formation before the deposition of the over- lying sandstone. The thickness of the equivalent beds in Randolph County is similar to that in the southern counties, the usual thick- ness commonly being about 60 feet. [To be continued | A CORRELATION OF THE PRE-CAMBRIAN FORMA- TIONS OF NORTHERN ONTARIO AND QUEBEC: H. C. COOKE Geological Survey of Canada, Ottawa In a paper recently published in the Journal of Geology? the writer correlated the pre-Cambrian formations of northern Quebec as well as facts obtained during the last ten years would permit. The formations most satisfactorily correlated were a number of scattered patches of sediments which had been given the local names Pontiac, Mattagami, Broadback, and Brock series. In composition, succession, structure, and external relations these scattered patches of sediments were deemed sufficiently alike to warrant substituting the general name Mattagami series for the local names. The position of the Mattagami series in the geologic — column is shown in the following succession: Mistassini limestone Mattagami series Unconformity Unconformity Diabase dikes Gabbro and anorthosite Intrusive contact Intrusive contact Cobalt series Granite-gneiss (around Lake Great unconformity St. John) Granite Intrusive contact Intrusive contact Nemenjish series (Grenville series ?) Lamprophyre dikes Abitibi volcanics (basalts, Intrusive contact andesites, rhyolites) During 1918 the writer had an opportunity to examine for the first time the Timiskaming series in the Kirkland Lake district of northern Ontario, and was so impressed by its similarity to the Mattagami series that further opportunity was taken in 1919 to investigate its relation to that series. The results indicate beyondany t Published by permission of the Geological Survey of Canada. 2 Journal of Geology, Vol. XX VII, Nos. 2, 3, 4, 5, 1910. 304 PRE-CAMBRIAN OF NORTHERN ONTARIO AND QUEBEC 305 reasonable doubt that the Mattagami series and the Timiskaming series of Kirkland Lake may be correlated and grouped together under a single name. LOCATION The area of Mattagami series nearest to the Kirkland Lake district is the Pontiac area lying south of the National Trans- continental Railway between the Ontario-Quebec boundary and the J Mes Fastmain fp Rup pert is Moose factor, ie rans Feary aR A io Cochrs Oe Wee nee ge Cobsa/t i Ly Timiskaming fed > Ke \ % S Scale of miles Fic. 1.—Index map showing location (hatched) of area dealt with in this paper Bell River. Between the two districts lies the Larder Lake dis- trict, 12 milesin width. The three districts together form a narrow strip of territory about 130 miles from east to west, and 25 miles from north to south (Fig. 1). The shaded area in Figure 1 shows 306 H.C: COOKE the general location of this strip within which lie the sediments whose correlation is dealt with in this paper. The Timiskaming Northern Ontario Railway passes through its western end, affording access to the important mining center of Kirkland Lake through Swastika station, and to Larder Lake through Dane sta- tion. The only access to the Quebec portion of the area is by canoe from Lake Timiskaming, Larder Lake, or the National Transcontinental Railway. | PREVIOUS WORK AND CONCLUSIONS Reconnaissances along easily accessible water routes were made as early as tgo0 by W. J. Wilson,’ W. G. Miller,’ L. L. Bolton,’ Walter McOuat,4 W. A. Parks,’ J. Obalski,° and J. F. E. Johnston,’ but the first detailed work appears to have been that of Brock® and Bowen,’ in 1908, at Larder Lake. In the four years following the whole of the area shown in Figure 1 was mapped in more or less detail by A. G. Burrows, and P. E. Hopkins,” E. L. Bruce,* M. E. Wilson,” R. Harvie, and J. A. Bancroft. These writers have reached very different conclusions regarding the age and external relations of the sediments in this strip of ter- ritory. Burrows and Hopkins (1912) consider the sediments of Teck, Lebel, and Gauthier townships to be a single series of inter- ~W. J. Wilson, Geol. Surv. Can., Sum. Rept., 1901, pp. 117A-130A; Geol. Surv. Can., Mem. No. 4, 19t0. 2W. G. Miller, Ont. Bur. of Mines, Rept. No. 14, pp. 261-68, 1905; Rept. No. 11, PP. 214-30, 1902. 3L. L. Bolton, Ont. Bur. of Mines, Rept. No. 12, pp. 173-90, 1903. 4W. McOuat, “Rept. of Prog.,’”’ Geol. Surv. Can., 1872, 1873, pp. 112-35. 5 W. A. Parks, Geol. Surv. Can., Sum. Rept., 1904, pp. 198-225. 6 J. Obalski, Mining Operations in the Province of Quebec, 1906, pp. 5-27; 19097, Pp. 42-56. \ 7J. F. E. Jolinston, Geol. Surv. Can., Sum. Rept., 1901, pp. 1330A-143A. 8 R. W. Brock, Ont. Bur. of Mines Rept. for 1907. 9N. L. Bowen, Ont. Bur. of Mines Rept. for 1908. t0 Burrows and Hopkins, Ont. Bur. of Mines Rept. for 1913. 1. L. Bruce, Ont. Bur. of Mines, Rept. for 1912. 12M. E. Wilson, Geol. Surv. Can., Mem. Nos. 17, 39. 3 R. Harvie, Report of Mining Operations in the Province of Quebec during rgro. 4 J. A. Bancroft, Report of Mining Operations in the Province of Quebec during 1912. PRE-CAMBRIAN OF NORTHERN ONTARIO. AND'QUEBEC 307 bedded conglomerates and greywackes folded into a tight syncline with vertical or very steep limbs. Miller correlates this series with a similar closely folded conglomerate near Cobalt, previously - called Timiskaming series. They recognize that it is much older than the Cobalt series, since the latter is found in Grenfell Town- ship, to the west of Teck, lying flat within a quarter of a mile of the folded Timiskaming. The Cobalt series is regarded as probably Upper Huronian or Animikean in age, and the Timiskaming as Lower Huronian.* The sediments continue without interruption (Fig. 3) from Gauthier Township eastward into McVittie and McGarry town- ships (Larder Lake district), where they are described by Brock and Bowen (1907) and later by M. E. Wilson (1908-9) as con- - sisting of slates, carbonate rocks, and greywackes interbedded with basic altered lavas and therefore as of Keewatin age. Some steeply dipping conglomerates found to the north and south of the slates and other rocks are mapped as isolated patches of Cobalt conglomerate deformed by local disturbance. Some four miles to the east of Larder Lake there outcrops the sedimentary series named the Pontiac schists by M. E. Wilson. These are separated from the Larder Lake sediments by a band of flat-lying sediments belonging to the Cobalt series (Fig. 3). The Pontiac series has been described by M. E. Wilson and J. A. Ban- croft as a series of interbedded conglomerate and greywacke, folded into a vertical position, overlying the Keewatin unconform- ably, and underlying the Cobalt series with great unconformity. The similarity of the descriptions of the Pontiac conglomerates and greywackes and the corresponding members of the Timiskam- ing series of Kirkland Lake is pronounced. The geological descriptions which have been briefly summarized appeared to the writer to indicate either that some error had been made in the study of one of the areas, or that there was more there than any of the geologists had recognized. Under the latter hypothesis it was conceived that there might be two ancient sedimentary series in the district, one of which was the better developed in each area, and that a different one had been t Ont. Bur. Mines, Rept. No. 22, pp. 123-27, 1913. 308 H. C. COOKE recognized by each set of investigators. ‘To determine if possible which of these hypotheses was the correct one, the writer In 1919 visited Larder Lake and made a study of the part of the area in doubt. INVESTIGATIONS AT LARDER LAKE (FIG. 2) A conglomerate outcropping prominently on the Larder town- site was the first thing to attract the writer’s attention. This conglomerate is mapped by Brock and Wilson as belonging to the . rl Wea a ea 5 Aad) } Scale of Miles Beaverkause CS ake & <0) True North La IG AUTHIER S Oe Tiff Yi YY UY ult Ho ST ° ° 6oe ie ° 4 wh Ps R S08. oy 6 el ioane & : : es Larder darris Maxwell (7s eas Cobale Timishamin Vilage ans Mine . SEMIRDINE lac o| SEF/ES Zags me, eee "8 LAP OW OCF feldspar "Keewatin, Lake eu "e| porphyry Gasalts and) L CRN Ahyolites Fic. 2.—Larder Lake area Cobalt series, but evidently with reservation on the latter’s part, as he describes it thus:* There is an area of mashed conglomerate on the north shore of Larder Lake, at Larder City, which has been intruded in a most complex manner by a hornblende lamprophyre, vogesite. The pebbles of this conglomerate differ from the normal type in that they consist entirely of quartz porphyry, rhyolite, and iron formation. .... The occurrence of the lamprophyre cutting the sheared conglomerate suggests that this conglomerate may be the equivalent of Dr. Miller’s Timiskaming series. Again, on page 37: In the neighbourhood of Larder Lake there are numerous areas of con- glomerate which have been greatly mashed in a direction parallel to the strike of the underlying Keewatin. These conglomerates might have any one of the three following relationships to the other rocks of the region: (1) They might be Keewatin conglomerates deposited between the volcanic flows of 1 Geol. Surv. Can., Mem. No. 17, p. 38. PRE-CAMBRIAN OF NORTHERN ONTARIO AND QUEBEC 309 that series in a manner somewhat similar to the interflow conglomerates which occur in the lower portion of the Keewanawan series. (2) They might belong to an older Huronian series, that is, a series younger than the Keewatin but older than the wndisturbed Huronian. (3) They might be portions of the ordinary flat-lying series which have suffered local disturbance. Wilson concluded to map all the mashed conglomerates under the third hypothesis, as locally sheared Cobalt series. The only evidence in support is given on page 38: It was also observed that in some outcrops where the mashed conglomerate has a considerable vertical thickness, the schistosity appears to diminish from the base upward, as if the contact of the flat-lying Huronian and the Keewatin might have served as a plane of deformation. This phenomenon can be seen in a conglomerate hill situated on the northern boundary of claim H.J.B. 21, in McGarry Township. A plane of contact very commonly does function as a gliding plane during folding, so that the rocks close to it are apt to be more schistose than those farther away without regard to the age of the series; but the foregoing inference would be valid only if the schistose strata passed gradually into flat-lying beds, which they do not. The writer found that this conglomerate forms part of the conglomerate of the Timiskaming series. To the east of Larder Village, across a small bay, Gold King Point projects into the lake. The greater part of this point is mapped as Keewatin on the early maps, with a small patch of conglomerate near the southeastern extremity; but the point was stripped clean of vegetation by fire three years ago, and the present exposures show clearly that the conglomerate is not a patch but a continuous band about 30 feet in width, dipping vertically and with its strike swinging from north 60 degrees west at the southeastern tip of the point to north near the Harris-Maxwell mine (Fig. 2). The conglomerate, like that on the Larder townsite, contains pebbles of basalt, rhyolite, jasper and iron formation. It is intruded by a feldspar porphyry like that of Kirkland Lake, and by dikes of lamprophyre. On the east side, where not intruded by porphyry or lamprophyre, it is in contact with massive basalt, exhibiting good pillow structure, and contains basalt pebbles. The rocks on the west of the conglomerate are not Keewatin, but well-bedded greywackes, showing fine lines of cross-bedding in 210 H. C. COOKE places; the dip and strike of the greywacke parallel those of the conglomerate. , It is evident therefore that these rocks, with those on the Larder townsite, are parts of a small tightly folded syncline, the axis of which has a north-south strike between the village and the Harris-Maxwell mine. It was also apparent that they lie un- conformably on the greenstones, as the rhyolite, basalt, jasper, and iron formation are all members of the underlying volcanic complex. 4 A mile and a half to the east, on Pearl Point (Fig. 2) normal 7 conglomerate and slate of the Cobalt series outcrops. The con- glomerate is massive, at least too feet thick, and crowded with pebbles of all sizes. From 75 to 90 per cent of the pebbles are granite, the remainder other rocks. The conglomerate dips 15 degrees to 20 degrees, and is overlain by the normal argillite of the series, fine grained, black, and well bedded, containing an occasional pebble. The composition, succession, and structure of these rocks is so utterly different from those at Larder Village only 14 miles’ away that it is difficult to conceive them to be of the same formation. Accordingly the writer returned to trace northward the band of conglomerate that passes across the Larder townsite. It runs somewhat east of north for about half a mile, then swings to the east across a drift-filled valley, on the other side of which it was easily picked up again and traced a short distance farther north, till, turning west, it passes beneath a large sand plain. Two miles to the west it has been mapped by both Wilson and Burrows on the shore of the Blanche River. About half a mile to the north of the Harris-Maxweil mine (Fig. 2) an interesting set of relations occurs. The basal band of conglomerate, composed as before mainly of pebbles of rhyolite, jasper; and iron formation, here striking south 4o degrees east and with vertical dips, is overlain conformably on the east by inter- banded greywacke and conglomerate, the conglomerates gradually becoming finer grained, passing into coarse grits and then into fine grits. Above the grits and greywackes occur some hundreds of feet of soft, slaty argillites, whose strike and dip parallel those of —- PRE-CAMBRIAN OF NORTHERN ONTARIO AND QUEBEC 311 - the conglomerate. All these rocks are greatly broken up by in- trusive dikes and masses of porphyry. Lying on the upturned edges of these rocks is a’second con- glomerate, a massive rock composed almost entirely of greenstone fragments with about ro per cent of granite pebbles, lying flat, and showing vague cross-bedding in places. This conglomerate occurs in irregular patches and knobs, evidently erosion remnants, lying indifferently on the older slates, greywackes, grits, and conglom- erates. The writer concluded from the composition, which is characteristic of the base of the Cobalt conglomerate, and from the lack of deformation that it is an erosion remnant of the Cobalt conglomerate found to the east in larger masses; the areal relations to the underlying sedimentary series show that there is a large unconformity between the two. The earlier geologic maps show a number of patches of con- glomerate, mapped as Cobalt series, lying to the north of the areas of Larder slate, in many cases some distance to the north and within areas of Keewatin. These conglomerates proved on examination to consist invariably of beds dipping to the south at angles varying from 60 to go, and frequently badly sheared. Their composition is like that of the conglomerate of the Larder townsite, in that the pebbles are largely of rhyolite, banded chert, jasper, and iron formation, with some basalt, although commonly the proportron of rhyolite is larger and that of iron formation smaller than on the townsite. The rocks to the north of the conglomerate band are invariably basalts except around Malone Lake, where rhyolite occurs; and almost invariably they exhibit pillow struc- tures and other characteristics of lavas. ‘The conglomerates were found to form, not isolated patches, but a strong band, continuous throughout the district, except where broken by intrusive masses of porphyry, and where, for a short distance to the east of Barber Lake, it has been obliterated by thinning and intense shearing. The rocks to the south of the conglomerate band were then carefully examined, as some of them had been previously mapped as Keewatin. in ~ 181 ean <= MOnthiye Ww catler Chali. sam) 7 _ THE UNIVERSITY OF CHICAGO PRESS, CHICAGO, ILLINOIS line POSITIONS OF ALL KINDS Never was the demand so great for qualified teachers and specialists. For ten years we have given our time and energy to this work. Write for our free literature. State qualifications briefly. Co-operative Instructors’ Association Marion - Indiana WATER REPTILES OF THE PAST AND PRESENT By SAMUEL WENDELL WILLISTON Late Professor of Paleontology in the University of Chicago Professor Williston, who is widely known as a student of extinct reptiles and as the author of American Permian Vertebrates, which has now become a standard work, presents in this new volume a summary, divested as far as possible of unnecessary scientific details, of our present knowledge concerning the reptiles of the seas, lakes, and rivers of past and present times. The numerous illustrations, in large part from the pen or brush of the author himself, include not only living types and twenty-four restorations of extinct forms, but also many figures elucidat- ing the structures and habits of the animals. es viti+252 pages, royal 8vo, cloth; $3.00, postpaid $3.20 ee The University of Chicago Press Chicago - - - - Illinois PUBLISHED PRICES Will be paid for The Journal of Geology We will pay the published prices for any numbers of the Journal of Geology. Vol. I, No. 1, to Vol. EX, No. 6, if sent to us in salable condition. Write clearly the name and ad- dress to which you want check mailed. The University of Chicago Press 5750-58 Ellis Avenue Chicago, Illinois The Geography of the Ozark Highland of Missouri By CARL ORTWIN SAUER, the University of Michigan (Publications of the Geographic Society of Chicago) The Ozark Highland of Missouri was selected for investigation because of its unusual wealth of geographic responses and because little is known concerning its con- ditions and possibilities. The purpose of such a study is twofold: to furnish an adequate explanation of the conditions of life in a given area, and to contribute proved statements which will aid in working out fundamental principles. The book is divided into three parts. The first is an outline of the environment. Only those things which are pertinent to an understanding of the conditions under which the people live are introduced. The second part considers the influence of environment on the settlement and develop- ment of the different parts of the highland. The third part is a study of the economic conditions as they exist today. In conclusion a forecast is offered of the lines along which the future of the region will be worked out. A valuable feature of the volume is the 44 figures in the text and 26 plates. xviti-+246 pages, 810, cloth; $3.00, postpaid, $3.20 The University of Chicago Press Chicago, Illinois VOLUME XXVIII NUMBER 6 THE SewNNAL OF GEOLOGY SEPTEMBER-OCTOBER 1920 DIASTROPHISM AND THE FORMATIVE PROCESSES XII. THE PHYSICAL PHASES OF THE PLANETARY NUCLEI DURING THEIR FORMATIVE STAGES T. C. CHAMBERLIN The University of Chicago In an article in the first number of this volume of the Journal a study of the relative densities of the moon, Mars, Venus, and the earth brought out the fact that the mean densities of these bodies not only rise in the order of their masses, but that the rate of increase itself rises with each unit-increase of mass.* This led to a study of the processes by which these bodies acquired their material, to see whether any part of the observed higher density could reasonably be assigned to greater proportions of inherently heavy material received during their formation. The conclusion was reached that larger proportions of /ight material entered into the formation of the large bodies than into the small bodies. The natural inference from this is that the higher mean densities now found in the larger bodies are due to some form of mass-effect that was sufficient to “The Order of Magnitude of the Shrinkage of the Earth deduced from Mars, Venus, and the Moon,” Jour. Geol., Vol. XXVIII (1920), pp. 1-17. Compare this with the theoretical deductions of Dr. A. C. Lunn, ‘‘ Geophysical Theory under the Planetesimal Hypothesis, in the Tidal and Other Problems,” Carnegie Institution Publication No. 107 (1909), particularly pp. 188 and 201. Compare also the very suggestive paper of Dr. Wm. D. MacMillan “On Stellar Evolution,” Astrophys. Jour., Vol. XLVIII (July, 1918), pp. 36-40. 473 474 T. C. CHAMBERLIN overbalance their higher content of light original material.t It was recognized, however, that the inquiry could not be regarded as having covered all the shrinkage phases of this mass-effect until the physical states of the four bodies during their formation were considered. Under the planetesimal hypothesis their formation embraced two phases: (1) the progressive concentration of those portions of the solar outbursts that were held together by their mutual attractions so as to act as unit assemblages and thus serve as collecting centers or nuclei; (2) the gathering into these nuclei of such scattered parts of the solar outbursts as were dispersed into planetesimal orbits from which they could be picked up only individually. The first process started with a gaseous body and followed the gaseous line of descent; the second was concerned with individual bodies and orbital dynamics. ‘This paper is confined to the first of these. . THE SUCCESSIVE PHYSICAL PHASES ASSUMED BY THE NUCLEI IN PASSING FROM THEIR ORIGINAL CONDITION TO THEIR FINAL STATES AS PLANETARY CORES There is a wide range between the largest planet and the smallest planetoid. ‘There may also be wide differences of view respecting the probable sizes of the nuclei. As I wish to leave the question of the nuclear masses freely open for the present, it seems best to treat broadly the whole group of solar dependents, including planets, planetoids, and satellites. There is the additional reason that their gradations, their likenesses, and differences, as well as the contrasts of their extreme developments, form the natural background for the special cases with wiles we are particularly concerned in this discussion. It is assumed that each one of the present planets, planetoids, and satellites started from a nucleus inherited from a solar out- burst. It is held that some of these nuclei were formed from the central portions of the solar outbursts, while others were merely segments detached from these. Some of the detached segments are supposed to have remained under the control of the central portions and become satellites, while others pursued orbits of their x “Selective Segregation of the Earth and Its Neighbors,” Jour. Geol., Vol. XXVIII (1920), pp. 126-57. DIASTROPHISM AND THE FORMATIVE PROCESSES 475 own and became planetoids. These are regarded as natural results of solar eruptions under exceptional stimulus from a passing body.! As the assigned result of this there arose a very significant series of solar attendants of cognate birth and linked together by gradations, though the extremes were quite highly differentiated. The series, as now presented, ranges from massive hot gaseous planets of low densities, down through intermediary forms, to quite small solid bodies of high densities and altogether devoid of appreciable atmospheres. In mass value the largest planet is several million times the smallest planetoid—probably we could say several billion times, if the lower limit of the planetoids were accurately determined. In Table I this great series is listed in the order of size, neglecting the common distinction between planets, planetoids, and satellites, which is immaterial in this particular study. The physical differences are brought out by groupings. It will be seen that the planets, planetoids, and satellites are notably mixed in the lower part of the list. The gradation would doubtless be much closer and the classes even more intermixed, if the sizes of all the smaller bodies were well enough determined to permit a strictly accurate arrangement of the smaller masses. There are now known to be 26 satellites and upward of 800 planet- oids, most of which seem to be less than 1oo miles in diameter. While in general there is a notable gradation, there is yet a wide gap between the giant group of gaseous planets and the terrestrial group next below, which are essentially solid but have gaseous envelopes. Within the latter group a somewhat notable difference in mass sets off the earth and Venus from Mars. The scant atmos- phere of the last allies it to the atmosphereless group below and its nucleus not unlikely belonged to that class. The differences in the groups suggest that the formative pro- cesses, though of the same type and initiated in the same way, entered in such different proportions into the actual formative work that they gave rise to very divergent results. This tallies with our earlier suggestions that the formative agencies embraced within t See the special cases of May 29 and July 15, 1919, outlined in this Journal, Vol. XXXVIII (February—March, 1920), pp. 145-49, or the original description by Pettit, in Astrophys. Jour., Vol. L (October, 1919), pp. 206-19. 470 T. C. CHAMBERLIN themselves opposing factors (VII of previous article)! so poised as to permit a ready shifting of dominance from one side of the TABLE I THE SOLAR DEPENDENTS GRADED BY SIZE AND GROUPED BY PHYSICAL PROPERTIES A. Tue Grant Group. HicHiy GASEOUS BopiEs (Densities low; diameters between 30,000 and 90,000 miles) Densities Disme pice ee 1.25 88,392 Satur se eee 0.63 74,163 Neptune....... 1.09 34,823 Wranuseeeeeeece 1.44 30,193 B. THE MEDIAL oR TERRESTRIAL GROUP. Sotm BopiEs BEARING ATMOSPHERES (Densities high; diameters between 4,000 and 8,000 miles) Densities Diometer: IBEW, couoccccc 5353 7,918 WENUS.cocco0ane 4.85 7,701 Miatsircerae ers eens 3.58 4,339 C. THe DiurtnutivE Group. ATMOSPHERELESS Sortm BopirEs (Densities high; diameters ranging from 3,600 miles down to the lower limit of estimating power) Bones) |OEariee Satellite. .| Jupiter’s III 3,558 Satellite. .| Jupiter’s IV 3,345 Planet. ..| Mercury 3,009 Satellite. .| Saturn’s VI 3,000 Tue DIMINUTIVE Group—Continued Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Satellite. . Planetoid Planetoid Planetoid Satellite. . Planetoid Planetoid Satellite. . Planetoids Satellite. . Satellite. . Satellite. . Satellite. . Bodies Jupiter’s I The Moon Jupiter’s IT Saturn’s VIII Neptune’s VIII Saturn’s V Saturn’s IIT Saturn’s IV Uranus’s I-IV Saturn’s II Saturn’s I Saturn’s VII Ceres Pallas Vesta Saturn’s 1X Juno Several exceeding Jupiter’s I Probably 7oo or more, ranging downward from Mars’s II Mars’s I Jupiter’s VI and VII Jupiter’s VIII and IX Diameters in Miles 2,452 2,100 2,095 2,000 2,000 I,500 1,200 1,100 500 to I,000 800 600 500 485 304 243 200 118 100 100 100 Io “small” (73 very small” Planetoids. Thesmallest order of planet- oids probably form the lower end of the series, ranging down to 1o or per- haps even 5 miles. balance to the other, thus giving rise to a series of widely varying effects which at the extremes even became contrasted. The pre- ponderance in the upper end of the series lay markedly on the 1 “Selective Segregation of Material in the Formation of the Earth and Its Neighbors,” Jour. Geol., Vol. XXVIII (1920), pp. 126-57. DIASTROPHISM AND THE FORMATIVE PROCESSES 477 side of gas accumulation, while in the lower part the dominant effect lay in the dissipation of all gas. In the middle ranges there was a closer approach to equipoise between these extremes and hence to a mixed product of the medial order. It is thus clear that the genetic processes, however alike basally, were capable of giving such different results as to make it necessary to study with care and patience the balancings between opposing influences and the differential effects of the shifting of these balances. THE CRITICAL CONDITIONS THAT CONTROLLED THE PASSAGE OF THE NUCLEI INTO COLLECTING CORES The original diversity of the nuclei is assigned to differences in the impulses imparted by the solar eruptions. The evolution of the nuclei, after being launched on their several careers, was critically dependent on the dynamic properties which they inherited individually. These now require attention. We need not dwell, however, on the giant gaseous planets, for they do not fall within the range of our present problem, nor do they seem to have ever passed through the more critical phases of the processes we are about to consider. They probably had, at the outset, nuclei massive enough to hold essentially all their own gases in spite of their molecular activity and to retain essentially all alien molecules that plunged into them.* To cover the whole field of the known solid bodies in a repre- sentative way, Table II is introduced. It gives certain essential dynamic properties for ten typical bodies, four natural and six ideal, so selected as to represent at convenient intervals the whole range from the earth—the largest known solid body—down to a ten-mile planetoid. Dr. W. D. MacMillan has been kind enough to make the computations for this table. It seems improbable that the nuclei of the earth, Venus, Mars, or the moon, even at their smallest stages, were so diminutive as the lower orders of ideal bodies in Table II, but these very small bodies are even more serviceable than larger ones in illustrating the critical conditions that attended their formation and measurably that of «The inevitable loss of such molecules as attained very exceptional speeds is neglected throughout this discussion. 478 T. C. CHAMBERLIN the larger bodies of the solid order. They thus serve to put to severe test our notions as to the formation of such bodies. It will not be surprising if we find that these small bodies lie on the precarious border that separates successful aggregation from dissipation into planetesimals. TABLE II DYNAMICAL PROPERTIES OF TEN REPRESENTATIVE BODIES OF THE TERRESTRIAL AND SMALLER CLASSES Tue Ten Bopies STATISTICAL PROPERTIES DYNAMICAL PROPERTIES a 6b C d e cf g h t 5 Paraboli Rete = & : Surface aTaDOrc | Ol Ncten= || Diameter i | Density Mass Graei Velocity | _ tion in Ss No. Name £2 \WWiaioe=an Earth=1 pee _ ils Miles Pee a aphee Fane: Earth |7918 5. 53|1-0000 I.00 6.95 4.91 1,240,000 II.....| Venus |7701| 4.85 ?]0.807? 0.85 6.33 4.48 I,156,000 | (836,000) IIl....| Mars |4330 3.58]0. 1065 0.36 3.06 2.16 588,000 (898,000) TWieee: Ideal |3407| 3.50/0.050 On27, 2.38 1.68 458,000 IV Neueteecns Moon |2160 3.34/0.0122 0.16 1.47 1.04 286,000 (50,000) Vie Ideal |1000) 3.30/0.001202 0.075 | 0.68 0.48 132,000 VII ...| Ideal | 500/(a) 3.30]0.0001503 0.0375 | 0.34 0.24 66,000 (b) 5.c0]0.000228 ©.057 | 0.42 0.30 76,000 VIII... | Ideal | 1oo|(a) 3.30/0.000001202 | 0.0075 | 0.068 0.048 13,200 (6) 6.c0/0.000002186 | 0.0137 | 0.123 0.087 16,000 IX.... | Ideal 50|(a) 3.30/0.0000001503| 0.00375] 0.034 | 0.024 6,600 (b) 6.50]/0.0000002960] 0.0074 | 0.067 | 0.047 8,200 XG Rr Ideal 10|(@) 3.30|1.202 10-9 | 0.00075) 0.0068 | 0.0048 1,320 (6) 7.00|/2.550xX 10-9 | 0.00160) 0.0144 | 0.0102 1,700 The selections are adapted to the earth as unit and the spheres of control are based on the earth’s distance from the sun. An ideal body 2) the mass of the earth is introduced between Mars and the moon to better grade the series, and for a like reason an ideal body 1,000 miles in diameter is introduced below the moon. The four ideal bodies, 500, 100, 50, and 1o miles in diameter, respectively, are selected to cover the range of the planetoids and smaller satellites. The largest of planetoids thus far measured satisfactorily is 485 miles in diameter (Barnard). Two hypothetical densities are assigned to each of these, the one to represent bodies supposed to be composed largely of stony matter, the other to represent those that may have a notable content DIASTROPHISM AND THE FORMATIVE PROCESSES 479 ofiron. In the 1o-mile body this higher density is put at 7, which is thought to be as dense as any such natural aggregate, inevitably more or less mixed and porous, would be likely to be. Column / gives the maximum acceleration of gravity at the surface of the given body, stated in percentages of the acceleration of gravity at the surface of the earth. Column g gives, in miles per second, the parabolic veloc- ity (=velocity required to give to a projected body a parabolic path= velocity capable of carrying a body to infinity=velocity acquired in a free fall from infinity = “‘velocity from infinity’’). In discussions of the limitations of atmos- _ pheres, this “velocity from infinity” has very commonly been used erroneously as “the critical velocity of escape,’”’ but by referring to column 7 it will be seen that a molecule shot away from these bodies may reach the limit of the body’s control very much short of an infinite distance. If one wishes to show that the molecules must escape, and desires to make his statement conservative by leaving a good “margin of safety” to cover defects in data and otherwise, the parabolic velocity is a very suitable criterion to use. If, on the other hand, one wishes to show that molecules will be retained, and desires, as before, to leave a margin of safety for retention somewhat above that, the figures in column / form convenient criteria. Strictly speaking, these represent the velocity required to give a particle a circular orbit at the surface of the body, and this velocity forms a dividing line between the ordinary collisional atmos- phere and the orbital ultra-atmosphere. The latter forms the transition stage through which molecules may escape from control with the least velocity. The velocity in a circular orbit has a fixed ratio to the parabolic velocity for the same point, viz., 1:;/2. The figures for the parabolic velocity and for the velocity of circular orbit or “‘velocity of retention’? each becomes lower as the points of reckoning rise above the surface. The minimum velocities required for escape lie between the velocity for circular orbit and the velocity of fall from the limit of the sphere of control and are dependent on the mode of escape. Column 7 gives the diameters of the spheres of control of the several bodies in competition with the sun at the earth’s distance. It is important to note the qualifying clause, for spheres of control vary with the distance from the controlling body. The actual spheres of control of Venus and Mars are given in parenthesis. For the present discussion spheres of control at the distance of the earth are most serviceable. In the case of the moon, the figure in parenthesis represents the moon’s sphere of control as against the earth, within whose sphere of control it revolves. It is worthy of note that the spheres of control at the lower end of the series, notwithstanding their diminution, still have notable dimensions. These spheres of control give concrete pictures of the areas over which the several bodies exercise collecting as well as holding power, while the figures in columns g and hk give data for realizing, in terms of velocity, how limited is the power of this influence in the smaller masses. 480 T. C. CHAMBERLIN The table would be additionally helpful if the central pressures, the central densities, and the central temperatures could each be given in terms equally trustworthy, but determinations of these properties rest on a much less secure basis. The central pressures can only be determined by assuming some law of downward increase of internal density. The actual rate of such increase is uncertain, beyond the fact that it must fall within certain rather broad limits defined by precession and other astronomical phenomena whose requirements are not precisely determinable. Laplace’s theoretical law of density is perhaps the most plausible and is the one commonly used in preference to such others as have been proposed. Using it, MacMillan finds the central pressure of the 1o-mile body, when assigned a mean density 3.30, to be only 11.8 lbs. per square inch, i.e., less than the pressure of the earth’s atmosphere. On the other hand, that of the present earth is 22,500 fons per square inch, or about 3,000,000 atmospheres. The results given by Laplace’s law are in general accord with those obtained earlier in this dis- cussion from a comparison of the moon, Mars, Venus, and the earth. However, reserve in placing implicit confidence in this law is to be observed, for by carrying the series of determinations upward from the earth on the same basis, MacMillan finds that at a radius of about 5,000 miles the central density becomes infinite. This seems to mean either that the law breaks down or else is rendered inapplicable by some intercurrent factor whose nature is as yet unknown. Dr. A. C. Lunn reached results of similar import in his geophysical studies under the planetesimal hypothesis in 1909.2 The suggestive correlation of the densities of the whole series of planets made by MacMillan in his paper “On Stellar Evolution” deserves thoughtful consideration in this connection.’ It is clear, then, that until some elucidation is found for this singular result so shortly reached after the dimensions of the earth are passed, it is unsafe to build important conclusions upon the law. “The Order of Magnitude of the Shrinkage of the Earth Deduced from Mars, Venus, and the Moon,” Jour. Geol., Vol. XXVIII (1920), pp. 1-17. 2 “The Tidal and Other Problems,” Carnegie Publication No. 107 (1909), pp. 201-2. 3 Astrophys. Jour., Vol. XLVIII (July, 1918), pp. 36-40. DIASTROPHISM AND THE FORMATIVE PROCESSES 481 In reaching conclusions respecting the central temperature there is not only the danger of error due to deducing it from com- pression computed according to this doubtful law, but there are other sources of uncertainty, among which one of the more serious is the unknown amount of heat removed by inherited eversive movements within the body, co-operating with ordinary convection — while the formative processes were in progress, and since. In the distinctly large bodies these might not perhaps rise to decisive value, but in the smaller orders of bodies the central temperature theoretically assignable to concentration might be so far dissipated by the combined effects of inherited eversive movements, convec- tion, conduction, and viselike mechanical action (discussed below) that it would have but limited effect on the physical state of the core into which the nucleus passed. A further serious difficulty in estimating central temperature arises from our ignorance of what part of the potential energy theoretically set free by compression went into endothermal re- organizations, what part became latent in forming solutions, what part was carried surfaceward by the forced ascent of these solutions, and what remained to increase the temperature. THE GROUP OF FACTORS THAT CONDITIONED THE PROCESS OF NUCLEAR CONCENTRATION The passage of the planetary nuclei from their original states as solar gases into their final states as the cores of planets, planetoids, and satellites was by no means so simple a process as gaseous condensation has usually been regarded. Beside the simple con- densing process, as usually considered, there were co-operating activities that radically modified the general tenor of the process. Four of these require consideration: I. Several types of motion were inherited from the solar eru piion, and these took the lead in determining the internal circulation. The thermal convection, as it arose, was superposed on these. Il. A sifting of the mixed molecules of the original gaseous matier set in almost as soon as it emerged from the sun and changed the mixture to the proper proportions for forming planetary cores. Ill. The formation of precipitates also set in early, and gradually changed the primitive gases into Brownian mixtures which themselves 482 T. C. CHAMBERLIN changed as time went on. In the smallest order of bodies this pre- cipitation, together with the escape of such molecules as remained free, went so far ultimately that the residues were reduced to clouds of precipitates which condensed in a way of their own. IV. Almost as soon as cores began to form, differential stresses, more tmiense below than above, were brought to bear upon them by external agencies which aided in working the lighter and more mobile materials toward the surface, thus developing increasing density and solidity im the central parts. In discussing these co-operating factors it will be helpful to have in mind concrete pictures of the deployment of the matter under study, fashioned in the form of spheres of control, for these best bring out the dynamics of the organizing work. The matter . in spheres of control may be very differently distributed, but it is to be regarded as occupying in some measure, however scant, the whole space. In adult organizations, usually the matter is highly concentrated toward the center and very sparsely distributed in the outer part. In the initial stages the distribution is likely to be heterogeneous with less difference between the outer and inner parts. Uniform distribution therefore becomes the most conven- ient standard of reference, though probably never realized in fact. Table II gives data from which selections may be made at pleasure in forming representative pictures. The primitive earth-mass, before sifting bean’ should have included (1) the light gases that later escaped and were never recovered, (2) the planetesimals that temporarily escaped and were later recovered, (3) the. nuclear portion that remained under self-control, and (4) minor factors that may be neglected. Let the whole be taken as having a mass about twice that of the present earth, without prejudice to a higher or lower final estimate. It would then, if properly distributed, have a sphere of control of the order of 1,500,000 miles in diameter and a mean density for the whole sphere of about o.oo11 on the air standard, or, let us say, approximately 1/1,000 of the density of the air at sea-level. As loss by sifting went on the sphere of control should have shrunk to a minimum, after which, when planetesimal accretion began to DIASTROPHISM AND THE FORMATIVE PROCESSES 483 overmatch the molecular escape, it should have grown to its present size. It is not best just yet to try to decide what this minimum may have been, but let it be placed as low as 5}, of the present mass of the earth to make the gap between our two pictures wide. Its sphere of control would then have approached 500,000 miles in diameter and the mean density for the whole sphere .oort1, on the air standard. The spheres of control are here computed on the supposition that the matter in each case is distributed in spherical form and that each concentric layer is homogeneous. Actual spheres of control are not strictly spherical and the distribution of matter at least in the early stages was probably not homogeneous. The figures given are themselves only convenient approximations, but they serve well enough to indicate the general order of tenuity. Only gravitative attraction is taken into account. The phenomena of comets’ heads imply that there is a supplementary force in such very diffuse bodies, perhaps electromagnetic, but that may be regarded as merely a “‘margin of safety”’ in this discussion. Among the points to be noted, though they cannot be discussed here, are: (1) the high degree of tenuity, which gives some notion of the extent to which matter may control itself in the terrestrial part of the solar domain; (2) the temperature produced by the expansion of the solar gases to this degree of tenuity; (3) the facilities for radiation afforded by this tenuity; (4) the nature of the internal movements in such tenuous bodies. I. The motions inherited from the solar eruption and thetr co- operation with convection in modifying the condensation of the nuclet.— The gaseous matter erupted from the sun inevitably carried into its new activities some measure of the turbulence that previously affected it, while the forces of ejection added to this their own differential impulses. In so far as these impulses had uniform effects on all parts of the erupted mass, they merely served to send the whole out into its orbital course. This does not specially concern us here, but we may note in passing that the uniform increments of motion discovered by Pettit in the solar eruptions of May 29 and July 15, ro19, seem to be singularly felicitous factors in promoting projection into orbits without those high tendencies 484 T. C. CHAMBERLIN to dispersion naturally assigned to simple explosion and that would be unfavorable to self-control.t We are here concerned only with those differential impulses that affected the relations of one part of an ejected mass to other parts. It is assumed that these differ- ential impulses were so graded that (1) they scattered into planetesi- mals a notable part of the projected mass, (2) they tore away from the central portions segments that were massive enough to hold themselves together, but not very firmly, while (3) the main: central masses retained a higher degree of self-control. Such a partition of effects seems the natural result of the mechanics of eruption. It seems also to fit the requirements of the bodies that now make up the planetary system. ‘The masses that retained their self-control were the nuclei of the organizations that were to follow, and constitute the theme to which we are here confined. Under such a range of impulses the nuclei probably graded from the largest and most strongly held down to small diffuse ones on the very limit of self-control, beyond which complete dissipation into planetesimals set in. They are therefore to be dealt with as a graded series rather than a single type. The question of control, however, was not so much a matter of mass as of balance between the force of gravity and of the motions involved. Three types of inherited motions need recognition: the turbulent, the vortical, and the rotatory. Turbulent motions were not only inherited directly from the sun, but must have been generated by un- balanced thrusts and drags incident to eruption. Eruptive actions almost inevitably give rise to more or less of vortical motion, or at least some form of eversive motion. In free interplanetary space, and in such tenuous bodies as those under discussion, motions of this type might persist long and be specially effective in discharging internal heat. All unbalanced differences of thrust and drag in the ejection of erupted masses would ultimately appear in the form of rotation and so rotation of some order could scarcely have failed Jour. Geol., Vol. XXVIII (February-March, 1920), pp. 145-49, or the original article by Pettit, Astrophys. Jour., Vol. L (October, 1919), pp. 206-19. 2Tn addition, the least projected parts fell back to the sun, while doubtless some particular parts received cumulative impulses and were thrown into anomalous courses, but these are neglected throughout this discussion. DIASTROPHISM AND THE FORMATIVE PROCESSES 485 to be inherited from the ejection. The amount of this primitive rotation, however, is not deducible directly from such rotations as the bodies now have, for the present rotations are assignable chiefly to the effects of planetesimal infall after the nuclei had become planetary cores. This later effect was conformable to a law of equilibrium under which the rotations were sometimes accelerated and sometimes retarded." When these inherited motions were strong enough to cause dis- sipation, the nuclei of course vanished into planetesimals, but when they were mild enough to be consistent with control, they became vital factors in the process of concentration. The normal system of thermal convection was gradually developed later and hence had to conform to the inherited motions already in control of the matter. In large nuclei the convectional motions might come in time to dominate the inherited motions, but in the smaller diffuse nuclei that were more quickly cooled it perhaps never came to be more than a secondary factor. At any rate, the dependence _ of the convective circulation on the inherited motions—merged mainly into rotation later—must have given rise to a distinctly gyratory system of circulation. This doubtless had some analogies with the circulations of the atmosphere and of the ocean, which, though essentially thermal, are profoundly affected by the earth’s rotation. A fundamental difference, however, needs notice. We are here dealing with hot bodies whose radiation is primary. The surface of a rotating body has its greatest convexity transverse to the equator, while the polar surfaces are relatively flat. The greatest radiation in proportion to the immediate submass there- fore takes place in the equatorial region. In addition to this the escape of molecules is aided by the centrifugal component of rota- tion, which is greatest at the equator and sinks to zero at the poles. It has already been noted that escaping molecules carry off ther- mal energy in relatively high amounts. The escape of molecules may then be regarded as a form of quasi-radiation. It is, therefore, a rather firm inference that the equatorial belt is the most effective cooling tract of a hot, rotating body, though this may easily be masked by the high radiation from all surfaces. t The Origin of the Earth (1916), p. 99. 486 T. C. CHAMBERLIN It is a notable fact that the equatorial belts of the sun, Jupiter, and Saturn rotate faster than portions of their surfaces on the same meridian in higher latitudes. This has been the subject of much speculation and has received different explanations, more than one of which may contain a measure of truth. One suggestion is that it is due to the infall of planetesimal matter. A closely allied suggestion is that it is due to the falling back of matter ejected from the sun into the planetary regions and drawn forward in the direction of their motion, so that on returning it carries surplus momentum acquired from the planets. These are not inconsistent with the suggestion here made that part of the acceleration may be merely a phase of circulation normally set up in such hot rotating gaseous bodies. In a hot fluid body of the volume and rate of rotation of the earth, a mass, cooling and sinking from the equa- torial surface, would—if it were free from contacts with surround- ing matter—acquire an orbital velocity before it reached the center, and hence would sink no farther because the centrifugal component of its motion would wholly offset the pull of gravity upon it. If forced below that depth, its centrifugal component would act as a buoyant force. Of course, the sinking mass never would be free from contacts, and so it would necessarily exchange energies with the contact matter. The sinking mass would thus act as an accelerating undertow for any matter that flowed in above it as it sank; so also it would tend to drag forward whatever was in contact with it on its sides and below. It is not difficult to work out a system of circulation actuated by such equatorial cooling and sinking. It would, however, undoubtedly be subordinate to the inti- mate turbulence that would spring from other factors. The axial tract would present a unique problem, for it would be little affected by rotation and would not directly be reached by the descending equatorial currents, for they would be restrained by the centrifugal component of rotation and turned northward and southward, completing their circuits by return from the higher latitudes with such deflections as rotation imposed. ‘This part of the circulation may be pictured as two vortex rings made up of spiral submove- ments trending downward on their contact sides at the equator and upward on their poleward sides. The axial tracts in themselves DIASTROPHISM AND THE FORMATIVE PROCESSES 487 would seem to invite a more direct and simple convection, but they might be specially subject to influence from the inherited motions. For example, if the rotation were west-east, like the sun’s rotation in which the mass participated before ejection, and there were a north-south axial movement as in the case of the eruptions of May 29 and July 15, r919, cited above, there might naturally be inherited from this an axial movement from one pole through the center to the other. The original tenuous state would apparently be favor- able to this, and, once established, it might be perpetuated as an effective form of central convection. A special interest attaches to this from its possible influence on the solid core as that gradually formed—but this cannot be discussed here. The point to be emphasized is the inevitable subordination— in the early formative stages—of the convection actuated by differ- ence of temperature to the inherited motions. ‘The circulation, far from being simple descent and ascent, was tortuous and involved, and the core-forming process must be interpreted on this basis. Il. The molecular sifting of the nuclet required to reduce the original solar gases to the composition of the present solid bodies.— The nuclei of the giant planets may be passed by, merely remark- ing that there is little reason to think they suffered much sifting; rather they seem to have been so massive from the outset that they retained all classes of molecules that came under their control. By far the larger number of the solid bodies of the solar system, however, are practically devoid of free gases, and seem to be formed almost wholly of stony and metallic matter. None of the terrestrial planets carry more than a very small percentage of free gases; apparently almost their whole substance consists of stony and metallic materials such as make up the main body of the earth and of meteorites. The original gases of all the bodies derived from the sun, large and small alike, should have had essentially the same composition. Spectroscopic analysis shows that the visible substance of the sun is an intimate mixture of many kinds of molecules. Unfortu- nately, their relative proportions can merely be inferred in a general way. The low density of the sun (1.40), notwithstanding its great mass, implies—even when its high temperature is considered— 488 T. C. CHAMBERLIN that the lighter elements form a notable factor, and the spectro- scopic evidence tallies with this. The low densities of the giant planets derived from the sun (Jupiter, 1.25; Saturn, 0.63; Uranus, 1.44; Neptune, 1.09) suggest a similar constitution with even more cogency, for they are much less affected by high temperature. If, therefore, solid stony or metallic bodies of high specific gravity were to be formed from outbursts of solar gases, the process must have involved the removal of large quantities of the lighter order of constituents. This sifting is precisely what the kinetic theory of gases applied to small bodies would lead us to expect. The process is essentially a form of evaporation, and so the planetoids and satellites, as well as the terrestrial planets with slight qualification, may be regarded as merely the residues of the selective aap onaTaa of much larger original bodies of mixed gases. If the original gases, after they were projected from the sun, occupied some large part of the spheres they could control, as indicated above, the escape of the lighter molecules would be rela- tively easy and prompt, at least from the smaller masses. If the nuclei became much condensed before the sifting was completed, the remaining escape might be slow, for the molecules could then only escape from the outer zone where free paths were open to them when they chanced to rebound in an outward direction with the requisite velocity. In so far as the original gaseous masses were affected by turbulence, or by vortical or other eversive motions derived from their ejection, the escape of the light molecules would be facilitated. The motions inherited from the original expulsion were probably such that the dominant tendency, in all but the more massive nuclei, was toward gaseous dispersion. Not only would the light mole- cules be likely to escape from control, but many ofall kinds. This is only another form of stating the primary tenet of the planet- esimal hypothesis, viz., that such dispersion was an inevitable effect of the solar eruptions, and a condition precedent to planet- esimal accretion later. There is merely the reservation that enough material was held under self-control to serve as collecting centers of the requisite orders of efficiency, but even this is not essential to an ultra type of planetesimal genesis. It seems, how- DIASTROPHISM AND THE FORMATIVE PROCESSES 489 ever, to be definitely implied by a posteriori reasoning from the existing bodies. The present line of attack shows that the nuclei, except the four of the giant order, were little more than the residues of the heavier material left by selective molecular action working on larger original bodies of mixed gases. This seems to apply to all satellites, to all planetoids, and, in qualified degrees, to all planets from the earth downward. The process of evaporation had the effect of reducing the energy of the residue per unit mass, and this, added to the inevitable loss by radiation, made control incréasingly secure and caused loss to diminish till it became negligible. III. The formation of precipitates and of Brownian mixtures, grading into quasi-gaseous clouds of precipitate aggregates.—As the original mixed gases emerged from the sun, expansion, abetted by radiation, must have promptly lowered the temperature, and this lowering of temperature doubtless led to the formation of precipi- tates. It is immaterial just here whether these precipitates were formed by simple cooling or by chemical action, or by both acting jointly. Nor is it of critical importance whether the precipitated particles were liquid or solid. It is highly probable that the earliest precipitates were formed of material such as later became the stony and metallic substances of the earth, of meteorites, and probably of all the small solid bodies. That such precipitates had begun to form even earlier is highly probable, for they are appar- ently now forming in the sun; at least the solar photosphere is commonly interpreted as a cloudlike zone of such precipitates. At the outset such precipitates would necessarily be minute and diffusely scattered, for under the law of diffusion of gases the par- ticular molecules that were precipitable at the temperatures exist- ent at that particular stage would be distributed sub-uniformly throughout the turbulent mixture of molecules which formed the gaseous mass, but aggregation into granules, chondrules,’ or other forms of concretions would doubtless at once ensue, after the analogy of the droplets and crystals of clouds. tJ] venture to name chondrules here to suggest that conditions such as these are perhaps those most likely to have given rise to these singular little aggregates found in the majority of meteorites. They are commonly of the size of a millet seed, but range up to that of a walnut and down to dustlike fineness. 490 T. C. CHAMBERLIN The minute precipitates thus scattered through the gas would serve as Brownian particles, and the increase of these would form a progressive series of Brownian mixtures. The minute precipi- tates would be jostled to and fro much as the free molecules were, except that, on account of their greater sizes and masses, they must have responded rather to combined molecular impacts than to single ones, while their lack of perfect elasticity must also have somewhat toned down these activities. An analysis of the conditions makes it clear that the Brownian evolution probably diverged very soon into two rather distinct lines, though they must have been united by numerous intermediate phases. One of these may be regarded as the typal line of gaseous descent; the other as divergent toward an alien type that combined a quasi-gaseous phase with a partially orbital factor. In the first the characteristic feature continued throughout to be that of a jostling assemblage, though the original high proportion of mole- cules gave place more and more to precipitates acting as Brownian particles. The gas in this case is presumed to have passed into the liquid phase and thence into the solid form. In the more divergent line the assemblage lost its free molecules largely, and in the ex- treme cases entirely, and became at best merely quasi-gaseous, with a trend toward orbital behavior. Though truly gaseous at the start, the molecular assemblage soon began to be seriously depleted by the escape of the more active molecules and the passage of the rest into precipitates and thence into aggregates, while these tended to lose their to-and-fro dynamics and take on circulatory, rotatory, or revolutionary dynamics. | Divergent as these trends were, they were readily reversible. When molecules escaped from a nucleus in which their habit was strictly gaseous, they usually took on a specific orbital habit and beame planetesimals; the accident of an encounter, however, might easily throw them back into to-and-fro oscillation. Notwith- standing such reversals, two quite contrasted systems of dynamics arose and were continually contending with one another in the processes that marked the passage of the nuclei into cores. The gaseous line of descent was obviously dominant in the nuclei of the giant planets. Perhaps it was also in the nuclei of the DIASTROPHISM AND THE FORMATIVE PROCESSES 401 earth and of Venus. But just how far down the scale it held its dominance may best be left an open question for the present. The considerations about to be offered imply that the second line of evolution was preponderant in the history of the small nuclei. In nuclei massive enough and quiescent enough to maintain high internal temperatures it seems probable that the precipitates would generally pass from the gaseous state directly into liquid droplets which would serve for a while as Brownian particles and gradually gather into liquid cores, which in turn would develop solid precipitates within themselves and ultimately collect into solid cores. In following this more typical line of gaseous descent, however, it is important to discard the old view that magmas are melts, and to replace it with the modern view, now well established, that magmas are mutual solutions. There are of course melts, and melts sometimes freeze, and so melting and freezing have some place in geological processes, but it is a rather trivial one compared with solidification by chemical processes. Even on the present surface of the earth, which for a hundred millions of years or more has been developing a temperature contrast between the exterior and the interior, simple refrigeration has little expression except in the form of thin crusts; the interiors of even surface lava flows or pools have solidified chiefly by crystallization brought about by saturation in the mutual solution. In a nucleus so conditioned as to sustain the progressive collection of a liquid core at its center by hot precipitates from enshrouding gases there is little ground for postulating even the trivial crust formation that takes place on lavas poured out on the present cold surface. Superficial refrigera- tion could scarcely have been more than a negligible process. Appropriate temperatures and pressures must of course have been very essential factors in core formation, but rather as imperative conditioning influences than as direct agencies. They were less intimate and ultimate factors than the chemical forces that served as the immediate actuating agencies. : Unfortunately, present knowledge of the precipitating processes deep within magmas is insufficient to predict with much confidence the history of even an ideal liquid core in a perfectly quiescent state, much less to forecast the solidification of a core actuated by ~ 492 T. C. CHAMBERLIN such a tortuous circulation as the case in hand seems really to involve. Inquiry should, however, at least be put on the right track by recognizing the later aspects of science and the physical realities of the case. It is at least safe to say that isolated crystals are habitually formed within magmas, not merely on their surfaces. In addition to this it is particularly important to recognize that the order of formation of minerals in magmas is not that of their melting- points, but rather singularly at variance with it. Some of the minerals commonly formed earliest, as magnetite, apatite, and ziroon, are higher in specific gravity than the average minerals formed later, and these are generally higher than the liquid from which they were separated. It is quite reasonable to suppose, as leading petrologists do, that the heaviest order of minerals, at any rate, if not the majority formed, would tend to sink through the mutual solution. The actual effectiveness of this tendency must, of course, be dependent on the viscosity of the magmas, the vigor of the circulation, and other conditions. Whether the heavy minerals would remain solid and collect at once at the center or be redissolved in the depths and continue longer in the circulation is doubtless to be left an open question for the present. But this and other questions are to be considered under the conditions of a tortuous circulation rather than those of a quiescent liquid. The tendency of the circulation must certainly have been to equalize the temperature and to favor a slowly progressive precipitation affecting large portions, if not all of the mass, rather than the mere surface. The heavier precipitates might then rather plausibly be assumed to collect where the combined effects of current and gravity offered them the most available resting places. If so, a core shaped to fit such conditions seems more probable than a strictly symmetrical sphere. If we turn now to the other type of nuclear evolution—in which the sifting action not only went to greater lengths, but the sifted residue was much more affected relatively by motions inherited from the expulsory action—it is well to recall at the outset that the range of cases stretches from the largest solid bodies notably affected by the sifting process downward to the very borders of DIASTROPHISM AND THE FORMATIVE PROCESSES 493 complete dispersion, such complete dispersion springing variously from inherited motions, from thermal convection, from molecular activity, or from divers combinations of these. If, as the natural- istic method insists, the solid bodies themselves are to be taken as the vestiges of the actual process, the observed range in size tallies with our previous suggestion that the limits which permit success along this line are actually reached in the existing series. Appar- ently our best method, then, is to consider the whole range for the sake of learning what were the inhibiting conditions at the vanish- ingend. We may then the better form an opinion of what probably took place nearer the middle of the great series where our interest chiefly lies. A naturalistic method is much to be preferred to a deductive treatment, for the latter is embarrassed by the multitude of possible assumptions. In pursuance of the naturalistic method let us seek some telltale feature that has been actually realized and make that our base of procedure. ‘The series of atmosphereless bodies furnish such a base. They tell us within what bounds the inhibiting limit lay for such gases as form atmospheres. In passing through the actual conditions of evolution they have been stripped of all gases as light and active as nitrogen or oxygen. Further than this, they have maintained that condition since. The condi- tions must probably have been most exacting in the hot genetic stages, and there has been chance for recovery since. Their present condition, with some reservations, may be taken as an approximate indication of equilibrium conditions. The graded list in Table I giving the range of planets, planetoids, and satellites, from the earth down, may be found convenient here. The case of atmospheric gases being thus approximately deter- minate, it remains to find at what stages the gases or vapors of such stony and metallic substances as make up the earth, meteor- ites and like bodies, would encounter their inhibitive limit. The basis for this lies in the fact that the molecular velocities of mole- cules vary inversely as the square roots of their molecular weights. The heavier we assume the molecules to be the more conservative our conclusions, so let us assume that the small nuclei were com- posed of molecules as heavy as those of the leading minerals in meteorites. The square roots of the molecular weights of the nine 494 T. C. CHAMBERLIN minerals commonest in meteorites, including iron, are 10, 10.59, 12.49, 14.60, 15.87, 16.18, 10-40, 16.70, 18-28. “hese arejtomme compared with the square roots of representative molecules that are not held by the atmosphereless bodies. We may take the molecules of oxygen and nitrogen as representing these, the square roots of their molecular weights being 5.66 and 5.2, respectively. The high temperatures at which alone the stony and metallic sub- stances occur in working quantities enter vitally into the case.t Making requisite computations, it appears that these heavier molecules would not be held under genetic conditions by the four lower orders of the bodies givenin Table II. This seems to force the conclusion that the planetoids and smaller satellites were not formed in a purely gaseous way. As this is a rather radical conclusion it is well to note that the premises have not been strained to reach this result but quite the opposite. The bodies have been taken at their full present masses, whereas only their nuclei were really involved during the critical genetic stages. The molecules are taken in their present complex state, whereas in their volatile state they were quite possibly simpler and hence more active. The attractive power at the surface of the present cold concentrated bodies was used, whereas the attractions at the surface of the expanded gaseous bodies would be much lower. Other concessions to conservatism were made. But this only excludes a direct or immediate formation by the gaseous method. It leaves open the question whether or not the cloud of precipitates into which the original mixed gaseous sub- stances naturally passed could have completed the work. I so, the genesis might have lain in the alien line of gaseous descent, though not in that of strict gaseous formation. It was noted earlier in the discussion that the gases of the stony and metallic substances must have begun to be precipitated soon after expulsion from the sun. In the small detached segments the precipitation must probably have gone on rather rapidly to com- tI am under obligations to Dr. Fred. E. Wright, of the Geophysical Laboratory of the Carnegie Institution of Washington, for information and advice on points involved here, as also to Professor W. D. Harkins, of the University of Chicago. In the statements made I have endeavored to avoid all uncertain ground and leave everywhere a margin of safety. DIASTROPHISM AND THE FORMATIVE PROCESSES 495 pletion, and the work of aggregation into granules probably fol- lowed closely after. The conditions for the escape of the molecules - that remained free would also be favored by the smallness of the bodies and the condensation of the precipitated portion. The escape of the free molecules would leave the precipitate aggregates with such internal motions as were inherited from the previous states. The last previous stage was that of a Brownian mixture whose internal motions did not differ radically from those of true gases, but the growing inelasticity must not be overlooked. The laws that would have governed the cloud of precipitates when first formed would not have differed very widely from gaseous laws. The inherited motions had, however, as we have seen, introduced a tendency toward an orbital development. In general the pre- cipitated particles in a Brownian mixture so conditioned would not fall directly to the center even if an open path were provided for them; on the contrary they would pursue elliptical orbits about the center. By interference they would undoubtedly at length reach the center but only through a delayed course with consequent dissipation of energy. Now the units—which at the start were perfectly elastic mole- cules—would by the precipitating and aggregating processes grow into granules many million times more massive, and in the process would become increasingly inelastic. By this change in the nature of the. unit there would have arisen a wide gap between even the heaviest of the free molecules and the average spherules, granules, or chondrules into which the precipitates passed. The velocities of the latter would-have been of so much lower order that there . seem no good grounds to doubt that the main mass of the latter would be susceptible of control and continued concentration under conditions that would be quite prohibitive of control as free molecules. The very process of molecular escape tended in itself to increase the gap between the units prone to escape and those prone to continue their concentration. In every collision from which a molecule escapes by rebound there is an equal reaction of the partner in collision in the opposite direction. The escaping mole- cule usually has the lesser mass and to give escape the rebound must 496 T. C. CHAMBERLIN be outward; the reacting molecule or granule therefore rebounds inward. The very process of dispersion was therefore mated with a concentrating process and the two divided their results between the - forming of nuclet on the one hand and of planetesimals on the other. On the residual side of the twin process the ultimate result was the formation of a cloud of precipitated granules from which all free molecules had escaped. The cloud of course had less mass than the previous mixed nucleus, but there was a proportionately larger reduction of dispersive activity. In the light of this we need next to consider further the holding power represented by the sphéres of control. The spheres of control in Table II are computed for the earth’s distance from the sun. They would be relatively larger farther out and smaller farther in. By reference to the table itewill be seen that the fields under control are by no means insignificant even for the smallest bodies represented. At the same time, reference to the adjoining columns of the table will show that the strength of control is distinctly limited. It is also to be noted that the velocities of retention and escape are given for the surfaces of the concentrated bodies as these now are, and that the velocities that can be controlled decline rapidly for points farther from the center. Now expansion does not affect the simple static holding power so much as it does the velocity that can be controlled. Within the limits of the sphere of control, and with some other qualifications, simple expansion or contraction does not affect the extent of the sphere of control. It is a principle of celestial mechanics that if a _body is spherical and if its substance is distributed either uni- formly or in homogeneous concentric layers its gravitative effect on bodies outside it is as though the whole matter were concen- trated at the center, and hence, of course, expansion or contraction is immaterial so far as relates to bodies on the borders and outside the body itself. If the body is not strictly spherical or homogeneous in concentric layers, the deduction will not strictly hold, but any departure will in general be measurably in proportion to the de- parture from sphericity or homogeneity, so that the principle may be used without radical error in respect to normal spheroids of revolution. Applying this deduction to the range of bodies rep- DIASTROPHISM AND THE FORMATIVE PROCESSES 407 resented in Table II, the sizes of the spheres of control will remain about as given whether the substance they contain takes the form of an expanded gas, or an open swarm of precipitated granules, or a compact solid body. This puts everyone in the way of modifying at pleasure the illustrations I offer. To the concrete pictures already given the following may be added as now more immediately serviceable. From the minimum radius of the sphere of control of the earth, 620,000 miles, let a depth of 20,000 miles on the outer border be left essentially unoccu- pied and the whole present substance of the earth distributed uniformly throughout the remainder. It would have a density of 0.001266 on the air standard. In the form of a cloud of granules, each half the mean density of the earth. and distributed uniformly, the empty space about each granule would be over a million and a half times the space occupied by the granule itself. If a ro-mile planetoid were converted into a cloud of granules uniformly distributed through its sphere of control, the cloud would have a density of o.oo11r on the air standard. If the gran- ules had the same density as in the planetoid, the average empty space about each granule would be more than two million times the volume of the granule itself. If therefore the clouds of granules were quite diffuse, they yet might be controlled by their mutual gravity, provided the dis- persive components of their internal movements were negligible. But with such wide distribution any appreciable dispersive move- ments would be fatal to control. To fashion a case of this order with a working margin of space, let the matter of a 1o-mile planetoid of density 3.3 be dispersed uniformly as granules of like density throughout the central 4 of its sphere of control, leaving the remaining 7 as empty space which the granules must cross to escape. The density on the air standard would be 0.00889, while the average empty space surround- ing each granule would be 280,000 times the volume of the gran- ule itself. Even in this case the velocity at the surface that would give escape if directed outward would be perilously low, not above a fraction of an inch per second. This reveals the critical nature of all this class of cases. To insure success in final concentration, the 408 T. C. CHAMBERLIN sifting process that preceded must have removed all constituents whose motions had any notable dispersing component; nor can any such component arise from mutual interaction without jeopardy. Almost the only line along which a body so small as a 10-mile plan- etoid could organize itself by the granular method seems to have lain in acquiring very early a higher central density and a less out- ward extension than that assigned. This might perhaps have been done by the reaction above noted. ‘The peril of dispersion and the narrow margin of control in such cases lead to the conclusion that the smallest order of planetoids and satellites lie near—or perhaps quite on—the limit of possible genesis by even this divergent phase of the gaseous line of descent. This conclusion tallies with the fact that no planetoids or satellites of the smaller order are known at the earth’s distance from the sun, or within it. Bodies of this type appear only at the distance of Mars and beyond. The dynamic conditions of this inner region are perhaps too adverse for this type of formation. In the outer region conditions are notably less restrictive, but even there they undoubtedly put lower limits on the size of bodies formed by the gaseo-granular method of assemblage. In the light of these considerations there seems little warrant for supposing that such bodies were ever formed in sufficiently great multitudes to have built up the earth or to have pitted the surface of the moon by their impacts. The number of lunar craters is estimated at 30,000. If each of these is the grave of an extinct planetoid, one might expect that a few living ones would have lingered to tell the story. The negative testimony of the heavens as to their existence in this region seems rather to favor the view that their restriction to the outer region implies that they are themselves witnesses to the limitation of this line of genesis in both place and frequency. With the general lines and limits of nuclear evolution thus defined, our remaining task is to find the median places between the two extremes that fit the earth, Venus, Mars, and the moon. It is clear from the present state of these bodies that much sifting of the original solar gases was required, for while the earth, Venus, and Mars hold envelopes of gases of moderate molecular weight, DIASTROPHISM AND THE FORMATIVE PROCESSES 499 they do not hold hydrogen and helium, which.abound in the sun in appreciable quantities. The sum total of the gases they do hold relative to the whole mass of these planets is very small. Even in the case of the earth, distinctly the most massive of the solid bodies, the sifting must have gone to very notable lengths. It is not impossible that the nuclei of all four bodies were so far sifted down as to exclude essentially all the atmospheric gases, and as a result their concentration fell ultimately into the precipi- tate line of descent. On the other hand, it is quite possible that the earth and Venus had atmospheres of some moment at all stages. In the case of the moon, there seems no escape from the view that its nucleus could not have formed in gaseous fashion, for the moon does not even now hold an atmosphere. In its original hot diffuse state a mass of so low an order as the nucleus of the moon could only hold its material in the precipitate form. The atmosphere of the adult Mars is so scant that its nucleus probably had no appre- ciable atmosphere. It is doubtful whether Mars could even now, in its full-grown state, hold the atmosphere it has if the planet were heated to the point of volatilizing its stony substances. The cases of Venus and the earth seem so nearly on the border line that it is not unreasonable to take either view as the evidence now stands. Further study may turn the scales one way or the other. So far as the shrinkage question is concerned, the matter narrows down to the possibility that the nuclei of the earth and Venus passed from their original gaseous states into planetary cores along the normal line of gaseous descent. If the main mass of the nucleus of the earth passed from the solar gaseous state into a cen- tral liquid magma and thence by chemico-crystalline action into a solid core, the process would have given special facilities for the adjustment of the matter in the interest of density. To that extent it would have forestalled later shrinkage that might other- wise have been recorded in diastrophic features. The record would not cover the full reality. The large amount of shrinkage deduced by our comparison of the moon, Mars, Venus, and the earth' would not be recorded even in the basal features of the earth’s configuration. These studies, however, imply that the unrecorded t Jour. Geol., Vol. XXVIII (1920), pp. 1-17. 500 T. C. CHAMBERLIN factor was not necessarily large relatively, even if the gaseo-molten phase of nuclear history did obtain and is given as generous an estimate as the data will warrant. It ought not to be overlooked, however, that the solid core, in its assigned formation by the deposition of crystals or other precipitates from the gyrating currents of the central circulation, would have been very likely to have incorporated inequalities of material and taken on asymmetries of form so as to have pre- sented a deformed foundation, as it were, for the later accretions. Such deformities would have been likely to have made themselves felt in the diastrophism of all that was built upon them. This is a phase of the subject which I hope to pursue further in the future. IV. EXTERIOR AGENCIES THAT AFFECTED THE PLANETARY CORES DURING THEIR FORMATION AND AFTERWARD The discussion thus far has been confined to agencies acting within the evolving masses. The evolution, however, was not free from influences that acted from without. One type of such action particularly requires consideration here, because it affected the successive adjustments and readjustments of material in the planetary cores. It will suffice to consider merely the case of the earth and the most typical agencies that affected it. The three agencies that lie back of changes of rotation, of nutation, and of the tides will sufficiently represent the whole. These agencies—and those of their kind here neglected—arose out of the same general processes as the planetary series itself and came gradually into func- tion as the planets themselves took form. They were more or less effective at all stages thereafter. One special effect was to bring into play the resources that lay in the mass-coherence of solids, an essentially new element in the evolution. The forces that produced the tides, the polar nutations and the changes in rate of rotation, not only caused changes of form that involved variations in the internal capacity of such inclosed spaces as there may have been, but caused differential stresses to permeate the growing cores from surface to center and call into action the viselike capabilities of stresses greater below than above. Those agencies which give rise to deformations of the DIASTROPHISM AND THE FORMATIVE PROCESSES 501 class known as zonal harmonics of the second order, such as the bulging of an equator and flattening of the poles, or the pulling out of polar cones and the flattening of the equatorial belt between, give rise to stresses much greater in the central parts than in the outer parts. Sir George Darwin" and others have computed these for an incompressible homogeneous earth and for certain com- pressible variations from this. In an incompressible homogeneous earth Darwin gives the differential stresses as bearing the ratios 8 at the center, 3 at the equatorial surface, and 1 at the poles. Ina compressible earth the surficial stresses are relatively lower and those at the center relatively higher. For a certain compressibility the surficial stresses disappear and the central stresses rise } in value.’ < Now the main tidal stresses come and go every twelve hours and the subordinate tidal stresses at other and generally longer intervals. While relatively small, they are constantly acting in a given direction, and this presumably has a certain kind of cumulative effect. This effect is doubtless chiefly felt by such molecules of the interior as are under stress and are about ready to change their attachments and so are responsive to the influence of even small strains. It is coming to be recognized that such individual molec- ular activities constitute a notable factor in rock metamorphism, glacial motion, and other geological changes of a very intimate sort. This has been set forth by Leith’ and other close students of the intimate nature of geological phenomena. Such persistent rhythmical oscillations of stress and strain as those of the tides seem well suited to aid effectually these individual molecular changes. The nutations of the poles represent pulsatory action whose periods are longer, but whose chief effects are probably of the same intimate sort. Changes of rotation, however, represent action of a much higher order of power and much greater length of period. In deformative potency, rotation has a competency of the first order. Changes of rate of rotation were probably most active and effective while t Scientific Papers, by Sir George Darwin, Vol. II (1908), pp. 476-81. ibd pS 05. 3 Leith and Mead, Metamorphic Geology (1905), pp. 173-76, and elsewhere. 502 T. C. CHAMBERLIN planetesimal accretion was in progress. The earth core was then youngest and least compressed, and so probably least rigid and most susceptible to the influence of differential stresses. I have elsewhere shown that the changes in rotation were probably oscillatory about a medial rate in conformity to a law of equilib- rium.t They may be regarded therefore as commanding influences both in respect to power and to the times and modes of application. The elevated poles and depressed equator of the rhythmical tidal deformations were transverse to the elevated equator and depressed poles of the rotational changes, and this transverse attitude no doubt lent facility to the kneading action which their rhythmical periods brought to bear on the interior of the earth. These co-operating agencies thus brought to bear on the whole interior of the solidifying earth a rhythmical series of differential stresses, most intense in the deeper parts and less intense toward the surface, and so admirably fitted to force the mobile and the lighter material toward the surface and to favor readjustments that brought about increased density and rigidity and probably also increased elasticity. It seems to me probable that this com- bination of strong mechanical stresses at distant intervals working with much gentler and more rapid rhythmical stresses has been the master-factor in controlling the secular reorganization of the earth’s interior, a gradual reorganization which I think has been in progress from the time solidification began down to the present day. The normal result, as I see it, would be a general gradation of concentrative effects from surface to center—taking form in appropriate gradations of density, rigidity, and elasticity, also graded from surface to center. The results of our comparative study of the earth, moon, Mars, and Venus tally perfectly with this view and make it theoretically logical and consistent. The steadily increasing density from smaller body to larger body, in spite of the high probability that the smaller bodies inherited the heavier molecules, points very definitely to reorganization under the influence of compression. The oscillating differential stresses, greater below than above, seem peculiarly well suited to aid in working out the. graded adjustments. t The Origin of the Earth (1916), pp. 95-110, 172-790. DIASTROPHISM AND THE FORMATIVE PROCESSES 503 The cumulative evidences of recent investigations on tidal, seismic, nutational, and other phenomena support this view with little less than demonstrative effect. The most of these are now quite familiar. There is space here merely to quote the latest numerical determinations (1917) that have come to my notice. Schweydar,' as the result of observational, experimental, and mathematical work on the tides, the polar nutations, and the trans- mission of seismic waves, concludes that the earth conducts itself as though it had a mean rigidity 25 times that of steel, that the constant of rigidity at the surface is about 3X 10'-dynes, that this increases in depth more rapidly than the density, so that at the center it reaches 30 X 10™ dynes, or ten times its value at the surface. The transverse seismic waves, as far down as the record permits a confident interpretation, indicate a definite gradation of density, rigidity, and elasticity. To insure that the total rigidity shall reach the mean value of 23 times steel, and at the same time be consistent with the rigidity known to prevail in the outer zone and the gradually rising rigidity implied by seismic waves as far down as their record is good, it seems clear that a high order of rigidity in the remaining central part is imperative. The old hypothesis of an iron core framed, among other reasons, to account for the high mean density of the earth—a purpose which it serves only clumsily—does not help much in meeting the still higher rate of rise of rigidity and elasticity toward the center, for iron is soft and malleable when hot. Nor does any special segregation of inherently heavy material in the earth, however helpful it may be in its place, fully satisfy the phenomena brought out by the comparative studies on the earth, the moon, Mars, and Venus. The whole evidence seems to point clearly to a systematic mass- effect, working on essentially the same material in all cases, aided, to be sure, but aided in only minor degree, by selective segrega- tion. In the heart of the earth very likely the segregation of the metallic from the stony material has gone much farther than in the outer parts, but I see little reason to think the two classes of mate- rial have been wholly separated from one another. A segregation =W. Schweydar, ‘Ueber die Elastizitat der Erde,’ Sonderabdruck aus Die Naturwissenschaften (1917), pp. 1-27. 504 T. C. CHAMBERLIN of the iron and allied metals into masses of moderate dimensions distributed through the stony material, after the fashion of the metallic and stony material in meteorites, would probably affect appreciably the transmission of transverse seismic waves, and so ~ account for the peculiarities of the record of such waves as come through the heart of the earth quite as well as the assumption of a purely metallic core. An original mixed constitution from center to surface kneaded into the present solidity by differential stresses whose central intensity is to their surface intensity in about the same ratio as the central density, rigidity, and elasticity is to the surface density, rigidity, and elasticity seems to fit the requirements of the case. NOTES ON THE MECHANICS OF GEOLOGIC STRUCTURES WARREN J. MEAD Structural Geology Laboratory, University of Wisconsin INTRODUCTION Since an early date in the development of the science of geology it has been recognized that secondary structural features are the results of failure or yielding of rocks under deformative forces, and students of geology have attempted to interpret these secondary features in terms of the forces and movements which produced them. Because of the great size and heterogeneity of the earth masses involved, the analysis of the casual mechanics of a given major structural feature is never a simple matter. The geologist, not having witnessed the production of the structural features at hand, or of any similar features, finds it difficult to view the prob- lem in perspective and in proper relationship to associated structural features. Discussions of the mechanics of deformation in geologic litera- ture on the whole indicate a rather elementary conception of the factors involved and a tendency to assume more or less arbitrarily a simple set of mechanical conditions, when the structures observed may be susceptible of several alternative explanations. A single structural feature or group of similar features is not necessarily indicative of the type of deformation involved. A group of inter- secting faults may be looked upon as a sequence of unrelated events, when, with equal or better reason, they might be con- sidered as essentially simultaneous and due to a single deformation. The two interpretations require a widely differing structural history of the region involved. An open fissure obviously due to tensional stresses (so far as the fissure itself is concerned) may be an incident in simple elonga- tion, shear, cross-bending, compression or shortening, or torsional 595 506 WARREN J. MEAD warping. A reverse fault implies conditions of shortening or compression but may in addition to this possibly be an incident in a general shearing movement, or,a phenomenon of simple cross- bending, or may be due to torsional warping. A series of folds may be due to shortening or compression in a direction normal to the trend of the folds or to a general shearing movement in a direction at a considerable angle to the trend. In general the shearing type of deformation has been largely neglected in analyses of the mechanics of geologic structures, both fractures and folds. It is in part the purpose of this paper to present the result of experimental work which illustrates the variety of mechanical explanations possible for a given structure, and incidentally to emphasize the extent to which many of these may be related to shearing, which the writer regards as an exceedingly common type of deformation in rock movements. It is further purposed to present and to illustrate experimentally analyses of the stresses involved in the various types of deformation. DESCRIPTION OF APPARATUS Several years ago the writer devised for use in the structural geology laboratory of the University of Wisconsin a simple type of apparatus for studying and demonstrating relations of fractures and of folds to the forces producing them. The apparatus is used in three forms as illustrated in Figures 1, 2, and 4. These are similar in construction, consisting of a rigid rectangular frame of gas pipe supporting two clamps between which a heavy sheet of rubber is stretched. One or both clamps may be moved in various ways by means of screws so that tension, compression, torsion, and shear, or combinations of these may be produced in the rubber sheet. The medium in which fractures are produced is a thin coating of paraffin applied to the upper surface of the tightly stretched rubber sheet and chilled until it is brittle. The paraffin coat is best applied by pouring the melted paraffin very rapidly and freely over the rubber sheet, which has previously been warmed practically to the melting-point of the paraffin. The hot paraffin is allowed to drain from the rubber. A little experience soon THE MECHANICS OF GEOLOGIC STRUCTURES 507 enables one to judge of the manipulation necessary to secure the thickness desired. The chilling can be accomplished by allowing cold tap water to flow over the uncoated side of the rubber. If chilled too rapidly or too much, cooling cracks will develop in the paraffin. Deformation of the rubber sheet produces systematic fractures in the paraffin bearing definite relations to the manner of deformation. Folds are produced by coating the paraffin with a thin layer of plastic wax and laying smoothly over this a very thin sheet of rubber, such as is used by dentists, or a sheet of tinfoil. When the rubber thus coated is deformed, folds are developed in the plastic wax and its coating, as described later. This type of apparatus has an advantage over the deformation of large masses of material in a compression machine of the piston type, as the forces are transmitted through the rubber and there- fore applied at every point in the coating. The thin coat of paraffin may be considered as representing the flat-lying rocks in the zone of fracture, having wide lateral extent as compared with thickness. The sheet of rubber might correspond to the deeper zone of rock flowage. Deformation of the paraffin-coated rubber is comparable to deformation affecting the zone of flow and the overlying zone of fracture simultaneously. A distribution of stresses throughout the deformed mass is obtained. The phe- nomena of repeated faults or extensive systematic joining and of repeated folds much more closely simulate nature than do the single fractures obtained in a testing machine or the single folds which develop ahead of the piston in the type of apparatus employed by Willis and others. An apparatus employing a stretched rubber sheet on which plastic layers were built up and deformed by the contraction of the rubber was employed by Alphonse Favre’ in connection with a study of rock deformation. In order to apply the compressive force he attached wooden blocks to the rubber to serve as buttresses or thrust blocks. ‘The net result was not essentially different from the results obtained by later investigaters employing apparatus of t Alphonse Favre, Archives des Sciences Physiques et Naturelles, Nouy. Pér., Tome 62 (Genéve, 1878), pp. 193-211. 508 the type used by Willis. WARREN J. MEAD A slight modification of the apparatus of Favre was used by Hans Schardt,’ who used various combina- tions of plastic and brittle layers in his studies of the mechanics Fic. 1.—Fractures produced by tension. A heavy sheet of rubber is held between two clamps and stretched by means of the screw at the end. It is then coated with paraffin which is allowed to chill until brittle, after which the rubber is further stretched by means of the screw, developing tension cracks in the paraffin. of mountain building. Still another modification of Favre’s work was employed by Stanislas Meunier,? who studied and de- scribed the fractures produced in a layer of partially set plaster on a contracting rubber sheet. These three investigaters con- fined their work to pure shorten- ing and paid no attention to stresses set up by tension, shear, or warping. EXPERIMENTAL RESULTS Fractures produced by ten- sion.—The apparatus (Fig. 1) consists of a frame with a rigidly attached clamp at one end and a movable clamp at the other which may be moved toward or away from the stationary clamp by means of a long screw. ‘To develop tension fractures the rubber sheet fastened at its edges in the two clamps is tightly stretched by means of the screw and then coated with paraffin which is allowed to chill until brittle. Tension is then applied by further stretching of the rubber by means of the screw. A typical set of tension fractures thus developed is shown in Figure r. t Hans Schardt, ‘Etudes géologiques sur le Pays-d’Enhaut Vaudois.” Troisiéme partie. A. Mécanisme des Dislocations. Chapitres xv—xvii. Planches VI-IX. Bulletin de la Soc. Vandoise des Sciences naturelles, Vol. XX, No. 90, 1884. 2 Stanislas Meunier, La Géologie Experimentale (Paris, 1899), p. 299. THE MECHANICS OF GEOLOGIC STRUCTURES 5°9 These are approximately at right angles to the direction of move- ment as is to be expected. They are open cracks perpendicular to the rubber sheet. mechanics involved would indi- cate. The paraffin, like rock, is less resistant to tensional stresses than to shearing stresses. It fails, therefore, by the development of breaks along planes which are perpendicular to the maximum stress. Fractures produced by com- pression.—The apparatus used for this purpose is the same as the one employed for tension. The compressive force is applied by releasing the screw and allowing the rubber to con- tract. This develops compres- sional stresses in the paraffin coat. The first breaks to ap- pear are small inclined thrust faults striking at right angles to the direction of shortening and dipping approximately 45° either way. (See Figures 2 and 3.) ‘These are followed by small number of vertical faults which strike at angles approxi- mately 45° to the direction of shortening (seen near the margin of the rubber sheet in Figure 2) and are due to the These results are what an analysis of the Fic. 2.—Fractures and faults developed by shortening or compression. ‘This is the same apparatus as shown in Figure tr. The heavy sheet of rubber is first tightly stretched by means of the screw, coated with paraffin which is made brittle by chilling, after which the rubber sheet is allowed to contract by means of the screw, thus producing compressional stresses in the paraffin. Figure 3 shows these compression fractures in detail. fact that lateral relief is afforded at these points by the ‘‘spread”’ of the rubber sheet. There also appear a number of vertical tension joints striking in the direction of shortening and apparently due also to the ‘‘spread”’ of the rubber. 510 WARREN J. MEAD An analysis of the mechanics involved leads to conclusions in accordance with the foregoing experimental results. Under simple compression, fracture takes place by breaks which develop in the planes of maximum shear.*. Therefore the paraffin should fracture along planes inclined at approximately 45° to the direction of the compressive force. The development of these inclined shear fractures requires actual displacement on the plane of fracture. Fic. 3.—Fractures and thrust faults produced by shortening or compression in the direction of the arrows. (See Fig. 2.) The white bands are fractures and thrust faults in the paraffin which strike at right angles to the direction of shortening and dip at angles of approximately 45° in either direction. This movement has a component parallel to the compressive force and also one at right angles to this force. Evidently, therefore, fractures can develop only in such an attitude as permits this movement to occur. In the central part of the rubber sheet the direction of easiest relief is upward or away from the surface of | the rubber and therefore we expect inclined fracture planes strik- ing at right angles to the direction of compression. Near the margin of the rubber sheet lateral relief is afforded, and we find ™C. K. Leith, Structural Geology, p. 16. THE MECHANICS OF GEOLOGIC STRUCTURES 511 vertical fracture planes striking at angles of approximately 45° to the direction of compressive force. This experiment in terms of earth movements is to be compared with a tangential shortening of an earth mass extending down into the zone of flowage, accompanied by side flowage or spread. This shortening is communicated to the rocks in the zone of fracture, resulting in inclined thrust faults striking normal to the direction IE pnp rer, cine ce emcee Na macaa tans at Bel Fic. 4.—Fractures and faults developed by shear or rotational stress. A heavy sheet of rubber is tightly stretched between the two clamps by means of the screw at the top and coated with a thin coat of paraffin which is made brittle by chilling. The parafiin-coated rubber sheet is then deformed by means of the screw at the left. The fractures developed by the shearing movement are shown in detail in Figure 5. of shortening, vertical shear faults striking at angles of approxi- mately 45° to the direction of shortening and vertical tension joints striking in the direction of shortening. _ Fractures produced by shear or rotational stress—The apparatus used for this purpose is shown in Figure 4. It has one clamp 512 WARREN J. MEAD which may be moved toward or away from the other by means of a screw. ‘The other clamp is mounted in a slide and by means of the second screw may be moved at right angles to the direction of movement of the first clamp. The sheet of rubber is tightly stretched between the two clamps, coated with paraffin, and deformed by means of the screw attached to the sliding clamp. This subjects the rubber sheet and its coat of paraffin to a shearing or rotational stress. The re- sulting fractures in the paraf- fin are shown in Figure 5. The direction of movement is indicated by the arrows and the amount of movement by the shape of the parallelogram which was, previous to defor- mation, rectangular. The first fractures to ap- pear in any one locality on the rubber sheet are usually tension cracks inclined about 45° to the direction of the shearing movement. These are at right angles to the di- ea z rection of maximum elonga- Fic. 5.—Fractures produced in paraffin tion and appear as vertical coat on rubber sheet by shearing. The open cracks. They are fol- arrows indicate the direction of move- lowed immediately by two ment and the shape of the figure shows Hooter eee sets of vertical faults with horizontal displacement, one set striking parallel to the direction of movement and the other parallel to the free edges of the rubber sheet. These represent two directions of non-distortion or two shear planes developed by the shearmg movement in which direction of relief is in the plane of the paraffin layer. Another set of faults, only two of which are shown in Figure 5, are thrust faults striking approxi- mately at right angles to the tension crack and inclined ap- proximately 45° dipping in either direction. These are due to THE MECHANICS OF GEOLOGIC STRUCTURES 523 compression in a direction at right angles to the direction of maximum elongation. The pattern of fractures developed in a series of experiments with varying thickness of paraffin coat is uniform in so far that the foregoing described set of fractures are always found. Their relative prominence, how- ever, varies with the thickness and brittleness of the paraffin coat. Fractures produced by tor- sional warping.—lt is difficult to form any estimate of the importance of this type of def- ormation, but it seems prob- able that torsional warping occurs and, therefore, that it merits consideration as one of the types of earth deformation. A warped surface may be considered as having been de- formed in two manners; namely, by the change in area which has engendered tensional or compressional stresses or both and by bending which has occa- sioned stresses characteristic of cross-bending. For purposes of the present analysis it appears best to consider these two phases of deformation inde- pendently. Changes in area due to war p- ing.—lf a sheet of rubber is held between two clamps, as in Figure 6, and subjected to tor- Fic. 6.—Fractures produced by tor- sional warping. A heavy sheet of rubber is tightly stretched between the two clamps by means of the screw, and coated with paraffin which is made brittle by chilling. Then the lower clamp is rotated by means of the handle at the lower end. This subjects the rubber sheet to tor- sional deformation and develops cracks in the paraffin. (See Fig. 7.) sion by turning one of the clamps (maintaining a constant distance between the clamps) the effect is to increase the area of the rubber 514 WARREN J. MEAD sheet. Plainly the center line of the rubber (the axis of torsion) remains unchanged in length, but the lateral margins are stretched, because the rotating of one clamp increases the distance between the ends of the two clamps. This stretching is maximum at the free edges of the rubber and decreases toward the center. The effect of - tension thus developed is illustrated in Figure 7, which shows the cracks developed in the paraffin layer as a result of torsional warping of the rubber. Distribution of the cracks shows plainly axis of i torsion dh | Dp Up rs aa Fic. 7.—Fractures produced in paraffin coat on rubber sheet by torsional warping. (See Fig. 6 and 10C.) the increase in the amount of tension as the margin is approached. The confluent nature of the cracks, resulting in a minimum num- ber of free ends, is an interesting feature. The rubber sheet may be so arranged that the free edges remain constant in length during torsion. If the rubber sheet thus mounted is subjected to torsion, there can be no change in the lateral margins. There is a change, however, along the center line because the turnable clamp approaches the other as it is turned, thus causing shortening or compression along the center line. This compression is maximum along the center line and decreases to zero at the edges. A paraffin coat on the rubber THE MECHANICS OF GEOLOGIC STRUCTURES 515 sheet thus deformed develops characteristic overthrust faults of the type shown in Figure 3, which are most numerous along the center line and decrease in abundance and displacement toward the margin. We have in these two types of deformation limiting cases, neither of which is probably realized under natural conditions. In the first case there is a net increase in area; in the second case a decrease in area. We may now consider an intermediate case in which the total area remains constant. This in terms of the 8 {e) iL T ' ‘ ' ' ' ' t ' ' H ‘ ' 1 ' ‘ ! ' ' ' ' ' ' i] ' ae Pp lo Fic. 8.—Illustrating changes in area due to torsional warping of a rectangular surface. In A the line O—P on the axis of torsion has remained constant in length resulting in a net increase in area from ABCD to A’OB’C’PD’. In B the lateral edges have remained constant in length and the result of warping is a decrease in area from ABCD to AO’BCP’D. In C the area has remained constant during warping resulting in elongation of the margins, shortening along the axis,and no change along the neutral lines E-N and F-M. rubber sheet would result in tensional stresses along the free margins and compressional stresses along the center line with a neutral zone of neither compression nor tension on either side of the center. These three cases are illustrated in Figure 8. In each case the rectangular area represents the outline of an unde- formed surface. The area with curved ends represents the surface 516 WARREN J. MEAD which has been deformed by warping and then flattened out for comparison with the original area. Cross-bending stresses resulting from torsional warping.—We have so far considered only the stresses resulting from change in area, and now turn to a consideration of the stresses due to warping or cross-bending. In Figure 9 a warped surface is represented in . isometric projection. In one set of diagonals the lines curve downward toward their centers forming a synclinal depression. In the other set of diagonals the lines curve upward toward their ~ centers and form an anticlinal elevation. DD ALLL SLL I LF Fic. 9.—An isometric representation of a warped surface If we consider a sheet of finite thickness to be thus deformed it is evident that on the upper surface compressional stresses will be developed at right angles to the axis of the synclinal depression and that tensional stresses will be developed at right angles to the axis of arching. At every point on the upper surface of this warped sheet there is a tensional stress and a compressional stress acting at right angles, each of them at an angle of 45° to the axis of torsion. On the lower surface it is evident that similar stress conditions exist but the tensional and compressional stresses are acting at angles of 90° to the similar stresses at the upper surface. These stresses are caused by cross-bending, and tensional condi- tions on one side mean compressional conditions on the opposite 517 THE MECHANICS OF GEOLOGIC STRUCTURES ‘euauousyd uorssaidurod oy} Jo spnyruseu pure uortsod ay} sayeoIpul q oul] papeys 24], ‘2 pue DP soul] papeys oy} Aq UMOYS SI SyovAI UOTsUa} Jo ad} BuNINseI YJ, "927 aswaasudsy Kup JUOJY SaITOF JULI[NSAL JO SOpNJLUSLUL BAT}L[OI PUL UOTIIIIP 9JROIpUl SMOIIW “g_{ PUL VW UI UMOYS sassorjs Jo s}uRyNsay “od "g Pur] ay} Aq poyRorput SI aavfuns 4amo] ay} UO podojadsgp SyIeII UOISUd} JO UOT}IEIIP eYy, “DP auUT| ayy Aq payeorpur st aapfuns 4aggn ay} UO SYIVIO UOISUd} JO MONOMIp BY, ‘927 assoasuv4y Kuy Fuojp sasseijs JO apnywuseu sayryas pue UOI}IAITP 9YVOIpUT SMOIIY “SuIdre [euoIsIo} Aq pasnvd Sulpuaq 0} enp “1dAxI v Jo aavfuns sadn ayy uo podojaaop sassayyg “gf "g 1% SoinyorIy UoIssordu0d Jo pue I puv PD je SYORIO UOISUd] SuI[Nser Jo opnytuseu dAtjeJaI pue uoToaIIp 94} 9} eoIpur Sour] pepeys AAvoy oY], *9227 aswaasun4) KUM FUuOJD Sasso1yS JO 9pNYUSeU dA1VLI9I PUL UOLJIEIIp a}voIpUL SMOITY (‘D8 “BLT 99S) “JURJSUOD poUreUII sey vorE [v}0} OY} ITY “SurdiemM [eUOoTsI0} 0} aNp vore ul SasueyD [eI0] Aq pasned sassa1jg “W—'Or “OI e) = Vv LiMop I an umop | da an ! : “Mop an UM op 518 ) WARREN J. MEAD side just as in the case of a simple beam under load, in which com- pression is developed on the concave side and tension on the convex side. Resolution of strésses due to change in area and to cross-bending— The stress condition at any point is the resultant of all of the stresses acting at that point. To determine conditions at a given point on the surface of a warped sheet it is necessary, therefore, to resolve the stresses due to cross-bending and the stresses due to areal change. In Figure 10A the stresses due to change of area caused by torsional warping are indicated in direction and relative Fic. 11.—Plaster of Paris positive of folds produced by compression or shortening magnitude by arrows. These represent the stresses occurring along any transverse line of the warped sheet. In Figure 1oB the stresses due to cross-bending occurring along any transverse line of the warped sheet are shown by arrows in a similar fashion. In Figure toC the resultants of the stresses in Figures roA and r1oB are shown. In Figure 1oC typical cracks developed by tension and compression are shown at a, b, andc. ‘The tension cracks a and c should begin at nearly go° with the edge at the margin, and gradu- ally curve to an angle of 45° at the neutral line. The curved line b indicates the strike of inclined thrust fractures which would be developed by compression. It is of interest to compare these curves with the tension cracks developed in paraffin on a rubber THE MECHANICS OF GEOLOGIC STRUCTURES 519 sheet, illustrated in Figures 6 and 7. In this experiment constant distance between the clamps was maintained and compression along the center line thus prevented. The tension cracks are evidently due to a combination of cross-bending and stretching and are in accord with the conclusions indicated in Figure roC. Depending on the relative intensity of the stresses due to change in area and those due to cross-bending, the curvature of the cracks would vary from the position in Figure 10A, due only to change in area, to the position shown in Fig- ure 10B, due only to cross- bending. Relative intensity of stresses due to cross-bending and those due to change im area.—With a given length of torsional axis the amount of cross-bending is a function of the angular displacement by torsion and is independent of the width of the warped sheet. The change in area (and therefore the tensional and compres- sional stresses), however, is a function of the width of the sheet as well as of the angle of torsion. It follows, there- fore, that in narrow strips Fic. 12.—Vertical view of reproduction deformed by torsion, cross- in plaster of Paris of folds produced by bending stresses may be domi- shearing deformation. The direction of movement is indicated by arrows and the nant while in wide areas a amount of deformation is shown by the small angle of torsion with a _ shape of the block. small amount of resultant cross-bending may develop relatively large tensional and com- pressional stresses. Cross-bending stresses also increase with the thickness of the individual beds. 520 WARREN J. MEAD The classical experiment of Daubrée,t in which he twisted a narrow strip of glass and obtained systematic sets of fractures at approximately 45° with the axis of torsion, is familiar to most students of structural geology. The fractures thus developed in the glass are evidently due to cross-bending. The writer has repeated the Daubrée experiment and found that one set of fractures in the glass developed as tension-cross-bending cracks on one sur- face and that the other set of fractures at right angles to the first developed as tension-cross-bending fractures on the opposite sur- face. Therefore, a brittle rock formation broken in the manner illustrated by the Daubrée experiment would show a conspicuous set of tension cracks at 45° to the axis of warping and the other set would not be apparent, as it would be developed from the under- side of the deformed rock stratum. In other words, if we look for a repetition of the Daubrée experiment in the field we should look for only one set of parallel tension cracks. Use of the apparatus in the study of folds —Previous experimen- tation in the reproduction of the structures of folded rocks in the laboratory has, so far as the writer is aware, been by means of the type of apparatus employed by Willis.? This type of appa- ratus with various modifications has been used by Hall, Lohest, Favre, Daubrée, Cadell, and others. In his investigation of the mechanics of Appalachian structure Willis used an apparatus of the piston or plunger type in which a series of wax layers of varying consistency were built up to resemble in their relative competence the rocks occurring in the Appalachian region. A load of shot was superposed to simulate the weight of overlying sediments and deformation was accomplished by forcing a piston or plunger against the end of the aggregate by means of a screw. Willis found that with flat-laying layers single folds were developed near the plunger and that repeated folds could be developed only when certain portions of the beds had an initial dip or when the first fold next to the plunger piled up material to such thickness and strength as to develop, as it were, an extension of the plunger which in turn caused a secondary fold in front of it. 1G. A. Daubrée, Etudes Synthetiques de Géologie Expérimentale, pp. 507-15. 2 Bailey Willis, ‘‘Mechanics of Appalachian Structure, Thirteenth Ann. Rept. U.S. Geol. Survey, Part 2 (1893), pp. 241-53. THE MECHANICS OF GEOLOGIC STRUCTURES 521 The apparatus of Willis permits deformation only by straight shortening or compression and does not afford means of studying the nature of folds developed by lateral shearing movements. It seems probable that the movements between great earth masses are in the nature of shears rather than simple straight-line com- pression. In other words, the application of a compressive force directly toward the point of maximum resistance would be less probable than the development of a couple which would cause what has been called a rotational stress. It does not appear to the writer that rocks have been folded by forces transmitted to them in a manner at all similar to the action of a piston against a more or less confined mass but that shortening of the earth’s crust has resulted from great compressive forces extending to some depth, and that the fracture and folded rocks within the zone of our observation have received from the rocks beneath, in a large measure, the force which deformed them. In other words, most of the faults or folds are the result of the riding or dragging of the upper layers by the underlying materials. The writer has attempted to apply to the study of folds the methods used in the study of fractures already described. The two pieces of apparatus shown in Figures 1 and 4 were employed but instead of the thin brittle coat of paraffin a very thin layer of plastic wax (made by mixing beeswax and Venice turpentine) was applied and over this, while the wax was still sticky, a sheet of tinfoil was carefully spread. When the rubber sheet was allowed to shorten, or was deformed by shear, the layer of wax and tinfoil developed a series of folds. It was found that a very thin sheet of rubber served the same purpose as the tinfoil. The purpose of the thin sheet of rubber or tinfoil is to supply a layer with a cer- tain small amount of competency. A layer of wax alone is entirely incompetent and follows the deformation of the rubber sheet without development of folds. A thin layer of tinfoil, sheet rubber, or waxed paper supplies the element of competency which results in the development of folds. Folds developed by pure shortening or compression.—The appa- ratus shown in Figure 2 was employed, with a layer of plastic wax covered by a sheet of thin dental rubber. The shortening was 522 WARREN J. MEAD produced by allowing the thick rubber sheet to contract. This resulted in the development of the system of folds illustrated in Figure 11. This figure represents a positive reproduction of the surface of the specimen in plaster of Paris. The original specimen was not of a nature to be easily photographed. This experiment and others of the same type show a set of folds striking in a direc- tion at right angles to the direction of shortening. All of these folds pitch at the ends and disappear. Depending on the thickness of the wax and the behavior of the rubber sheet, shortening is accomplished by a few large folds or by a larger number of smaller folds. An interesting overlapping of the folds is noted. When- ever a fold terminates by pitching, another fold appears over- lapping it and continuing the necessary amount of shortening. The experiments demonstrate very well that pitching folds do not necessarily mean cross-folding but that they are developed in flat-lying beds with perfectly even application of shortening stresses. ; Development of folds under conditions of shear or rotational stress.—The apparatus used is illustrated in Figure 4, except that in place of a thin layer of paraffin, a very thin layer of plastic wax was applied to the very tightly stretched rubber sheet and over this a sheet of tinfoil was carefully spread, care being taken to secure perfect adhesion of the tinfoil to the underlying wax. Deformation was accomplished by causing the slidable clamp to move parallel to the other clamp by means of the screw. This resulted in a set of folds in the tinfoil-covered wax layer, shown in Figure 12. The illustration is from a photograph of a posi- tive reproduction of the specimen in plaster of Paris. The direc- tion of shearing forces is shown by the arrows and the amount of deformation by the shape of the parallelogram which was originally rectangular. The folds, it will be noted, have their axes parallel to the direction of elongation‘of the mass. All of them are pitch- ing folds. They illustrate the phenomenon of repeated folds. Like the previously described experiment they demonstrate that pitching folds. do not necessarily mean cross-folding or shortening in the direction of the axes. As a matter of fact, in this experi- ment tensional stresses existed in the direction of the axes of the THE MECHANICS OF GEOLOGIC STRUCTURES 523 folds and actual open ruptures in the tinfoil occurred across some of the folds but were not preserved in the process of casting. A rather striking similarity between the folds illustrated in Figure 9 and the structure of the southern Appalachians is appar- ent and the writer ventures to suggest that perhaps certain of the characteristics of the southern Appalachian structure, such as repeated folds, pitching folds, and repeated thrust faults, may receive a certain amount of new light by viewing them with the conditions of the above-described experiment in mind. Whether the deformation was accomplished by straight compression between the oceanic segment and the continental mass or by a shearing movement between these great segments cannot be said. It seems, however, that the latter is rather more probable mechani- cally and that the foregoing experiment demonstrates that the folds of the Appalachians could have been produced by such shearing movement. PRELIMINARY DESCRIPTION OF A NEW SUBORDER OF PHYTOSAURIAN REPTILES WITH A DESCRIP- TION OF A NEW SPECIES OF PHYTOSAURUS 1B, (Ca (CANSIS, University of Michigan, Ann Arbor The Triassic beds exposed in a narrow strip along the eastern edge of the Staked Plains in western Texas have not been inten- sively studied and are now grouped together under the name given by Drake, the Dockum beds. Drake recognized an indefinite tripartite division of the beds which may later be distinguished as recognizable horizons. The vertebrate remains which have so far been recovered from these exposures are all members of the reptilian order Parasuchia and of the stereospondylus stegocephalia — and are indicative of an upper Triassic age. . Summary accounts of the reptilian forms previously described have been given by McGregor and Mehl. In the year 1917 the author, while on a hurried trip across the plains, found near the now little used crossing of the Blanco or Catfish River on the road from Spur, in Dickens County to Crosbytown, in Crosby County, a series of vertebrae, with portions of the dorsal armor, of a reptile which appeared to be of the usual phytosaurian type and were referred, in the museum catalogue, to Phytosaurus buceros Cope with question. In 1919 a chance was afforded to revisit the locality and considerably more of the same specimen was recovered. The material now in hand includes most of the vertebral column with exception of the posterior part of the tail, much of the dorsal armor and the skull, lacking the anterior end of the nose and the lower jaws. ‘The skeleton was found in a sandy clay mixed with abundant fragments of vegetation and the bones were coated with a tough layer of gypsum which has in places penetrated and rotted the bones and in other places left tJ. H. McGregor, Mem. Am. Mus. Nat. Hist., Vol. IX, Part XI, 1906; M. G. Mehl, Jour. Geol., Vol. XXIII (1015), p. 129. 524 NEW SUBORDER AND NEW SPECIES OF PHYTOSAURUS 525 them in singularly good condition. The bones were disturbed when deposited so that only a few of the cervical and of the dorsal vertebrae were found in series. That the bones are those of a single individual is indicated by the fact that they were found in a single mass and that no other bones were found, even after a careful search, within a half-mile of the small patch where they were found, but that there is some room for doubt is indicated by Ua 7 bo ie WY A he WAZ SP d eZ, mx Fic. 1.—Lateral view of the skull of Desmatosuchus spurensis, X% Fic. 2.—Upper view of the skull of Desmatosuchus spurensis, X+ the fact that it seems impossible to fit two of the isolated cervicals into the series and that there are more dorsals than are usually found in the Parasuchia. ‘These points are significant because the skull and the dorsal armor are so radically different from any form yet found while the vertebrae are typically Parasuchian. If, though it seems impossible at present, the remains are finally shown to belong to more than one individual, the skull must remain the type of the new sub-order and the vertebrae and dorsal armor must be described separately as anew form. 526 E.C. CASE The character of the skull is shown in Figures 1, 2, and a. It is astonishingly small for the size indicated by the vertebral column. ‘The basi-cranial region is distinctly phytosaurian in size and the arrangement of the elements but the skull as a whole shows several marked peculiarities. ‘There is a single large lateral temporal fenestra with no trace of an incipient or disappearing upper fenestra. The quadrate is so far reduced that it occupies a depression bounded by the opisthotic and the quadrato-jugal (?); the large, laterally directed orbits are preceded by large, elongate antorbital vacuities and these by smaller, elongate narial openings which lie entirely on the side of the nose; the teeth are small (as indicated by the sockets and a single poorly preserved tooth), are of equal size throughout the length of the maxillary, and were set in deep sockets; the posterior surface of the skull is completely closed except for the large foramen magnum and two small openings, amounting only to foramina, which occupy the position of the posterior temporal openings. Fic. 3.—Posterior view of the skull The anterior end of the skull is °% Desmatosuchus spurensis, X% missing but there is every indication that the nose was short. The close anchylosis of the bones and the coarse sculpture of the upper surface of the skull prevents the determination of many of the sutures; such as have been located are indicated upon the figures. As it has been shown by v. Huene and others that Aétosaurus possessed upper temporal openings and as such openings are present in the Proterosuchia, their absence in the form here described is alone sufficient to indicate its isolated position. The vertebral column so closely approaches that of the Para- suchia that it need not be discussed in a preliminary description. The dorsal armor consists of four rows of plates, two on each side of the median line. The rows are made up of incomplete rings covering the dorsal portion of the body. Whether there was any armor on the sides or the abdomen is uncertain; a single NEW SUBORDER AND NEW SPECIES OF PHYTOSAURUS 527 small plate of irregular form such as occurs on the lower part of the body of the Parasuchia was found with the specimen but its position is uncertain. The dorsal armor is most conveniently described as a series of transverse arches. ‘The first five are not metamerically arranged with relation to the vertebrae; they cover the first ten vertebrae, rather more than the whole cervical series; posterior to this the plates correspond in number and position with the vertebral seg- ments below. Lach arch consists of a pair of median plates with a low blunt spine or knob, and a pair of outer plates which carry spines of varying size and form. ‘The plates of the first five arches have only a slight sculpture, those posterior to the fifth have an increasingly heavy sculpture of pits and ridges which soon becomes very coarse. The form of the various plates is shown in Figure 4, which is a semi-diagrammatic restoration of the dorsal armor. Except for the three posterior dorsal arches and the median caudals there is evidence for all the parts. The arrangement of the plates is such as is dictated by the form and characters of the units, none having been found in sequence posterior to the cervical region. The first five arches show a gradual increase in the size of median plates and a rapid increase in the size of the spines upon the lateral plates. In the fourth arch the median and lateral plates of the right side are co-ossified and in the fifth arch the median plates of both sides and the lateral plate of the left side fit well together; in these arches there can be no doubt of the position and direction of the spines and by analogy with these the position of the other plates and spines is determinable. The fifth arch is most astonishing in the development of the relatively enormous spines which extend outward almost horizontally and curve forward over the third and fourth arches. In all of the arches of the cervical region the plates are thickened at the point of juncture in the same arch and thinned anteriorly and posteriorly to permit of the overlapping from before backward and allow for slight vertical movement of the head and neck; any lateral move- ment, as indicated by the form of the plates and the faces of the vertebral centra, must have been very slight. Fic. 4.—Restoration of the dorsal armor of Desmatosuchus spurensis, X s'4 NEW SUBORDER AND NEW SPECIES OF PHYTOSAURUS 529 Posterior to the fifth arch the median plates become shorter antero-posteriorly and elongate laterally to accommodate the broadening of the back. The lateral plates have a distinctly angular form, the inner half of the base shorter and nearly hori- zontal, the outer half longer and extending outward and down- ward at a slight angle. The spines are directed outward and slightly forward. In the pelvic region is placed a broader arch followed by a narrower one; this arrangement is only tentative, but the plates can hardly go elsewhere and it has been pointed out to the author by Professor Alexander Ruthven that such an arrangement of scales and scutes is not uncommon in living forms. Only the anterior portion of the tail is represented and only by plates of the lateral series. It is altogether probable that the median plates gradually disappeared as in the living Crocodilia. The lateral plates have well-developed spines and the outer half of the base is elongated and extends downward almost directly in the same plane as the outer side of the spine, indicating that the sides of the tail were elevated and flattened, not rounded. The length of the tail is uncertain. The exact position of this peculiar form is undeterminable as yet; its phytosaurian affinities are beyond question, but whether it is to be regarded as a highly specialized form of a more primitive type retaining characters of the primitive single-arched reptiles and placed near the origin of the Parasuchia, or a specialization of a more advanced type, remains to be determined after more com- plete preparation and the assemblage of the material in a mount. An attempt to obtain more material is in progress. For this new form, indicating the skull as the characteristic portion of the holotype, I would suggest the name Desmatosuchus spurensis, with the family and sub-order Desmatosuchidae and Desmatosuchia. ON A NEW SPECIES OF PHYIOSAUR FROM THE DOCKUM TRIASSIC BEDS OF TEXAS While on a trip through a portion of the Triassic beds of Texas in the summer of r919 the author was privileged to examine the collection of Mr. George D. Doughty, of Post City, Texas. In this 530 E. C. CASE collection is the posterior portion of the skull of a large Phytosaur preserved in a soft yellow clay. The bones are somewhat rotted and broken but, as is common in specimens preserved in such a matrix, the sculpture and the sutures are exquisitely preserved. Mr. Doughty was so kind as to loan this specimen to the author for study and illustration. As preserved and collected, the pos- terior part of the top of the skull and the left side are present as far forward as the nares and the posterior end of the antorbital opening. The lower part of the skull is lost*and the occipital region is represented by a detached fragment. The skull was distorted in fossilization so that the left temporal region is pushed ‘outward at a decided angle. Figures 5 and 6 show the sculpture of the bones and the position of the sutures so clearly that extended description is unnecessary. The lateral temporal opening is rhomboidal in outline with the long axis inclined obliquely forward and back. Its upper border is formed by the squamosal and postorbital. The squamosal is fairly flat on the upper surface and marked with a low relief of broad rugosities. The descending process on the outer side is smooth and widens toward its connection with the quadrato-jugal. The posterior end is widened and the inferior face is marked by two shallow, elongate, depressions, probably the location of cartilagi- nous attachment to the opisthotic. Just anterior to these there is a deep pit which received the head of the quadrate. There is a total lack of any descending, hooklike process posterior to the quadrate and defining the otic notch. The parietals are flat or slightly concave; the posterior ends are broken off but there is clear evidence that the posterior ends were below the level of the squamosals and that there was no elevated arcade defining the posterior border of the upper temporal opening. The quadrate was erect and its greatest breadth lay almost at right angles to the squamosal and the quadrato-jugal. Its anterior face is sharply concave, the inner border extends more forward than the outer and is very thin; apparently only the lower part of this inner edge united with the pterygoid. The quadrato-jugal is peculiar in its mode of articulation with the jugal and the squamosal; the anterior and upper edges are NEW SUBORDER AND NEW SPECIES OF PHYTOSAURUS 531 split and the lower edge of the squamosal and the posterior edge of the jugal are deeply dovetailed into the quadrato-jugal. This relation is clearly shown by the distinct and perfectly preserved sutures. From this peculiar relation it follows that the quadrato- jugal occupies very nearly as large an area on the inner side of the Fic. 5.—Upper view of the skull of Phytosaurus doughtyi, Xt temporal arch as on the outer, though it forms but a small portion of the posterior lower edge of the temporal opening. The attach- ment of the quadrato-jugal to the quadrate is by a broad, flat suture, interrupted by a good-sized quadrate foramen. The jugal extends forward in a nearly straight line, not rising sharply or even bending upward to form a notch below the lateral temporal opening. It is not certain, but possible, that an anterior 532 E. C. CASE extension of the jugal took part in the posterior edge of the ant- orbital opening. On the inner side of the anterior lower angle of the jugal are two articular faces which are separated by a deep groove; this when opposed to the maxillary would form a large foramen. hij / Lyf, Y WP ZZ ee a Fig. 6.—Lower view of the skull of Phytosaurus doughtyi, X% The orbits are nearly circular and are separated from the lateral temporal opening and the antorbital opening by strong bars. The anterior bar, formed by the lachrymal, is without sculpture but there is no indication of the opening extending backward in a depression on the surface of this bone. The nasals are nearly flat in their posterior part but rise sharply to the posterior edge of the nares, which open upon a considerable NEW SUBORDER AND NEW SPECIES OF PHYTOSAURUS 533 elevation. The nasals are short, not extending, apparently, anterior to the openings, and they do not separate the nares. The septum is formed by thin, paired plates which rise from below, as in most of the Phytosaurs; these are apparently the mesethmoids. The surface of the nasals, together with the surface of the frontals and parietals, are deeply sculptured. The nasal canals extend sharply backward as well as downward. On the under side of the specimen the walls of the brain case. are broken and lost from a line above the otic region. The bones which form the remnant of the brain case extend forward in sharp processes which lie in grooves on the lower surface of the frontals; these are probably the alisspenoids. The channel for the forward extension of the olfactory portion of the brain is well defined. At the point of junction of the postorbital, postfrontal, and parietal there are deep pits, transversely elongate, on either side. The function of these is obscure; they are perhaps connected with the orbitopineal process of the brain described by Cope in the cast of the brain cavity of Phytosaurus buceros.* MEASUREMENTS MM. Posterior edge of nares to posterior end of squamosal........ 406.8 Mamomsquamosal toibase of quadrate’, .5 2.2 .0...+-26-0065- 233.6 Wadthvacross lower face of quadrate......:.2...:2....-::-- 83 lintseronoitalespace wmarrowest on... c'4 sh acne oRlss ve ave stee ele 62 Center of foramen magnum to end of opisthotic............. 124.7 A consideration of this brief description and the accompanying figures will show that this specimen differs at least specifically, from any yet described, and the name Phytosaurus (Machaeroprosopus) doughty1 is proposed for it in honor of the discoverer. From the published figures of Phytosaurus kapffi given by v. Huene’? this specimen differs notably. The quadrate fits into a pit on the lower side of the squamosal and is supported above by the opisthotic and below by the pterygoid; the latter is attached to the lower half of the quadrate and does not appear on the posterior lateral face of the quadrate as suggested in Huene’s Figure 15. t American Naturalist, Vol. XXII, p. 914. 2 Geol. u. Paleontolog. Abhdig., N.F., Bd. X, 1891, Figs. 14-16. 534 | ES OR CASE The opisthotic and pterygoid join the quadrate by separate attach- ments in a manner quite different from that suggested by Huene, Figure 16. From this and all other described Phytosaurs it differs . in the absence of a descending hook from the squamosal posterior to the quadrate, outlining the otic notch. From Paleorhinus' it differs in the more posterior position of the nares, these lying over the antorbital opening as in the Phyto- saurus type rather than anterior to it as in the Mystriosuchus type. The nasals are shorter and do not separate the nares nor extend anterior to them. No otic foramen is present. The quadrate foramen is entirely on the posterior face of the skull. The pre-frontal is much shorter and lies almost entirely anterior to the frontal. The lachrymal is stouter and forms a broad bar between the orbit and the antorbital opening. There is no evidence of a depression extending to the lachrymal from the posterior edge of the antorbital foramen and the lachrymal had little or no connection with the maxillary. The lower edge of the lateral temporal opening is convex upward, not concave. From Angistorhinus it differs in the quadrangular rather than the oval form of the lateral temporal opening; in having the parietal-squamosal arcade depressed; in that the orbits are rounded instead of oval; in the position of the orbits which are more anterior in reference to the lateral temporal opening; in that the parietals meet anteriorly in a recess rather than in a point; and in that the opisthotic is more spatulate at the outer end. 5 From Machaeroprosopus it differs in the greater distance between the center of the orbit and the posterior end of the nares, 74mm. instead of 4omm. The nasals do not include nor extend in front of the nares. The frontals are longer and narrower; the jugals probably take little part in the posterior edge of the antorbital foramen. ‘The quadrate foramen issmall. The pterygoid is absent but it does not appear that there could have been any extension in the form of a wedge or hook between the quadrate and quadrato- jugal, as described and figured by Mehl. ‘ Bibliographies of the American Triassic Phytosaurs are given by McGregor, Mem. Am. Mus. Nat. Hist., Vol. IX (a906), Part XI; and Mehl, Jour. Geol., Vol. XXIII, No. 2, rots. NEW SUBORDER AND NEW SPECIES OF PHYTOSAURUS 535 From Phytosaurus buceros it differs in the more rounded and shorter posterior process of the squamosal. The lateral temporal opening is larger and more quadrangular in outline. The antorbital opening was more rounded at the posterior end. The jugal passes forward in a nearly straight line, not bending sharply upward, to form a notch in the lower border of the temporal opening. Mehl in his account of the skull of Machaeroprosopus' has indicated his belief that the form described by him is congeneric with Cope’s Phytosaurus buceros; but because he agreed with Jaekel that Cope’s form is a distinct genus and because Jaekel’s name, Metarhinus, is preoccupied, he suggested the new name Machaeroprosopus. It is far from certain that the European type of Phytosaurus does not occur in North America; Huene has rejected Jaekel’s distinction between Phytosaurus buceros and the European forms, and to the author it seems very doubtful on present evidence that such a distinction is justified. For this reason he prefers to refer the new species to the genus Phytosaurs (Machaeroprosopus ?) with Cope’s buceros and Mehl’s validus and gracilis. t Quart. Bull. Univ. Oklahoma, N.S., 103 (1916), p. 21. THE HEART? MOUNTAIN OVERTHRUST, WYOMING? D. F. HEWETT United States Geological Survey CONTENTS SUMMARY OF RESULTS INTRODUCTION SURFACE FEATURES STRATIGRAPHY McCutLiock PEAK EXPOSURES PHYSICAL FEATURES OF THE OVERTHRUST AGE OF THE OVERTHRUST TERTIARY DEFORMATION AND SEDIMENTATION IN THE BIGHORN BASIN ILLUSTRATIONS Fic. 1.—Sketch geologic map of region near Cody, showing Heart Moun- tain overthrust. Fic. 2.—Cross-sections of the Heart Mountain overthrust Fic. 3.—The McCullock Peak region viewed from the southwest, near Cody, Wyoming. Fic. 4.—Bighorn limestone (?) overlying Wasatch (or Bridger ?) beds on a hill near East Peak, McCullock Peak region, near Cody, Wyoming. SUMMARY OF RESULTS The recognition in 1919 of blocks of Madison limestone (Missis- sippian) overlying beds of the Bridger epoch (middle Eocene) in the McCullock Peak region, 12 miles east of Cody, Wyoming, shows that the overthrust fault recognized by Dake (2) in 1916 is much more extensive than first suspected. Dake mapped the fault in a belt 30 miles long, within which the extent of overthrust was estimated at 16 miles. He also noted the existence of thrust faults along the east front of Beartooth plateau, where pre-Cambrian rocks overlie ‘““Red Beds” (Chugwater formation) but did not assume continuity with the Heart Mountain overthrust. The residuals on McCullock Peak show that the extent of overthrust is at least 28 miles and indicate that the fault should be traceable t This spelling approved by United States Geographic Board, 1909. 2 Published by permission of the director, United States Geological Survey. 536 THE HEART MOUNTAIN OVERTHRUST, WYOMING 537 over the entire eastern edge of Absaroka Range, perhaps for 125 or 150 miles. On the basis of exposures in the mountain region, Dake was able to conclude that the overthrust took place after the deposition of his Fort Union (?) (early Eocene) beds and before the outbursts of the “early basic breccias (Neocene or upper Miocene) ”’ described by Hague. The McCullock Peak outliers show that the overthrust is probably post-Bridger (middle Eocene). It may be recalled that Richards and Mansfield (15a) concluded that the Bannock overthrust in southeastern Idaho was developed before the deposition of Wasatch beds (lower Eocene). Similarly Veatch (16a) concluded that the Absaroka overthrust in south- western Wyoming was developed after the deposition of the Almy and Fowkes beds (lower Wasatch) and before the deposition of the Knight formation (upper Wasatch, lower Eocene). On the north, Willis (17a) concluded from physiographic rather than stratigraphic evidence that the Lewis overthrust was developed in mid-Tertiary time, and was completed before the Miocene epoch. Although further work will undoubtedly determine more closely the periods at which these four overthrusts were developed, it appears highly probable at present that, although they lie in a belt scarcely 500 miles long, they did not take place simultaneously. The Lewis overthrust may have been nearly simultaneous with the Heart Mountain overthrust, however. A brief reconnaissance by the writer in the region west of Cody studied by Dake confirmed his conclusion that the overthrust beds were deeply eroded before the outbursts of the early basic brec- cias of the Yellowstone Park region (upper Miocene). Evidence obtained by the writer previously between Owl Creek and Wood River, however, indicates that there was conformable deposition of the Bighorn Basin Wasatch and the overlying tuffs and breccias of that region, which are here tentatively correlated with the “early acid breccias” of the Absaroka Range (lower Eocene). It is concluded that the overthrust took place after the deposition of the “early acid breccias”? and before the outbursts of ‘early basic breccias” and is, therefore, middle Eocene or early Oligocene in age. 538 D. F. HEWETT INTRODUCTION It is the principal purpose of the present article to set forth observations in the McCullock Peak region that bear upon the extent and age of the Heart Mountain overthrust. It will be apparent, however, that such a profound structural feature must consistently fit into the complex Tertiary history of sedimentation and orogenic movements in the region, so that it seems advisable to set forth here some of the problems of the region and some of the data that must be adjusted to a correct and comprehensive interpretation of that history. As the writer has devoted parts of the field seasons of 1911, 1912, 1913, 1916, and ro19 to detailed investigations in three fifteen-minute quadrangles that lie between Cody and Thermopolis, southeast of the region in which the over- thrust has been observed, some conclusions must be stated without giving much of the evidence on which they are based. SURFACE FEATURES The Bighorn Basin (4, 5, 14) is an elliptical area of low relief with few conspicuous hills or mountains, surrounded on the east, south, and west by mountain ranges. For purposes of physio- graphic, stratigraphic, and structural description, it may be con- sidered as made up of two parts, a central part 75 miles long by 45 miles wide, which largely coincides with the area of Wasatch and younger beds, and a border belt 10 to 20 miles wide, which lies between the central part and the mountain ranges that surround the basin. These ranges include the Bighorn Mountains on the east, the Bridger and Owl Creek ranges on the south, and the Absaroka Range and Bear-tooth Plateau on the west. Figure 1 shows some of the features and geology of an area in the north- western part of Bighorn Basin, which extend from the eastern edge of the Absaroka Range, across the border belt to the center of the basin. Figure 2 shows two sections across this region. The central part of the basin contains large areas of flat uplands that lie several hundred feet above the nearby valleys. There are also extensive areas of bad lands where erosion has cut back into the uplands. The central part contains three conspicuous elevated areas that rise above the uplands, and that range from 1,200 to THE HEART MOUNTAIN OVERTHRUST, WYOMING 539 x é = x) aS) Sd mH 9 S] iy 4 4 (Largely ay ph li: S=— cal y hy. ak jd y b A LT 4 Wy [4s _ i, SA LS ZA!) \ Fic. 1.—Sketch geologic map of region near Cody showing Heart Mountain overthrust. compiled from maps by C. L. Dake and C. A. Fisher.) 540 D. F. HEWETT 1,500 feet above the nearby streams; Squaw Buttes (6,200 feet), Tatman Mountain (5,800 feet), and McCullock Peaks,’ three in number and approximately equal in elevation (6,200 feet). Heart Mountain (8,080 feet) is a conspicuous peak that lies west of the central part of the basin, where it merges with the border belt. The border belt is largely made up of long stretches of flat upland terraces that rise gently toward the mountains. Most of the streams have cut broad terraced valleys below the uplands. The rocks exposed in this belt range from the Chugwater formation (“Red Beds’’) to the Fort Union formation (lower Eocene), and attain a maximum thickness of about 15,000 feet on the west side of the basin. The successive formations are brought to the surface in a series of pronounced folds whose axes are roughly parallel to a median trough in the central part of the basin. Dips that range from 15° to 60° are common along the flanks of the folds (14a). In a broad way, the mountains that limit the basin on the east, southeast, and south, the Bighorn (4), Bridger (4), and Owl Creek (3) ranges respectively, are rather simple smooth ridges that coincide with extensive anticlines. The mountains west of the basin (10) are high and rugged and present an imposing front toward the basin. They coincide roughly with an area of volcanic tuffs, breccias, and flows that are a part of the extensive field of vol- canic rocks which covers northwestern Wyoming and eastern Idaho. » STRATIGRAPHY It will be sufficient at this place to state briefly the general features of the Paleozoic and Mesozoic sections and such details of the Fort Union, Wasatch, and younger rocks as bear upon the age of the overthrust and the physical conditions surrounding the process. The commonest underlying pre-Cambrian rock in this region is a rather homogeneous red granite which is locally cut by diabase dikes. The Paleozoic and Mesozoic sections are separable into three groups on the basis of lithology and degree of induration, which measure their strength. The first and strongest group includes the Paleozoic limestones and associated sandstones and quartzites, Commonly referred to as McCullock Peak. 541 THE HEART MOUNTAIN OVERTHRUST, WYOMING —_— _ ———_ ————___ —___. SuWad HOOTIND sw JSNAYAVAO ULLJUNOT JARIFT VY} JO SuUOT]IaS-sSOIQ—'e “OTT aG-D ANIT SNOW NOLoaS ed, Di SS Rea ( “ZN 2araro wp 2 = ae ———————aae_—sd”” _ ~~S877IIFZIAO § JO FSP : iw uvaa> NLW cdaaHs a@-V ANIT SNOW NOILOAS NLW 3yWNSaLLivy 542 DE EWE le that include the beds from the Bighorn limestone (Ordovician) to the Embar group (Permian and Pennsylvanian), the sum of whose average thicknesses in this region is about 1,280 feet. Most of this thickness is beds of dense homogeneous gray limestone that range from 5 to 50 feet thick. ‘This group is to be regarded as the most competent unit in the entire section. The second group includes several distinct sandstone formations, such as the Chugwater (Triassic) about 750 feet thick, and Mesa- verde formation (Upper Cretaceous), which is largely sandstone and about 1,200 feet thick. Both of these formations include some shale. As the Chugwater formation overlies the Paleozoic lime- stones and sandstones, it would tend to increase the strength of those beds. The third and weakest group includes the remaining formations. The Deadwood formation (Cambrian) made up of shale and sand- stone is 700 feet thick. The Mesozoic and Tertiary formations (12, 14) that are made up of thin soft sandstones, locally con- glomerate, sandy shale, and shale include the Sundance, Morrison, Cloverly, Thermopolis, Mowry, Frontier, Cody, Meeteetse, Lance, Fort Union, and Wasatch formations, 13,600+ feet thick. It will be noted that the most competent unit is that which includes beds that range from the Bighorn limestone (Ordovician) to the Chugwater sandstones (Triassic) about 2,030 feet thick. If this section be compared with others of regions in which large overthrust faults have occurred, the thinness of the competent part of the section is impressive. The characteristics and structural relations of those beds only that may bear upon the period of overthrusting will be presented here. The recognition of the Fort Union formation on the west side of Bighorn Basin is based upon numerous collections of leaves that have been studied and identified by F. H. Knowlton. The base is considered to be a conglomeratic sandstone, which is locally unconformable on underlying beds that range from the middle part of the Meeteetse formation (roughly equivalent to the Judith River formation of the Montana group) to the top of the Lance formation (“‘Ceratops beds’’). This unconformity has only been recognized in one locality east of Oregon Basin for a distance of THE HEART MOUNTAIN OVERTHRUST, WYOMING 543 15 miles where a maximum of about 2,000 feet of beds were here eroded before the deposition of the lowest Fort Union beds. The contact of these beds with the underlying Lance formation in a belt 50 miles southeast, however, yields no evidence of unconformity. The unconformity indicates local warping and erosion of the Lance and older beds before the deposition of the lowest Fort Union sandstone. . The top of the Fort Union formation in this region is a persistent unconformity at the base of beds that yield a large mammalian fauna until recently called Wasatch but now known to be charac- teristic of the Wind River formation (8, 9). The unconformity is readily recognizable at every locality where dip cross-sections of the beds may be seen, but in strike sec- tions it can only be detected by ‘close study of the lithological features. Within these limits, the Fort Union formation attains a maxi- mum thickness of more than 5,250 feet and is made up of many beds of pale yellowish buff and white sandstone alternating with gray, olive, and red shale. The sandstones of the lower 200 feet commonly contain lenses of pebbles of many rock types. Black and gray chert predominate but red and gray quartzite, pale pink porphyry, gray sandstone, and silicified wood are common. Pink granite and coal pebbles are sparingly present but limestone has never been found. The chert pebbles have yielded an interesting collection of invertebrate fossils which are characteristic of the Madison and Embar formations. At least three coal beds occur in the lower 600 feet of the formation on the west side of the basin. Thus far, in this region, the formation has yielded a single verte- brate bone, but no invertebrate fossils. No bentonite or volcanic ash have been recognized in it. There are good reasons for believing that the beds here con- sidered as the Fort Union formation are part of an extensive sheet of sediments spread over a large area of Wyoming and Montana, at least as far west as the Rocky Mountains, and eastward into the Dakotas. They probably covered the site of the present Bighorn, Bridger, and Owl Creek Ranges. These beds are so involved in the folds of the border belt of the basin that it is concluded that these vw D. F. HEWETT as well as the long anticlines which coincide with the Bighorn, Bridger, and Owl Creek ranges, were developed after the beds were laid down. The folds are broken by many normal faults of small displacement, none of which appear to pass into the overlying Wasatch beds. The term Bighorn Basin Wasatch has long been considered to include the sandstones and alternating olive and red clays of the central part of the Bighorn Basin, where they attain a maximum thickness of about 2,500 feet. These beds include light brown and white sandstones, gritty arkose and pale olive, gray, and red clays. The sandstones locally contain pebble zones of re-worked Fort Union materials, with the addition of limestone and granite which are absent or uncommon in those beds. Although bentonite has been reported east of Meeteetse (5), no unaltered volcanic tuff has yet been recognized. The beds are nearly horizontal over large areas in the center of the basin, and although the range along the border is commonly 3° to 10°, dips as high as 20° are known (09). The beds of the central part yield a large vertebrate fauna which has been studied from time to time. Only a few invertebrate fossils are known in the beds (5). The recent careful faunal studies of Granger and Sinclair show that the Bighorn Basin Wasatch contains beds that range from ‘‘ Paleocene” (their Clarks Fork beds) to uppermost lower Eocene (their Lysite or Upper Wind River beds) (7, 8, 9). The early work of Eldridge (6) as well as the later work of Fisher (5) showed the presence of a persistent nearly horizontal layer, composed of sandstone and gray and red shale underlying volcanic tuffs in the region between Meeteetse Creek and Owl Creek in the southwest part of Bighorn Basin, and they were con- sidered to be Wasatch. These beds outcrop along the south edge of the Meeteetse quadrangle and the west edge of the Grass Creek quadrangle which have been studied by the writer. In addition to the alternating olive and red shale and sandstones with local chert and quartzite pebble lenses, which are characteristic of the Wasatch deposits of the central part of the basin, there are thin beds of dark carbonaceous shale, carbonized plant remains, and THE HEART MOUNTAIN OVERTHRUST, WYOMING 545 thin lenses of coal. The thickness of the layer ranges from 125 to 250 feet. It lies on a surface of low relief cut across the folded Cretaceous rocks, and as the base rises from an elevation of 6,000 feet near Owl Creek to 7,200 feet near Wood River, a distance of 21 miles, it appears to have been slightly warped since deposition. North of Cottonwood Creek a narrow east-west strip has been down-faulted about 360 feet. Throughout this region, these beds are apparently conformably overlain by paper-thin carbonaceous shales that weather white, and these, in turn, by pale greenish volcanic ash and light brown tuff, locally indurated. About too feet above the typical Wasatch sediments, the well-stratified fine material is succeeded by coarser, cross-bedded light brown tuff and still higher by heterogeneous fine and coarse brown andesitic breccia that makes up the masses 3,000 to 4,000 feet thick in the region east of the Washakie Needles. Except for the reference to lava flows, which are not known between Wood River and Owl Creek, the following statement by Black- welder might be considered an accurate description of conditions in the southwest part of Bighorn Basin (1a): At the northwest end of the Wind River Range, where it articulates with the mountains of Yellowstone Park, thick beds of volcanic ash and agglomerate with interbedded glassy lava flows rest upon the pre-Tertiary folded rocks, but are themselves younger than the Wind River Eocene. Traced eastward to Horse Creek, the Washakee Needles, and the valley of Owl Creek, this thick volcanic series is found to rest conformably upon the striped clays of the Wind River formation, with which they intergrade through gray, plant- bearing shales and greenish volcanic sandstones containing petrified logs. A closer examination of the volcanic beds shows that some of them are massive agglomerates, devoid of stratification, whereas other beds are distinctly stratified, cross-bedded, and occasionally interrupted by lenticular sheets of coarse gravel, suggestive of stream channels. The conditions indicated are those which would be found upon low gradient river plains adjacent to active volcanoes. The stratified tuffs have yielded several collections of leaves, one collected by Mr. N. H. Darton (3a) on the Middle Fork of Owl Creek and one by the writer south of Sunshine. Both are con- sidered characteristic Fort Union material by F. H. Knowlton. No vertebrate or invertebrate fossils have yet been found in the supposed Wasatch beds, or the overlying tuffs or breccias. 546 D. F. HEWETT The relations between the beds described above, which may be referred to as the border Wasatch, to the Bighorn Basin Wasatch might be uncertain if it were not for the existence of remnants on hilltops in several parts of the border belt. ‘These remnants have the same lithologic features as the border Wasatch and some of the beds rather high in the section of the Bighorn Basin Wasatch in the Tatman Mountain section. A longitudinal section through three of these remnants from a point south of Sunshine northeast toward the basin shows that they are part of a river channel which extended from the foothills of the border belt out into the basin with a gradient of 25 to 70 feet to the mile. It is the writer’s opinion that the border Wasatch was laid down at the same time as some of the beds that make up the Lost Cabin or Tatman Mountain beds (upper Wind River) of Sinclair and Granger in the central part of Bighorn Basin. The particular significance of the border Wasatch beds is that they indicate that there has been little if any local folding of the Cretaceous and Fort Union beds of the border belt since upper Wind River beds were laid down. The border Wasatch has, however. been broadly warped and locally faulted. Only tentative conclusions concerning the age and correlation of the stratified tuffs and breccias that overlie the border Wasatch can be made at this time. In considering this problem the writer has had the benefit of informal discussion with Dr. J. P. Iddings, who examined a large part of the area of the Yellowstone Park (11) and that covered by the Absaroka folio (10). It may be recalled that along the headwaters of Lamar River (Cache Creek) and Clark’s Fork (Republic Creek) several varieties of light colored andesitic tuffs and breccias, locally distinctly stratified, underlie the darker, basaltic breccias that cover a large area east of Yellowstone River and Lake, and north of the latitude of Greybull River (10). Although locally, the lower group (“early acid breccias”) appears to merge upward with the upper group (“early basic breccias’’), elsewhere there is evidence of considerable erosion of the lower group before the deposition of the upper group. The lower group yielded a large flora, considered by F. H. Knowlton to be Fort Union (lower Eocene), whereas the flora of the upper group, likewise large, is considered to be upper Miocene. From THE HEART MOUNTAIN OVERTHRUST, WYOMING 547 the evidence of this region, one must conclude either (1) that the period of erosion intervening between the deposition of the two groups of breccias was brief and therefore that the floras are mis- leading, or (2) that the intervening period of erosion was rather long, probably persisting from upper Eocene time through Oligocene and lower Miocene time, as the floras indicate. Dr. Iddings agrees with the writer that the latter conclusion is more acceptable at present. He further considers that the lithology and flora of the tuffs and breccias between Owl Creek and Wind River warrant the tentative correlation of them with the “early acid breccias”’ of the Yellowstone Park region. It is apparent, however, that considerable additional work must be done before the relations of the volcanic rocks of the southern Absaroka Range are proved. MCCULLOCK PEAK EXPOSURES The foregoing statement of the lithologic features and structure of the Fort Union and Wasatch beds of the Basin permit a more careful consideration of the exposures near McCullock Peak and the relations of the overthrust recognized by Dake west of Cody. In previous descriptions of this region, the name “*McCullock Peak”’ has been applied to the central part of a high rugged area, about 4 miles long by 2 miles wide, 12 miles east of Cody. (Figure 3 shows a view of the McCullock Peak area from the west.) In the following description of the region, three culminating summits will be referred to as West Peak, Middle Peak, and East Peak, | respectively. Although extensive gravel-covered terraces extend south from the peaks for several miles, the entire elevated region is completely surrounded by typical bad lands, and the rock exposures are uncommonly good. West Peak lies about two miles northwest of Middle Peak and is connected with it by a sinuous rugged ridge. It is the culminating point of a rugged bad lands area carved from nearly horizontal Wasatch beds. It is almost devoid of vegetation and the terraced ridges of alternating gray, olive, and red shale and sandstone present an impressive picture from the east, north, and west. It was not visited by the writer, but several attempts to ascend it by horseback from the west are known to have failed. 548 D. F. HEWETT Middle Peak may be readily ascended from the south along a gravel-covered terrace of gentle gradient. The summit is smooth and the adjacent slopes, except on the north and west, are covered with grass and dwarfed sage-brush. If viewed from the west, the summit appears to be a part of the terrace which extends south- ward, but if critically viewed from the south it appears to be a smooth hill that projects about 200 feet above the extension of Fic. 3.—The McCullock Peak region viewed from the southwest, near Cody, Wyoming. West Peak lies on the left; East Peak rises above the terrace on the right; Middle Peak is the low cone just left of East Peak. the terrace. In other words, it appears to be the residual mass of a more imposing hill that was in existence while the plain of which the terrace is a remnant was being formed. Good rock outcrops are confined to the west, north, and northeast slopes of Middle Peak, where nearly horizontal sandstones, locally arkosic, alternate with pale olive and red clays. The material is similar to that which makes up the bulk of the Bighorn Basin Wasatch. The THE HEART MOUNTAIN OVERTHRUST, WYOMING 549 fossils which are described below were collected from a 50-foot zone of gray, olive, and red clays that are well exposed along a narrow ridge 1,000 feet northeast and 150 feet stratigraphically below the summit. The fossils were first referred to J. W. Gidley of the United States National Museum, but, as field parties from the American Museum of Natural History have closely studied and made numer- ous collections from these beds, it seemed advisable to refer the fossils to them also. The statements of Mr. Gidley (a) and of Mr. Walter Granger (0b) are attached: a) A preliminary report was furnished to Dr. Stanton on November 25 (unofficially). It was as follows: “No. 1. The largest tooth is a right upper third or last molar of Helaletes; cf. manus Marsh. Not known outside the Bridger horizon. “No. 2. The next smaller tooth is a last left upper molar of Eohippus sp. “No. 3. The fragment of jaw containing two teeth I have not been able to definitely determine.”’ In addition, I may now say that a further study of the small jaw fragment seems to warrant referring it to Hemiscodon pucillus Marsh, with which it agrees almost exactly in size. This determination, however, cannot be made positive without comparing it with the type, which is probably in the Yale Museum collection. This is a Bridger species. It would thus seem that the three specimens represent a Bridger fauna, although the Eohippus tooth suggests Wasatch rather than Bridger affinities. b) I have examined the three specimens of mammal teeth from Eocene beds near the top of McCullock Peak, Wyoming, and submit the following determinations: (1) Helaletes, a last upper molar of the right side. (2) Eohippus, a last upper molar of the left side. (3) Tetonius or Absarokius, a fragment of the right mandible, containing the first and second molars. The Lophiodont genus Helaletes has hitherto been recorded only from the Bridger, but it might reasonably be expected from the uppermost levels of the Lower Eocene since another Bridger perissodactyl, Hyrachyus, has recently been found in the upper horizon of the Wind River (Lost Cabin beds). The Hyracothere tooth shows characters most closely approached by the smaller specimens of Eohippus from the upper Wind River, as well as by specimens from the upper Huerfano, which is now regarded as the probable equivalent of the almost barren Bridger A. The tiny jaw fragment of the Tarsiid cannot definitely be assigned to either of the genera mentioned because it lacks the diagnostic front teeth. 550 DE ED EVV GEHL Tetonius, the type species of which is Anaptomorphus homunculus Cope, is recorded from the Gray Bull and Lysite horizons of the Bighorn and Wind River basins respectively, while the closely related Absarokius is from the Lysite and Lost Cabin horizons of the Wind River. The jaw appears to be of a new species, somewhat more progressive than any described form of either genus. These three specimens seem to represent a fauna intermediate between that of the upper Wind River and that of Bridger B. It may belong to the base of the Bridger (Hor. A), the mammalian fauna of which is practically unknown, correlation with the upper Huerfano being made on a single specimen of Titanothere. In any event the McCullock Peak horizon is close to the border line between the Lower and Middle Eocene. Dr. W. D. Matthew, who has also examined these teeth, concurs in the above identifications and in the conclusions regarding the age of the beds in which they were found. Blocks, bowlders, and smaller fragments of dense cream to buff limestone are found on the summit of Middle Peak, as well as on the ridge that extends south and in the ravines cut below it. In an area on the summit about 600 feet square, there are no less than twenty blocks more than 3 feet in maximum dimension, and the largest is 5510 feet. Most of the blocks are rudely rectangular and appear to be bounded by bedding planes and joints, although the surfaces are pitted and grooved by the solvent action of water. The blocks are irregularly distributed and there can be no doubt that they are not in place. The size, shape, and composition of the blocks as well as their distribution, are similar to those described by Granger and Sinclair (9@), to which a glacial origin was ascribed. Fossils collected from one of the blocks have been examined by Dr. George H. Girty, who reports that the following species are present: Cliothyridina, crassicardinalis, Eumetria Verneuliana, Schuchertella aff. Chemungenis, Spirifer centronatus, Triplophyllum sp. Dr. Girty states that this is a characteristic Madison fauna (Mississippian). This collection may be compared with the following collection, also identified by Dr. Girty as belonging in the Madison lime- tone. It was obtained about 50 feet above the southwest base of the block of limestone that forms Chalk Mountain, west of Cody, where it overlies beds of Cretaceous age: Triplophyllum ex- cavatum, Schuchertella aff. Chemungensis, Camarotoechia metallica (2), THE HEART MOUNTAIN OVERTHRUST, WYOMING 551 Spirtferina soliderostris, Cliothyridina crassicardinalis, Eumetria Verneuliana, Platyceras sp. ; South Peak lies about 8,000 feet southeast of Middle Peak and is formed by the conjunction of three smooth ridges that extend northwest, southwest, and east, respectively. The space between these two peaks is a smooth, rolling flat covered with sagebrush, and no rock outcrops were recognized in it. Just below the summit of this peak and well distributed around it there are six rugged outcrops of cream-colored limestone, each from 50 to 75 feet long. The limestone outcrops do not exhibit bedding but are highly shattered and several are entirely made up of angular fragments of limestone, one to ten inches in diameter, which are imbedded in a fine sand of similar material. The only fossils noted in these outcrops are a few crinoid stems, but the texture resembles that of the Madison limestone blocks on Middle Peak. The best evidence of the character of the beds under the lime- stone is obtained at the head of a ravine on the northeast side of the peak, where the smooth surface near the summit merges with ' bad lands that show horizontal Wasatch material. No fossils were found at this locality. The evidence at South Peak indicates that a triangular cap of crushed Madison limestone, about 600 feet long and 400 feet across the base and about 80 feet thick, overlies horizontal Wasatch beds. The summit of South Peak lies at the northwest end of an area about 3,000 feet in diameter within which there are eleven smaller and lower hills each of which shows outcrops of limestone breccia more than 50 feet long. Figure 4 shows one of the largest of these outcrops, in which the bedding is still preserved. It is 200 feet long, about 25 feet thick and, although it yielded no fossils, the texture indicates that it is probably part of the Bighorn limestone (Ordovician). Farther southeast these low hills merge with gravel- covered flat terraces that extend to the east and southeast toward the center of Bighorn Basin. Near the center of the hills, capped by limestone, there is a flat depression 500 feet in diameter at the northwest edge of which is Markham Spring, at an elevation of 5,950 feet. Even at the end of August, in the dry season of rg19, it yielded about two gallons a minute of clear non-alkaline water. 552 D. F. HEWETT Another similar spring lies about 3,000 feet southwest at a lower elevation. ‘These springs appear to be supplied from a thin mantle of mingled limestone breccia and sand from the Wasatch beds that becomes sufficiently saturated with water during the wet season to yield a small flow throughout the year. No other perennial springs are known within ten miles. VS SABO IR Fic. 4——Bighorn limestone (?) overlying Wasatch (or Bridger?) beds on a hill near East Peak, McCullock Peak region, near Cody, Wyoming. The significance of the McCullock Peak exposures may be briefly summarized. Before Dake’s work was done in 1916, the McCullock Peak exposures might have been as puzzling as Heart Mountain to the geologists who examined that region some years ago (5, 6), and who considered that the block of Madison limestone which forms its summit was a plug bounded by a circular fault. In the light of Dake’s work, there can be little doubt that the blocks of limestone in the vicinity of South Peak are parts of a layer of limestone that was thrust from the west, probably from the region THE HEART MOUNTAIN OVERTHRUST, WYOMING 553 near Sheep Mountain between the north and south forks of Sho- shone River, 28 miles west. It is impossible to imagine that the McCullock Peaks blocks are somehow related to another over- thrust or to some obscure and complex structure, for the structure of the beds between the peaks and Sheep Mountain is completely shown in the canyon of Shoshone River (14). The recognition of mammalian fossils characteristic of the Bridger formation of western Wyoming under the blocks of Madison limestone shows conclusively that the period of overthrust was no older than these beds. PHYSICAL FEATURES OF THE OVERTHRUST Sufficient information is not yet available to make a compre- hensive statement concerning those features of the overthrust that are needed to interpret the conditions under which it developed. It may be that the thick masses of “‘early basic breccias”’ cover so much of the mountainward portion that a satisfactory explanation can never be given. The known exposures permit the following summary of its features: 1. The lithology of the beds and few fossils collected in the mountainward area indicate that the base of the overthrust is uniformly the base of the Madison limestone. Dake states that the base of the Heart Mountain block is Madison limestone. On the other hand, although the few bowlders on Middle Peak yield Madison fossils, the lithologic features of the East Peak remnants resemble the Bighorn limestone. The base of the overthrust is, therefore, probably not the same horizon throughout. A limestone breccia is present at the base of all of the blocks that were examined. In the few days available in the mountain region, the writer was unable to confirm Dake’s conclusion that there are two surfaces of overthrust. Although the presence of Sundance fossils under the Chalk Mountain Block appears to be evidence that such is ithe case, the exposures along South Fork, where a lower surface s also mapped, do not appear to demand this interpretation. 2. The elevation of the base of the remnants of the overthrust block decreases from about 7,200 feet on Carter Mountain to, about 6,800 feet on Sheep Mountain, then increases to about 554 D. F. HEWETT 9,500 feet on the divide between Rattlesnake and Dead Indian Creeks. The base of the Heart Mountain remnant stands at about 7,200 feet and the McCullock Peak remnants from 6,100 to 6,200 feet. Rattlesnake Mountain, a persistent ridge which attains an elevation of 9,100 feet and lies between these two groups of rem- nants, was examined but does not appear to retain any remnants of the overthrust block, although it once overlay the mountain. It cannot be stated assuredly yet whether the differences in elevation here indicated represent the form of the original surface of over- thrust or whether the surface has been subsequently warped. Although parts of the region have been warped since the deposition of the bedded tuff near Owl Creek, tentatively correlated with the “early acid breccias” (lower Eocene), the writer believes that a large part of the noted differences represent the form of the surface over which the block was thrust. 3. The structure and attitude of the remnants of the over- thrust block indicate that it was a relatively simple, unfolded layer of rock. The thickness can only be conjectured. The entire Paleozoic, Mesozoic, and Fort Union sections above the Deadwood shale (Cambrian) are about 17,000 feet thick. In attempting to estimate the probable thickness of the block, it must be borne in mind that it was largely removed by erosion before the outburst of “early basic breccias” (upper Miocene). The writer would tentatively estimate the thickness near the mountains at 15,000 feet, but the eastern edge was probably much thinner. 4. The surface upon which the overthrust moved is cut across a sharply folded belt of rocks that range from the pre-Cambrian granite to beds that appear to represent horizon A of the Bridger formation. No part of this surface that has been studied appears to be a fracture across the beds, but it is probably a surface of erosion. To this extent it resembles Willis’ interpretation of the Lewis overthrust (17). 5. Dake mapped the overthrust in an area thirty miles long and the McCullock Peak exposures indicate a minimum thrust of twenty-eight miles. Only casual consideration of the regional geology is needed to convince one that the overthrust must have involved a much larger area, over which it can probably be traced. THE HEART MOUNTAIN OVERTHRUST, WYOMING 555 6. The relations of the border Wasatch and the residuals east of Sunshine indicate that the border belt of folds had assumed: their present form and were truncated by erosion before these beds were laid down. Apparently, therefore, the folds were devel- oped considerably earlier than the overthrust. AGE OF THE OVERTHRUST As a result of his work Dake concluded that the overthrust was younger than certain beds along the north and south fork of Sho- shone River, tentatively correlated with his Fort Union (?) forma- tion. He also concluded that the overthrust was older than the “early basic breccias.” . The short time available to the writer in this region in 1919 only permitted the following conclusions: 1. The beds exposed along the north fork of Shoshone River under the “‘early basic breccias”’ (13), and locally under remnants of the overthrust block resemble lithologically those 1,000 to 1,500 feet above the base of the Fort Union formation in the Bighorn Basin and lack the arkoses which are rather characteristic of the Bighorn Basin Wasatch. On the other hand they yield collections of leaves, among which F. H. Knowlton has recognized ““Aralia notata Lesq.’’ which though considered to be a Fort Union species, has not yet been recognized in the Bighorn Basin Fort Union, but is present in most of the collections from the “early acid breccias”? in Yellowstone Park and the tuffs north of Owl Creek. The beds yield numerous bone fragments, largely turtle skutes, not yet recognized in the Bighorn Basin Fort Union, but common in the Bighorn Basin Wasatch. Although existing evidence is conflicting, the writer is inclined to consider that the beds are to be correlated with the border Wasatch. 2. The overthrust is older than the ‘early basic breccias”’ because the breccias locally lie in channels cut 200 to 300 feet below the overthrust surface. 3. The period of erosion that followed the overthrust and - preceded the deposition of the ‘‘ early basic breccias”’ was sufficiently long to destroy most of the beds that made up the thrust block, for the thickest remnant, that which makes up the summit of 556 D. F. HEWETT Sheep Mountain, is only 1,000 feet thick and there are large areas over which the block is completely removed and ‘early basic breccias”’ rest on Wasatch beds (?). TERTIARY DEFORMATION AND SEDIMENTATION IN THE BIGHORN BASIN The table opposite p. 556 is presented at this time in the hope that it may aid in a better understanding of the relation of the Heart Mountain overthrust to the other deformations of the Bighorn Basin. It is not considered necessary to present in this paper any more of the evidence on which the conclusions are based. REFERENCES 1. Blackwelder, E. ‘‘Post-Cretaceous History of the Mountains of Central western Wyoming,” Jour. Geol., XXIII (1915), pp. 97-117, 193-217, 307-40; (a) p. IIo. : 2. Dake, C.L. ‘“‘The Heart Mountain Overthrust and Associated Structures in Park County, Wyoming,” Jour. Geol., XXV (1918), pp. 45-55- 3. Darton, N. H. ‘‘Geology of the Owl Creek Mountains, Wyoming,” Senate Document No. 119, soth Congress, Ist Session (1906), 47 pp.; (a) p. 25. “Geology of the Bighorn Mountains,” U.S. Geol. Survey, Prof. Paper No. 51 (1906). 5. Eldridge, G.H. ‘‘A Geological Reconnaissance in Northwest Wyoming,”’ U.S. Geol. Survey, Bull. No. 119 (1894). 6. Fisher, C. A. ‘“‘Geology and Water Resources of the Bighorn Basin,” U.S. Geol. Survey, Prof. Paper No. 53 (1906). 7. Granger, W. “On the Names of the Lower Eocene Faunal Horizons of Wyoming and New Mexico,” Amer. Mus. Nat. Hist., Bull. No. 33 (1914), pp. 201-7. 8. Granger, W., and Sinclair, W. J. “Eocene and Oligocene of the Wind River and Bighorn Basins,” Amer. Mus. Nat. Hist., Bull. 30 (1911), pp. 83-117. 9. ———. “Notes on the Tertiary deposits of the Bighorn Basin,” Amer. Mus. Nat. Hist., Bull. 32 (1912), pp. 57-67; (a) pp. 64-66. to. Hague, A. ‘Absaroka Folio” (No. 52), Geol. Atlas, U.S. Geol. Survey (1899). . 11. Hague, A., Iddings, J. P., Weed, W. H., and Others. “Geology of the Yellowstone National Park,’’ U.S. Geol. Survey, Mon. No. 32, Part II (1899). 12. Hewett, D. F. ‘The Shoshone River Section, Wyoming,” U.S. Geol. Survey, Bull. No. 541¢ (1014). PRELIMINARY CORRELATION OF TERTIARY DEFORMATION AND SEDIMENTATION IN BIGHORN BASIN GroLocic AGE ABSAROKA-OWL CREEK MountTAIN FRONT BiGHoRN BASIN Border Belt Central Part Deposition or Erosion Deformation Deposition or Erosion Deformation Deposition or Erosion Deformation 5 Early basic ‘a Basic tuff ( ?) (?) = | Upper breccias (local) Q (widespread) Ss | | Lower Erosion Erosion (?) Erosion (?) vo 5 8, Erosion Erosion Erosion (?) 3 Upper ORION Normal faults Birpsion Normal faults Erosion! (?) (local) 5 (local) Heart —(Unconformity—|— Heart —(Unconformity Heart (2) Mountain — (2) ) Mountain ?)) Mountain overthrust’ overthrust , overthrust py |) ible Ray acg Andesitic tuffs Andesitic tufls (?) g (local). (local) (local) a (Conformity) — —(Conformity) -(?) - Border Wa- Border Wa- ig Tatman Mtn. oA satch (local) satch (local) aie |eebeds 28 = Q'S | Lost Cabin beds |S © &) 7 ocal depres- Lower Erosion Erosion © § 4 Lysite beds “aia ll © ston P 3 = Gray Bullbeds | 326 26 | Sand Coulee 3a teal beds n | —(Unconformity) ~(Unconformity) —(Unconformity (?) ) Extensive nor- Local normal a mal faults faults ; Bas Large folds Small folds Depression o8 Widespread Local uplift 8 io uplift aw : aS Fort Union (?) Sinking (?) Fort Union Widespread Clark Fork beds of Sinclair Widespread eo sinking and Granger sinking Fort Union (Local uncon- (Local uncon- formity) ——— formity) (?) ‘ Local warping Local warping (2) Tertiary (?) - = — Lance formation Sinking Lance formation (2) Sinking (?) Lance formation Sinking formability with the underlying cretaceous beds. « The Lance formation is here placed in the Tertiary (?) to conform with the usage of the U.S. Geological Survey, although the local evidence indicates con- Ta 14. hy, 16. T7. THE HEART MOUNTAIN OVERTHRUST, WYOMING 557 Hewett, D. F. “Sulphur Deposits in Park County, Wyoming,” U.S. Geol. Survey, Bull. No. 540-R (1913). Hewett, D. F., and Lupton, C. T. “Anticlines of the Southern Part of the Bighorn Basin, Wyoming,” U.S. Geol. Survey, Bull. No. 655 (1917); (a) pp. 16-41. Richards, R. W., and Mansfield, G. R. ‘“‘The Bannock Overthrust, a Major Fault in Southeastern Idaho and Northeastern Utah,” Jour. Geol., XX (1912), pp. 681-709; (a) p. 704. Veatch, A. C. ‘‘Geography and Geology of a Portion of Southwestern Wyoming,” U.S. Geol. Survey, Prof. Paper No. 56 (1907); (a) p. 109. Willis, B. ‘Stratigraphy and Structure of the Lewis and Livingston Ranges,” Mont. Bull. Geol. Soc. Amer., XIII (1902), pp. 302-52; (a) p. 344. SUMMARIES OF PRE-CAMBRIAN LITERATURE OF NORTH AMERICA EDWARD STEIDTMANN University of Wisconsin INTRODUCTORY In r915, summaries of pre-Cambrian literature from 1909 to 1915 were published in the Journal of Geology. The summaries published herewith bring this review nearly down to date. For 1919, attention has been given only to those publications which are likely to contain important papers on the pre-Cambrian. Emphasis in these summaries is placed on stratigraphic facts and problems. I. LAKE SUPERIOR REGION AND ISOLATED PRE-CAMBRIAN AREAS OF THE MISSISSIPPI VALLEY In the Lake Superior region notable contributions to the stratigraphy of the pre-Cambrian have been made by Wolff, Grout and Broderick, Hotchkiss, and Allen and Barrett. The tendency has been to subdivide the iron formations of the leading districts into several units and to connect the occurrence of major ore deposits with certain of these units. In the main productive portion of the Mesabi district, Wolff from his extended experience in drilling and mining has recognized four divisions of the Biwabik iron formation. Grout and Broderick have recognized similar units in the eastern, less-productive portion of the formation. Hotchkiss has found an unconformity in the Ironwood iron forma- tion of the Gogebic district and has recognized two units in the upper division and three in the lower. He has also found an unconformity between the iron formation and the overlying Tyler slates. The recognition of these new unconformities, Hotchkiss believes, does not call for a revision of correlation. Allen has discovered an unconformity cutting the Upper Huronian 558 PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 550 formations of Van Hise and Leith in the eastern part of the Gogebic district. Between the Keweenawan and this unconformity, Allen finds a series of clastic sediments which are neither Keweenawan nor Upper Huronian. Allen decides that the Upper Huronian of Van Hise and Leith is to be correlated with the Middle Huronian of the Marquette district. On much less convincing evidence than in the Gogebic district he also concludes that a Middle Huronian is found in the Menominee district. Following these studies, a revised correlation of the Lake Superior region is offered by Allen in which all the important iron formations of the region are designated as Middle Huronian. Allen has made important studies of the region between the Penokee and Iron River districts, but owing to the drift cover of the area he has not obtained the facts for a satisfactory correlation of the Iron River with the Penokee and Marquette districts. He has also studied the Gwinn district and the eastern extension of the Menominee district. | R. C. Allen™ believes that the pre-Cambrian rocks of the Gwinn district located about sixteen miles south of Marquette, Michigan, comprise two unconformable series which he correlates with the Upper and Middle Huronian respectively. The succession accord- ing to Allen is shown on page 560. The graywacke and conglomerate near the middle of the sedi- mentary succession constitutes the evidence on which Allen bases his conclusion for unconformity in the Huronian system. The conglomerate contains fragments which resemble the underlying sediments including iron ore. The two unconformable series appear to be structurally concordant however. His correlation of the lower series with the Middle Huronian is based on the fact that the Lower Huronian is not known to contain iron formation of the type in the Gwinn district. Allen and Barrett? believe that the acid mica schists of Wolf Lake are the metamorphosed equivalent of the Paint slates of ™R. C. Allen, ‘Correlation and Structure of the Pre-Cambrian Rocks of the Gwinn Iron Bearing District of Michigan,” Jour. Geol., Vol. XXII, No. 6 (1914); also in Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 161-64. 2R. C. Allen and L. P. Barrett, “‘The Paint Slate and the Wolf Lake Granite, Gneiss and Schist,’’ Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 131-39. 560 EDWARD STEIDTMANN the Iron River and Crystal Falls districts. ‘The granite intrusive into them, they believe, is of the same age as their Presque Isle granite of the Gogebic district. Wolf lake is situated between the Iron River and Gogebic districts. Quaternary—Pleistocene glacial deposits Ordovician and Cambrian—limestone and sandstone Algonkian—Keweenawan, probably represented by certain basic dikes which cut all formations Slate, ferruginous slate, | Equivalent of Michi- Upper Huronian chert and quartzite, quartz- { gamme slate Princeton series ite quartzite of Mar- Conglomerate graywacke Equivalent of Goodrich quette district Gray slate Black slate Equivalent of Negaunee Unconformity Iron formation iron formation and Siamo Middle Huronian Gray slate slate Gwinn series Black slate Ones \Equivalent of Ajibik slate Unconformity Archean system Laurentian—Granite and greenstone, mainly granite Keewatin Allen and Barrett* find that the pre-Cambrian rocks at the Little Lake Hills about seven miles east of the Gwinn district of Michigan consist from the base upward of conglomerate, arkose, and quartzite separated by an unconformity from a conglomerate quartz slate and quartzite respectively. The strata appear to be concordant, but the conglomerate near the middle of the column is clearly basal. The lower series is correlated with the Middle Huronian Gwinn series of the Gwinn district; the upper series with the Upper Huronian—Princeton series of the aforementioned dis- trict. The absence of iron formation in the Little Lake Hills, Middle Huronian is thought by the writers to indicate the depth of erosion of the Gwinn series before the Princeton series were laid down. tR.C. Allen and L. P. Barrett, ‘Evidence of the Middle-Upper Huronian Uncon- formity in the Quartzite Hills at Little Lake, Michigan,” Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 153-59. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 561 Allen* finds a series of magnetic belts in a region covered by Paleozoic rocks and glacial drift extending east of Waucedah on the Menominee range to Escanaba, a distance of about twenty- eight miles. From Waucedah, the buried eastward extensions of two productive iron formations have been traced for six miles by magnetic survey. The detached magnetic belts east of this limit may or may not be equivalent to the iron ranges at Waucedah. They undoubtedly represent belts of folded sedimentary rocks which may include iron formation. Allen and Barrett? describe the Conover district of northern Michigan, about forty-five miles southeast of the Gogebic range, as a drift-covered area showing several magnetic belts. Drilling has shown that the underlying rocks are slates intruded by granites. The authors believe that the slates are the equivalent of the iron- bearing slates of the Iron River district. Allen and Barrett state that the existence of the Manitowish - range is indicated by a series of strong, parallel linear magnetic belts about twenty-five miles southeast and parallel to the Gogebic range. Drilling has shown that the underlying drift-covered rocks are mica schists intruded by granites. Allen and Barrett? report that the Vieux Desert district of Wisconsin and Michigan lying about forty miles southeast of the Gogebic range is a deeply drift-covered area showing several faint magnetic belts. As shown by drilling, the underlying rocks are acid gneisses and schists. Allen and Barrett’ correlate the strongly magnetic iron forma- tion of the Marenisco range with the Ironwood formation of the =R. C. Allen, “Relative to an Extension of the Menominee Iron Range Eastward from Waucedah to Escanaba, Michigan,” Econ. Geol., Vol. TX, No. 3 (1914), pp. 236-38, 1 map; also in Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15. 2R. C. Allen and L. P. Barrett, ‘‘Geology of the Conover District,” Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 123-29. 3R. C. Allen and L. P. Barrett, ‘‘ Geology of the Manitowish Range,” Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. IZI-17. 4R. C. Allen and L. P. Barrett, “‘Geology of the Vieux Desert District,’ Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 119-21. 5R. C. Allen and L. P. Barrett, ‘Geology of the Marenisco Range,’ Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 65-86. 562 EDWARD STEIDTMANN Gogebic district. Both are classified by them as Middle Huronian. The succession in the Marenisco range, they state, is as follows: Keweenawan Diabase Igneous contact Intrusive granite Intrusive greenstone Huronian Middle Huronian Slate (Animikie) Extrusive lavas Iron formation Quartzite and graywacke Unconformity Northern Area—granite and greenstone Archean {Southern Area—mica schist, green schist, and amphibolite (may be Huronian) The Marenisco range lies three to twelve miles south of and is parallel to the Gogebic range. . Allen and Barrett’ describe the Turtle iron range about eighteen miles southeast of and parallel to the Gogebic range of northern. Wisconsin and Michigan. ‘The succession of the range as stated by the authors is as follows: ALGONKIAN: K f Intrusive diabase eweenawan '\ Granite and greenstone Middle Huronian Granite (Animikie) Effusive Agglomeratic and ellipsoidal Greenstone Black slate and graphitic schist Tron formation Huronian Quartzite and mica schist Unconformity ? Mica schist (may be Middle Huronian) Lower Huronian Dolomite and dolomitic quartz- ite Quartzite Unconformity ARCHEAN: aan { Mica schist and | green schist =R. C. Allen and L. P. Barrett, ‘Geology of the Turtle Range,”’ Mich. Geol. Surv. Pub. 18 (1915), Geol. Ser. 15, pp. 86-109. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 563 Allen and Barrett' find that a group of dominantly clastic sediments overlie with marked unconformity the Upper Huronian of Van Hise and Leith in the Gogebic district. They also find that a granite intrudes the Upper Huronian. Their threefold division of the rocks between the Keweenawan and the Archean of the Gogebic, they claim, identifies this succession with the three Huronian divisions of the Marquette district. They desig- nate as Middle Huronian, the Upper Huronian of Van Hise and Leith in the Gogebic district and extend this change to the corre- lation by these authors to every other district of the Lake Superior region, outside of the Marquette district. The revised correlation table which Allen and Barrett offer places every important Lake Superior iron formation excepting that of the Vermilion district in the Middle Huronian. In 1919, Allen? extended to the Menominee district the three- fold classification of the Huronian which he had found applicable to the Gogebic district of northern Michigan. His revised correla- tion of the Huronian of the Menominee district is shown on page 564. In previous correlations of the Menominee, notably that of Bayley in Monograph 46 of the U.S. Geological Survey, the Middle and Upper Huronian of Allen were regarded as a conformable suc- cession designated Upper Huronian. Van Hise and Leith, in rorr, recognized a Middle Huronian quartzite, but later Leith is said to have given up this revision. Allen bases his separation of the Upper Huronian of Bayley on the fact that some drill holes have shown what is interpreted as a basal conglomerate between the Hanbury slate and the Curry iron formation. Many drill holes do not show this conglomerate. He also appeals to the fact that the Randville dolomite in different places is covered by various formations ranging from the Traders iron formation to the Han- bury slates. Bayley had accepted one of the alternative explana- tions for this fact, viz., that the formations overlying the Randville were all conformably deposited on a very uneven surface of the tR. C. Allen and L. P. Barrett, ‘‘Contributions to Pre-Cambrian Geology,” Mich. Geol. Surv. Pub. t8 (1915), Geol. Ser. 15, pp. 13-164, 12 pls., 11 figs., maps. 7R. C. Allen, “Correlation of Formations of Huronian Group in Michigan,” Am. Inst. Min. and Met. Eng. (1919), No. 153, pp. 2579-94. 564 EDWARD STEIDTMANN Randville. Allen emphasizes this relation of the Hanbury slate as evidence of unconformity with the Curry iron-formation group. No proof of angular discordance between the Hanbury slates and the Curry iron formation seems to have been found by Allen. His recognition of the Loretto slate is based on the fact that on a certain Period Epoch Stage Formation Epi-Huronian | Revolution Emergent Granite interval -Quinnesec Eruptive contact, basic extru- Upper Huronian sives, sills and dikes Hanbury Great slate series with beds of conglomerate, quartzite, graywacke, ferruginous chert, and impure limestone. Thickness ? Emergent | interval Loretto Slate 400 feet Macale Curry Iron formation 100 to 200 feet Huronian H : Brier Ferruginous, siliceous banded uronian slate 300 to 400 feet Traders Conglomerate, quartzite, and iron formation 150 feet Emergent | interval Lower Huronian | Randville Dolomite, cherty dolomite, and talcose facies 1,000 to 1,500 feet. Conglomerate, arkose, graywacke, and quartzite 1,200 feet Great Archeo|zoic interval forty-acre lot, a few drill holes passed through what he assumes to be the basal Hanbury conglomerate and then through a slate before striking the Curry iron formation. It appears that Allen’s revision is based solely on a very local occurrence of a fragmental rock above the Curry iron formation. Broderickt presents a detailed classification of the beds of the Biwabik iron formation in the eastern part of the Mesabi range. He retains Wolff’s general classification into Upper slaty tT. M. Broderick, “‘ Detail Stratigraphy of the Biwabik Iron Bearing Formation, East Mesabi District, Minnesota,” Econ. Geol., Vol. XIV (1919), pp. 441-51. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 565 beds, Upper cherty beds, Lower slaty beds, and Lower cherty beds. Broderick’ interprets certain negative magnetic lines of the Duluth gabbro as due to a certain angle of inclination of the mag- netic formations with the horizontal. According to Cayeux’ traces of crinoids are found in the iron formations of the Gogebic, Mesabi and Menominee ranges. They consist of circular, quadrilateral, and hemispherical bodies larger than the oélites and of polygonal cells whose walls are composed of iron. The cells occur with and without alignment. Grout finds that siliceous pegmatites formed on all sides of the basic Duluth gabbro, but that distinct dikes occur only outside the contact. He infers that the two magmas separated in the liquid state. Grout* presents interesting petrographic descriptions of the Biwabik iron formation of the eastern part of the Mesabi district. He concludes that originally the iron formation was a shallow water deposit formed mainly by organic processes. Grout’ proposes the name lopolith for intrusions like the Duluth gabbro whose floors and roof sag downward toward the middle. Evidence is introduced for concluding that the main mass of this intrusion was along a plane of unconformity. Sug- gestions are made that the method of intrusion of the lopolith is different from that of a laccolith. Grout believes, as indicated by his diagrams, that the dip of the pre-Cambrian beds around the western rim of Lake Superior 1s due to settling rather than compression. It would be interesting to find out whether the nature of the fracture and flow cleavage structures in these formations checks with this view. tT. M. Broderick, ‘“‘Some Features of Magnetic Surveys of the Magnetic Deposits of the Duluth Gabbro,” Econ. Geol., Vol. XIII (1918), pp. 35-40. 2 L. Cayeux, ‘‘ Existence de restes organique dans le roche ferruginenses associées aux minérais de fer huroniens des Etats-Unis,” Acad. Sci. Paris Compt. rend., Vol. 153, PP. 910-12. 3F. F. Grout, ‘‘The Pegmatites of the Duluth Gabbro,” Econ. Geol., Vol. XIII (1918), pp. 185-97. 4F. F. Grout, ‘‘The Nature and Origin of the Biwabik Iron Bearing Formation of the Mesabi Range, Minnesota,”’ Econ. Geol., Vol. XIV, (1919). pp. 452-64. 5 F. F. Grout, ‘“Lopolith, an Igneous Form Exemplified by the Duluth Gabbro,” Am. Jour. of Science, Vol. CXCVI (1918),"pp. 516-22. 566 EDWARD STEIDTMANN Grout’ states that the Duluth gabbro lopolith shows difierentia- tion into the gabbro and granite families and discusses the prob- lem of the processes of differentiation. Hore? presents descriptions of the most important copper lodes of Upper Michigan, and reviews the literature of the region. He concludes that the ores are replacement deposits formed by chloride solutions liberated with the formation of the traps in which they occur or with which they are associated; that they have not been modified except in very minor ways since they were formed; but that the rocks in which they are formed were farther tilted since the ores were formed. Hotchkiss, Bean, and Wheelwright? map a part of the pre- Cambrian area of Ashland, Bayfield, Washburn, Sawyer, Price, Oneida, Barron, Rusk, and Chippewa counties. The chief aim of the work is to show the distribution of iron-bearing formations. Since the area is nearly all drift-covered, magnetic surveys furnish most of the facts. The pre-Cambrian sediments are classed as Barron quartzite and undivided Huronian. Keweenawan traps and granites and gneisses probably of various ages are found in the area. The report has notable chapters on field methods used in work of this type and on the nature and interpretation of magnetic data. ) Hotchkiss? has made an important contribution to the study of the stratigraphy and structure of the Gogebic iron district of northern Wisconsin and Michigan. ‘The influence of stratigraphy and structure on the formation of the ores is also discussed by him. Many new facts and relationships are presented. Although he recognizes several new unconformities in the succession, he does not believe that the facts now known warrant any fundamental tF, F. Grout, ‘A Type of Igneous Differentiation,’ Jour. Geol., Vol. XXVI (1918), pp. 626-58. 2R. E. Hore, “Michigan Copper Deposits,’ Mich. Geol. Surv. Pub. 19 (1915), pp. 19-161, 18 pls., 16 figs. 3W. O. Hotchkiss, ‘‘Mineral Land Classification Showing Indications of Iron Formations,” Wis. Geol. Surv. Bull. No. 44 (1915), 378 pp., 8 pls., 39 figs. (incl. maps). 4W. O. Hotchkiss, ‘‘Geology of the Gogebic Range and Its Relation to Mining Developments,” Eng. and Min. Jour., Vol. CVIII (1919), pp. 443-52, 501-7, 537-41. 577-85. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 567 revision of the classification by Van Hise and Leith. The essentials of the stratigraphic classification by Hotchkiss follow: Keweenawan sandstones and conglomerates overlain by basic flows Unconformity | Graywacke slates Tyler graywacke slate } Iron carbonate slates o-2 miles thick Pabst member-cherty and fragmental slate beds Unconformity Upper Ironwood {Anvil wavy-bedded ferruginous chert member formation \Pence even-bedded ferruginous slate member Slight unconformity Norrie wavy-bedded ferruginous chert member Yale member—interbedded ferruginous cherts and ferruginous slates Plymouth member—wavy-bedded ferruginous chert (most mines located on this member) Lower Ironwood formation Palms quartzite 400 feet to 800 feet Unconformity Bad River cherty dolomite with quartzite below in eastern part of district Unconformity Granite and green schist both of igneous origin The unconformities within the Ironwood formation and between the Tyler slate and the Ironwood formation have not been described before. Their existence is inferred from basal conglomerates and evidences of erosion of the members underlying the unconformity. Lane? finds that the strata on the east side of the Copper range were uplifted and the eastern sandstone deposited on them. ‘The Trap range subsequently overrode the sandstones in places several hundred feet. Leonard? states that pre-Cambrian granite struck in wells of the Red River valley is the only known pre-Cambrian rock of North Dakota. tA. C. Lane, ‘Abstract,’ Bull. Geol. Soc. of America, Vol. XXIV (1913), p- 718. 2 A. G. Leonard, ‘‘The Geology of North Dakota,’ Jour. Geol., Vol. XX XVII (t919), Pp. 1-27. 568 EDWARD STEIDTMANN Leith? suggests that the unconformity at the base of the Cam- — brian was developed by a process of cut and fill, and that the common occurrence of late pre-Cambrian terrestrial sediments is more than a coincidence, but is related to the development of the basal Paleozoic unconformity. Nebel? presents a petrographic study of certain basal portions of the Duluth gabbro and its contact effects. Powers’ reports that drilling in the east-central portion of Kansas has shown the existence of pre-Cambrian granite of con- siderable relief occurring along a north-south line. Wolff reports that the average thickness of the Mesabi iron formation is six hundred and twenty feet, and that it consists of four divisions which from the top down are as follows: Upper slaty horizon, Upper cherty horizon, Lower slaty horizon, and Lower cherty horizon. The ores occur chiefly in the two cherty horizons and in the Lower slates. Marked differences exist between the ores of the various horizons. =C. K. Leith, “‘Relations of the Plane of Unconformity at the Base of the Cambrian to Terrestrial Deposition in Late Pre-Cambrian Time,” Congrés Géologique International, XII. Session Canada, pp. 333-37- 2M. L. Nebel, ‘‘The Basal Phases of the Duluth Gabbro Near Gaamicneenn . Lake, Minnesota, and Its Contact Effects,” Econ. Geol., Vol. XIV (1919), pp. 367-402. 3 Sidney Powers, ‘‘Granite in Kansas,”’ Am. Jour. of Science (4th ser.), Vol. XLIV (t917), pp. 146-50, 1 fig. 4 J. F. Wolff, ““Recent Geologic Developments on the Mesabi Range, Minnesota,” Am. Inst. Min. Eng. Bull. No. 118 (1916), pp. 1763-87, 14 figs. [To be continued] Notice to Subscribers @. The total income of the Journal of Geology, including subscriptions and contributions from _the University and from individuals, is practically the same from year to year while the cost of labor, paper, printing, and publishing has become constantly greater. _ @ During recent years many publishers of peri- a odicals have been forced to make large increases ‘in subscription prices in order to ‘meet the in- creased cost of manufacture. | : @. It has been decided in the case of the Journal of Geology to meet the demands of present condi- _ tions by changing the Journal to a bi-monthly basis of issue. This will not effect the editorial policy nor the value of the contributions to the Journal. @ This plan, which enables us to maintain the same price to our subscribers, will be inaugurated January 1, 1921, commencing with Volume XXIX.’ FINE INKS 4x? ADHESIVES For those who KNOW level of typing must go ‘to the Rer ington. Drawing Inks Eternal Writing Ink Engrossing Ink s 2 9 Taurine Mucilage r | g a Ins Photo Mounter Paste . Drawing Board Paste . Liquid Paste Office Paste f Vegetable Glue, Etc. Are the Finest and Best Inks and Adhesives Emancipate yourself from the use of corrosive and ill-smelling inks and adhesives and adopt the Higgins Inks and Adhesives. They willbea revelation to you, they are so sweet, clean, well put up, and withal so efficient. _ At Dealers Generally The Setpetrane Remington a cor respondence, the Key-Set Remingt for form work, and the Remingt Accounting Machine (Wahl Mecha ism) are the latest achievements clerical labor saving. : s Remington Typewnter. Comoanad (Incorporated) 374 Broadway CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street ‘Brooklyn, N.Y. The Gakic of Vancouver and Vicinity By EDWARD M. J. BURWASH _ 14 half-tones, 10 line drawings, and 2 colored maps A thorough and authoritative survey of the region, made. more interesting by the remarkable photographs, which include a profile of the Grouse Mountain spur, the Lions” from Mount Brunswick, a view of Red Mountain from Black Tusk Mountain, Mount Garibaldi from the south, a glacial section from a cutting on Keith Road, and others equally inter- _esting. One colored map shows the ‘topography of Vancouver range and the other is a geological map of Vancouver and ae vicinity. s 112 pages, paper covers; $1.50, postpaid $1.60 | THE UNIVERSITY OF CHICAGO PRESS ~ CHICAGO x soe ILLINOIS: vial JME XXVIIL : ie : net NUMBER 7 7 EDITED BY \ au “THOMAS Cc CHAMBERLIN AND ROLLIN D. SALI 6 Cala es With the Active Collaboration of NAL vi Ot aw ART WELLER, Invertebrate Paleontology ALBERT JOHANNSEN, Petrology eos Lee ROLLIN T. CHAMBERLIN, Dynamic Geology ) 4 ’ ASSOCIATE EDITORS ; \RCHIBALD GEIKIE, Great Britain JOHN C. BRANNER, Leland Stanford Junior University ARLES BARROIS, France RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. RECHT PENCK, Germany uy WILLIAM H. HOBBS, University of Michigan NS REUSCH, Norway FRANK D. ADAMS, McGill University ia DEGEER, Sweden CHARLES K. LEITH, Uni ity of Wi " W. EDGEWORTH DAVID, Australia gi AR admin ea 9 WALLACE W. ATWOOD, Harvard University \RLES D. W. ALCOTT, Smithsonian Institution WILLIAM H. EMMONS, University of Minnesota . H P. IDDINGS, Washington, D.C. ARTHUR L. DAY, Carnegie Institution OCTOBER-NOVEMBER, Lae CLARENCE N, FENNER 569 poe PHYSICAL PROPERTIES OF ICE - - - + - Moronort MatsvuyAMA 607 “OF METAMORPHIC ROCKS ia ye Rios cak leew Nee ose Gorpon Carison” 632 UMMARIES OF PRE- CAMBRIAN LITERATURE OF NORTH AMERICA _EDWARD STEIDTMANN 643 ree ee rion ate ot 2 Se oe Se a ee THE. UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS, U.S.A. : " THE CAMBRIDGE UNIVERSITY PRESS, Counts THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoro, FUKUOKA, SENDAI THE MISSION BOOK COMPANY, SHANGHAI THE JOURNAL OF : x) ote BDIFED BYS THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of STUART WELLER ~ : eM JOHANNSEN | ak Jnvertebrate Paleontology |. t bie oe Peaaleny FS ROLLIN T. CHAMBERLIN Ks faa Dynamic Geology ~ March 15, May 1, June ts, Aperist ie Sboterber ts, November 1, December T5ae {The Le price is $4.00 per year; the price of single copies is 65 cents. Orders for service of less. than a year will be charged at the single-copy rate. {] Postage is prepaid by the publishers on all orders es the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Ha Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. { Postage is charged extra as follo For Canada, 30 cents on annual subscriptions (total $4.30), on single copies 4 cents (total 69° ce for all other countries in the Postal Union, 53 cents on annual subscriptions (total $4. 53), on sir copies 11 cents (total 76 cents). {] Patrons are requested to make all remittances Pen to 0 The L _versity of Chicago Press i in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, Louie E. C4: ‘Years subscriptions, including postage, 21s. each; single copies, including postage, 3s. 6d. each. ee For Japan and Korea: The Mariven KaWuclikiK veta: 11 to 16 Nihonbashi Tori Sanchi a Tokyo, Japan. Yearly subscriptions, including postage, Yen 9.00 each; single copies, ‘includ g postage, Yen 1.50 each. For China: The Mission Book Company, 13 North Szechuen Road, Shanghai. ‘Yearly s scriptions $4.00; single copies 65 cents, or their equivalents i in Chinese money. pias! extra, mailed direct oustle of Shanghai, on yearly subscriptions 53 cents, on single copies 11 cents: 3 Claims for missing numbers should be made within the month following the regular ignite of pu cation. The publishers expect to supply missing numbers free anly when losses have been sustained transit, and when the reserve stock will permit. Business correspondence should be addressed to The University of Chicago Pee Chicago, Communications for the editors and manuscripts should be addressed to the Hdtier of % JouRNAL oF GEOLOGY, the University of Chicago, Chicago, Ill. une Contributors are requested to write scientific and proper names with particular care ala n cit to follow the form shown in the pages of the Journal. ' paint Separates, if desired, must be ordered in advance of publication. Separates of leading article: r ‘printed as ordered, of which 50 (without covers) will be furnished gratis, the actual cost of the rema (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depe: the amount of work in remaking the pages into forms, Disney Paper, tindiae etc. Le Net ye of cuts and the amount of work required upon them. From time to time supplements to the Journal will be issued. One on T Pe ue Feldspars, by H. L. copa (about 100 pages) va appear nous, and will be sent Pie es a subscribers. July 15, zo18, y Among articles to appear in early numbers of the Journal of Geology are the following: Diastrophism and the Formative Processes. XIII. The Planet- esimals and their Infall. By T. C. CHAMBERLIN. Geologic Reconnaissance of the Southern Part of the Taos Range, New Mexico. By JOHN W. GRUNER. Summaries of Pre-Cambrian Literature of North America. Papers II-VII. By Epwarp STEIDTMANN. A Pleistocene Peneplain in the Coastal Plain. By H. F. CLELAND. The Mineralography of the Feldspars. By H. L. Attn. To be published as a supplement to the Journal of Geology. Features of a Body of Anorthosite-Gabbro in Northern New Y ork. By Wix11Am J. MILLER. A Glacial Gravel Seam in Limestone at Ripon, Wisconsin. By F. T. THWAITESs. The Physical Chemistry of the Crystallization and Magmatic Differ- entiation of Igneous Rocks. By J. H. L. Vocrt. A New Form of Diplocaulus. By M. G. MEHL. Volcanic Earthquakes. By CHARLES DAVISON. The Mechanical Interpretation of Joints. By WALTER H. BucHER. Types of Rocky Mountain Structure in Southeastern Idaho. By GEORGE RoGERS MANSFIELD. POSITIONS OF ALL KINDS Never was the demand so great for qualified teachers and specialists. For ten years we have given our time and energy to this work. Write for our free literature. State qualifications briefly. PUBLISHED PRICES Will be paid for Co-operative Instructors’ Association Marion - - Indiana The Journal of Geology We will pay the published The Origin of the Earth prices for any numbers of By Thomas C. Chamberlin the Journal of Geology, A new chapter in “‘The Story of the Birth Vol. I, Not 25 to Volek of the Earth.” No: 6) if Sent) teuusmem xii+ 272 pages, 8vo, cloth; $1.75, postpaid salable (condition miaaae Pree. clearly the name and ad- dress to which you want check mailed. One of “‘“The University of Chicago Science Series”’ The Duiversiey of Chicago Press Chicago - - - - Illinois The University of Chicago Press Librarians s If you are not 5750-58 Ellis Avenue Sa aa ee UN Chicago, Illinois Library Catalogue Cards announcing new pubiications, ask to be put on our mailing list. The University of Chicago Press CHICAGO ILLINOIS to be put on our mailing list. If you do not receive our advertising matter regularly, ask THE UNIVERSITY OF CHICAGO PRESS Cambridge Gniversity Press Fossil Plants. A Text-Book for Students of Botany and Geology. By A. C. SEWARD, M.A.,F.R.S. In4vols. Demy 8vo. Vol.I. With 112 illustrations. $3.60. Vol. II. With 266 illustrations. $5.60. Vol. III. With 253 illustrations. $6.50. Vol. IV. With 190 illustrations. $6.50. Palaeontology—Invertebrate. By Henry Woops, M.A., F.G.S. Crown 8vo. ‘Fifth edition. Revised and enlarged, with 173 illustrations. $3.75. Petrology for Students: an Introduction to the Study of Rocks under the Microscope. By Atrrep Harker, M.A., F.R.S., F.G.S. Fifth edition, revised. Crown 8vo. $2.75. A Treatise on Crystallography. By W.J. Lewis, M.A. Demy 8vo. $4.50. The Principles of Stratigraphical Geology. By J. E. Marr, Sc.D., F.RS., F.G.S. Crown 8vo. $2.00. An Introduction to Geology. By J. E. Marr, Sc.D., F.R.S., F.G.S. Crown 8vo. $1.30. Published by the Cambridge University Press, England The MacMillan Co., Agents in the United States 64 and 66 Fifth Avenue, New York City VOLUME XXVIII NUMBER 7 THE PewmweNA OF GEOLOGY OCTOBER-NOVEMBER 1920 THE KATMAI REGION, ALASKA, AND THE GREAT ERUPTION OF 1912 CLARENCE N. FENNER Geophysical Laboratory, Carnegie Institution of Washington In June, 1912, Mount Katmai was the scene of one of the greatest volcanic eruptions known in history. The material ejected, mostly in the form of pumice and fragmental glass, formed deposits whose total volume has been calculated to amount to nearly five cubic miles.t. At the town of Kodiak, a hundred miles away, this fragmental material (‘‘ash”’) fell to the depth of nearly a foot, and nearer to the volcano hundreds of square miles of territory were completely devastated. General knowledge of the effects of this eruption has been derived chiefly from the explorations of several expeditions sent out by the National Geographic Society, the first one under G. C. Martin and later ones under R. F. Griggs.* The expedition of the summer of 1919 was planned on a considerably more ambitious scale than former ones, and an invitation was extended to the Geophysical Laboratory to co-operate in the scientific work. «G. C. Martin, ‘‘The Recent Eruption of Katmai Volcano in Alaska,” National Geographic Magazine, Vol. XXIV (February, 1913), No. 2, p. 131. 2G. C. Martin, Nat. Geog. Mag., Vol. XXIV (February, 1913), No. 2, p. 131; R. F. Griggs, Nat. Geog. Mag., Vol. XXXI (January, 1917), No. 1, p. 13; and Vol. XXXII (February, 1918), No. 2, p. 115; Ohio Journal of Science, Vol. XIX (1918), ie, 569 570 CLARENCE N. FENNER Under this arrangement E. T. Allen, E. G. Zies, and C. N. Fenner joined the party for the purpose of studying the chemical and geological phenomena. Previous to our departure for the Katmai region the publications of Dr. Martin and of Professor Griggs and the information obtained from conversations with them were of much assistance to us in making plans for the trip, and, while we were on the ground, Professor Griggs’s knowledge of the region and its phenomena continued to be of great service. The party spent about two months and a half in the field, which is about as long a working season as is practicable. Since the return to Washington, much time has been given to the study of the materials collected, and a full report will be pub- lished later. In advance of such publication, however, it has been thought that a shorter article, descriptive of some of the features of chief geologic importance, may be of interest, and is here presented. Necessarily in this brief treatment, many matters to which attention has been paid in our work will be omitted entirely, and in the case of others the basis for conclusions will be presented in brief form only. Fuller discussion must be reserved for the more comprehensive articles to follow. TOPOGRAPHY AND GENERAL GEOLOGY The Katmai country is situated near the base of the Alaska peninsula—that long arm which extends southwestwardly from the southern shore of continental Alaska and, with the Aleutian Islands, reaches nearly to Kamchatka (Fig. 1). Previous to the eruption of 1912, the Katmai region, though difficult of access, was not entirely unknown. On the Pacific side of the volcanic range was the small native village of Katmai, not more than twenty miles from the volcano. On the Bering Sea slope, at the head of Naknek Lakes, was the similar village of Savonoski. Between them ran a trail which had probably been traveled by the natives for many years, and more recently had been used fairly frequently by white men as a means of crossing the peninsula. In 1898 J. E. Spurr, of the United States Geological Survey, led a party over the Katmai trail, but observed nothing which might be considered to indicate that the preliminary pro- ° THE KATMAI REGION, ALASKA 571 cesses leading to the eruption were at work except that near the summit of the Pass the party experienced several earthquake shocks. Spurr’s record" is of much value, however, for the infor- mation it gives on the general character of the country. Their route lay through the midst of the area that was subsequently devastated. Further reference to this report will be made later. Noknek Lake [sar Arm As iter ") FAmahk Gay = 2. Cape Dktugitar eB { [are Buy SHELIK OF |S TRAIT LEGEND a L SINY Glacier, Contour interval SOO: feet By 0 8 Church 1916 Streams choked with pumice, and quicksands | | Statute Miles i 1Ss°s0" % isa? 40) Fic. 1.—Sketch map of the Katmai region. By courtesy of the National Geographic Society. In the Katmai country and its vicinity the volcanic mountains occupy a comparatively narrow strip of territory approximately parallel with the coast line, bounded on both the northwest and southeast sides by areas of predominantly sedimentary rocks. These sedimentary areas exhibit many forms of mountainous tJ. E. Spurr, “A Reconnaissance in Southwestern Alaska in 1808,” Twentieth Annual Report, U.S. Geological Survey (1898-99), Part VII, pp. 31-264. J 572 CLARENCE N. FENNER relief, and the higher summits are but little inferior in elevation to those of the volcanic belt, but dissection has here reached a stage at which broad valleys of moderate slope have been developed. The line along which arise the active or recently active volcanoes is one of the longest and most clearly defined volcanic chains in the world. At its northeast end the farthermost volcano whose character is definitely known is Mount Redoubt, though Mount Spurr, Black Peak, and Double Peak, from their position and charac- teristics, appear to prolong the range still farther to the northeast. These all lie well in the interior, among characteristically conti- Fic. 2——The Katolinat Mountains, between foot of Valley of Ten Thousand Smokes and head of Naknek Lakes. These mountains show sections of Upper Jurassic shale and sandstone, several thousand feet in thickness, in horizontal strata. Postglacial canyon in foreground. Photograph by J. D. Sayre, 1018. nental structural features. Thence the range runs southwestwardly to the base of the Alaska peninsula, and follows the latter through- out its length. Here its course lies through a region of nearly horizontal sediments at only a moderate distance from the edge of the continental shelf, but where, at about the end of the Alaska peninsula, the edge of the shelf curves to the northward, the line of volcanoes continues without deviation and strikes off across oceanic deeps of 1,000 to 2,000 fathoms. The Aleutian Islands and their volcanoes form the summits of a narrow, steep-sided ridge, with great depths of water on both sides. The well-defined character and continuity of this volcanic belt were noted by I. C. THE KATMAI REGION, ALASKA 573 Russell, who says: ‘This belt of igneous activity is nearly 1,600 miles long. .... It is so narrow and well defined that two parallel lines drawn on a map of Alaska, twenty-five miles apart, may be made to include nearly every volcano in the belt that is known to have been active in historic times.’ It seems that such a linear distribution must indicate a major fracture in the earth’s crust, and we might expect to find plain Fic. 3.—View across canyon of Katmai River, from lower slopes of Mount Katmai, looking at Barrier Range. These mountains consist of shale and sandstone, believed to be of Upper Jurassic age, in beds gently inclined away from the observer, with some igneous intrusives. Photograph by D. B. Church, 10916. evidences of dislocation of strata or even profound disturbances associated with it. On the contrary, very little evidence of this kind is apparent in the region explored by us. On the northwest side of the.belt masses of Upper Jurassic sediments (Spurr’s Naknek series), 5,000 feet at least in thickness and possibly much more, lie in horizontal, undisturbed strata, whose continuity may often be ™T. C. Russell, Volcanoes of North America, p. 268. 574 CLARENCE N. FENNER followed by the eye for miles along the mountain sides. A typical mountain block of horizontal sediments is shown in Figure 2. On the southeast side of the range the sediments are still the shales and sandstones of the Upper Jurassic, little different lithologically or paleontologically from those on the other side. Structurally, however, this fact is observable—that they dip fairly uniformly away from the range at angles of 10 to 15°. Some typical views Fic. 4.—Looking up canyon of Katmai River from Prospect Point. On the left, Mount Katmai in the background, and its lava slopes and cliffs in the middle distance; on the right, the sediments of the Barrier Range. Just above the river level and at the foot of the lava cliffs, surfaces of glaciated sandstone (shown in Fig. s) were found. Photograph by D. B. Church, 1916. are shown in Figures 3 and 4. This difference of attitude of the beds on the two sides of the range may indicate block-faulting and tilting, but this seems remarkably slight evidence to be the only indication of a break of such great length and reaching to profound depths. That profound depths have been reached is indicated by the manner in which the break extends without deviation across fundamentally different surface structures. There is, however, a possibility that the volcanic chain may be situated not directly along the surface trace of the major fracture, but THE KATMAI REGION, ALASKA 575 along a system of accompanying breaks. Such a relation appears to be not uncommon in other volcanic districts. The Katmai group of volcanoes has evidently been built up on a platform of Upper Jurassic sediments, and several features show that comparatively recent flows have produced marked changes of topography. For instance, the canyon of Katmai River (shown in Fig. 4) is a narrow defile connecting open valleys above and below. It is evident that a former open valley here was invaded by floods of lava coming down from Mount Katmai, which shifted — Fic. 5.—Lava flow of basic andesite overlying glaciated sandstone and a small remnant of till (at pick), at foot of the lava cliffs shown in Figure 4. Photograph by R. F. Griggs, 1919. the river over upon the lower slopes of the Barrier Range.- The river has again cut nearly to grade along a narrow canyon, and beneath the lava-flows may be seen beds of till and sandstone surfaces grooved and polished by glacial action (Fig. 5). In this vicinity lava-flows of apparently post-glacial age measure 1,500 to 2,000 feet in thickness (see Fig. 4). The lavas of the group of cones that we are considering seem to be predominantly basic andesites. In Mount Katmai itself the suc- cession of flows that have built up the cone is now revealed in the great crater pit, and they appear to be of medium to basic character. The fragments of old rocks thrown out with the new lava in the 576 CLARENCE N. FENNER recent eruption are likewise of this composition, as are also the materials composing the bowlder beds at the rim of the crater (which probably represent a ground moraine whose components were transported by glaciers from the now annihilated upper slopes of the mountain). The testimony from all sources is concord: ant and demonstrates with a reasonable degree of certainty that Katmai is predominantly andesitic throughout. In addition to the lavas that have built up these cones, however, there seem to have been other products thrown out by them. In several places near the lower end of the Valley of Ten Thousand Smokes recent stream-cuttings show peat beds interstratified with many narrow bands of siliceous pumice. Probably the rhyolitic lava ejected in the latest eruption of Katmai was not the first highly siliceous differentiate evolved from the underlying magma. THE VALLEY OF TEN THOUSAND SMOKES AND ITS GREAT ASH DEPOSIT According to Spurr’s description the portion of the Katmai trail immediately to the northwest of the Pass ran for several miles through a wooded valley of varied topography. During the activities of the eruption, the floor of this valley was covered with a thick deposit of ash and pumice, which in most places has buried every detail of the former topography, and whose surface now forms a gently sloping plain. Thousands of fumaroles have found vent through this deposit and are sending out exhalations of hot gases and vapors. Professor Griggs, who discovered and described these remarkable features, has given to this valley the name “Valley of Ten Thousand Smokes” (Figs. 6 and 7). This ashy deposit covers the old floor of the valley to a great depth (possibly several hundred feet in certain areas) and extends up over Katmai Pass. Its distribution is shown on the map of the valley. From the very first explorations of the region by the National Geographic expeditions, Professor Griggs recognized that this deposit is quite distinct from the widespread ash-falls due to the explosive ejection of material from Katmai crater, and that it must be accounted for by the operation of other processes. Because of the fact that, when first discovered, certain of its characteristics THE KATMAI REGION, ALASKA 577 were thought to imply that its formation was the result of the extrusion of a mass of semi-fluid mud, this deposit was termed ‘‘ the great hot mud-flow,”’ and has been so described.* To the Geophysical Laboratory members of the 1919 expedition the evidence seemed opposed to the idea connoted by the term sate 7 THE VALLEY OF TEN THOUSAND SMOKES .~ FROM A SURVEY BY THE KATMAI EXPEDITIONS OF THE NATIONAL GEOGRAPHIC SOCIETY’ BERT ¥ GRISGS. DIRECTOR ~ ~ os ~~ TEN THOUSAND ‘< Fic. 6.—Topographic map of the Valley of Ten Thousand Smokes and adjacent region, from surveys made by topographers of National Geographic Society’s ex- peditions. ““mud-flow,” and early discussions among us led to the expression of the opinion by Dr. Zies that the evidence was much more in harmony with the idea of the movement of a dry, highly heated mass of sand and pumice than of a water-bearing mud. This t Professor Griggs’s article in the Ohio Journal of Science (Vol. XTX [December, 1918] No. 2, p. 117) gives an interesting description of this deposit and its remarkable features. 578 CLARENCE N. FENNER suggestion seemed from the first to have decided merits and later investigations served to strengthen it. Moreover, as evidence of various kinds accumulated, a much more complete conception of the attendant processes was afforded. The make-up of the deposit itself, its situation with respect to the configuration of the landscape, and various striking effects produced by it demand that certain definite characteristics should be attributed to it at the time of its appearance, and prescribe Fic. 7.—Looking northerly down the Valley of Ten Thousand Smokes. Photo- graph by R. F. Griggs, 1917. rather rigid limitations to one’s ideas as to its possible derivation. Observations show plainly that, in the first place, this material was not thrown violently into the air to descend over the general landscape, but that it was restricted very definitely to topographic depressions. In point of time, it was one of the first manifestations of activity, for it is covered by the subsequent ash-falls. The thorough manner in which vegetable material engulfed by it was carbonized and the indications of brush fires started by it can hardly be explained except on the supposition that it possessed a high temperature, probably near incandescence. In many places THE KATMAI REGION, ALASKA 579 adjacent to it but beyond its borders, fallen trees lie as if over- thrown by a violent wind accompanying it. This is observable along the margin of the deposit and also on those slopes of the Katolinat Range that lie at the foot of the valley and face up the valley in the direction from which the flow advanced. Katmai crater could hardly have been its source, as. physical obstacles stand in the way of distribution from that point, and Fic. 8.—The great sand-flow of the Valley of Ten Thousand Smokes, overlain by stratified ash from the Katmai ash-fall; view taken at junction of Knife Creek (at left) and River Lethe (at right, nearly concealed by steam cloud). The stratified terraces just above stream-level are not part of the sand-flow but are the result of recent stream deposition. Photograph by E. G. Zies, 1or19. glaciers that still cover the slopes of Katmai on this side would probably show noticeable effects from the movement of such an incandescent avalanche over their surfaces. The distribution of the material is such that there seems to be almost no escape from the conclusion that it originated within the valley itself and that we must look for its source in vents situated on the floor of the valley or on the lower slopes of the mountains at its head. Such 580 CLARENCE N. FENNER vents, however, would tend to be concealed by the material that they themselves extruded and by later materials from the ash- falls, and we are not able to point out the exact location of such vents with certainty. It is rather by a process of deduction that our opinions as to their position have been reached." The vents of extrusion may well have been located along the fissures that are now the seats of fumarolic activity. Support is Fic. 9.—Carbonized stumps and tundra. The foreground was covered by the sand-flow, but the standing trees in background were beyond its reach. Photograph by E. G. Zies, 1919. given to this supposition by evidences of former more vigorous activity of a mildly explosive kind at some of these localities. Possibly the newly formed crater of Novarupta in the upper part of the valley was one of the vents, differing from the others only in that it was of larger size than most and that its activity con- t Professor Griggs had previously, in one of his articles, expressed the opinion that the material must have been extruded from fissures within the valley. See article, ‘The Great Hot Mud Flow of the Valley of Ten Thousand Smokes,” by R. F. Griggs, in Ohio Journal of Science, Vol. XIX, No. 2, p. 130. THE KATMAI REGION, ALASKA 581 tinued into a stage not represented elsewhere, by which a plug of viscous lava was extruded. Views of this sand-flow or sand-avalanche and of some of the effects produced are shown in Figures 8-10. Considering further the origin of the sand-flow, we suppose that -rhyolitic magma, charged with dissolved gases, rose to the surface in the newly. formed vents. According to general observation the usual course for such a magma is either to retain its gases and form a flow of obsidian, or to evolve them with explosive violence and scatter the disrupted particles to a great distance. In this Fic. 1o.—Trees prostrated as if by wind accompanying sand-flow, though beyond reach of the avalanche of sand itself. Photograph by C. N. Fenner, roro. instance, however, it apparently pursued an intermediate course, and produced, by moderately forcible disruption, an outward- spreading and forward-moving torrent of incandescent sand and pumice, each particle of which was surrounded by and partially suspended in gases which it continued to give forth during its impetuous flow. An artificial reproduction of the properties that are believed to have characterized this ashy material at the time of its extrusion may be obtained by igniting the powder of basic magnesium car- bonate. The substance boils in a manner extraordinarily like a liquid, and the gases evolved buoy up the solid particles. In this 582 CLARENCE N. FENNER condition the mixture exhibits the lack of coherence and readiness to flow that characterize liquids. The exact counterpart of this deposit does not seem to have been described among volcanic phenomena elsewhere, but the avalanches of incandescent sand and ash that formed prominent Fic. 11.—Specimen of banded pumice (4X4 inches) from the deposit in the Valley of Ten Thousand Smokes. Adjacent bands show marked differences in composition, as indicated by a silica content of 74.70 per cent in the case of a light band, and 60.40 per cent in an adjacent dark band. The structure is believed to be due to digestion of foreign material. features of the eruptions of Pelée and La Soufriére in the Antilles in 1902 seem to offer many close analogies. The following quota- tion from the Encyclopedia Britannica gives a statement of the essential features that have been observed in the Peléan eruption: Its distinctive character is found in the sudden emission of a dense black cloud of superheated and suffocating gases, heavily charged with incandescent dust, moving with great velocity and accompanied by the discharge of immense volumes of volcanic sand, which are not rained down in the normal manner, THE KATMAI REGION, ALASKA 583 but descend like a hot avalanche... . . So much solid matter was suspended in the cloud, that it became too dense to surmount obstacles and behaved + rather like a liquid. Though one may find in the detailed descriptions of these sand- avalanches certain differences from the results seen in the area of the valley flow, they seem to be of degree rather than of kind, and the analogies are striking. Some of the pieces of pumice in this deposit show a banded or variegated structure, such as is illustrated in Figure 11. The difference of composition of adjacent bands is easily apparent, and in the specimen figured, determinations of silica by Dr. Allen have shown 74.70 per cent in a white band and 60.40 per cent in a dark band. It is believed that these structures are due to a process of partial solution of basic rock in the new siliceous magma. The very limited degree of mixing of solutions shown by these specimens hardly permits us to suppose that the solvent action was long continued; therefore we must look for the source in matter which became involved in the magma when it was near or at the surface and just prior to its foaming-up into pumice. There are several possibilities that should be considered. We might suppose that the sedimentary series beneath the valley had been previously injected by rocks of this description and that these were encountered by the new magma and material absorbed from them; or that the floor of the valley was composed of an old lava flow of basic com- position, which contributed material; but the supposition that, for a number of reasons, appears to me the most probable is that the source to which we should look is the deposit of lava bowlders of glacial origin that covered the floor of the valley to a great thick- ness. The fragments of undissolved andesite found with the ash are probably of the same origin, while the pieces of shale that are quite common in places were doubtless derived from the under- lying Naknek sediments. All of these may be duplicated in the lava and ejecta of Novarupta. THE FUMAROLES The fumaroles, which are now the most active volcanic features of the region, usually find vent through the unconsolidated deposits 584 CLARENCE N. FENNER that cover the floor of the valley. In most places they are restricted ’ to the valley floor and few are found on even the lower slopes of the adjacent mountains, but the country within a radius of a mile and a half of Novarupta forms an important exception. In this area, hill and valley alike have been greatly shattered, and are crossed by many steaming fissures. This includes not only the portions of the valley to the east and west of Novarupta, but also Baked Mountain, Broken Mountain, Falling Mountain, and some of the lower slopes of Trident. This area was undoubtedly a scene of the greatest activity during the eruption and is still the site of many fumaroles. Evidently the strains that were here set up in the outer crust have been of sufficient magnitude not only to cause fissures to break through the valley floor but also to shatter the adjacent mountains. In most places, however, they are so restricted to the floor that this topographic depression was evi- dently a controlling factor, and hence a moderate depth for their place of origin is implied. The phenomena suggest what might be expected from the injection of a sill under a rather small thick- ness of cover. The fumaroles were the chief subject of investigation by Dr. Allen and Dr. Zies. They made many measurements of tem- peratures and collected samples of gases. for analysis, and much of interest may be expected when their work is completed. At present the account will be confined to a slight description of a few of the features of the fumaroles. In temperature they run from below the boiling-point of water to a heat more than sufficient to melt lead and zinc. The highest temperature found was 645° C. Among the evolved gases, water usually forms more than 99 per cent. The remainder is mostly hydrogen sulphide, nitrogen, carbon dioxide, and methane. Hydrochloric and other acids are probably present, though in small amounts. Around the vents sulphur is often found in quantities as a sublimation product, and pyrite in finely divided form is very common. Ammonium chloride also has been collected, as well as crystallized hematite and magnetite, and study of the crusts brought home will probably reveal other fumarolic sublimates. The vents are frequently alined along fissures half a mile to a mile THE KATMAI REGION, ALASKA 585 in length. The velocity of the outpouring gases is seldom very high; commonly their escape is attended by a hissing sound at small vents and a subdued roaring at large ones. A close view of a vent of moderate size is shown in Figure 12. It should not be inferred that all of the water that is evolved is of magmatic origin. The ashy and pumiceous material that forms the upper part of Fic. 12——Fumarole No. 42. Baked Mountain and Broken Mountain in the background, and the volcanic range in the far distance. Photograph by E.G. Zies, 1919. the conduits is soaked with water, and considerable quantities must be vaporized and carried out with the hot gases. FALLING MOUNTAIN In the upper part of the valley and not far from Novarupta is Falling Mountain, so called from certain remarkable phenomena which it exhibits. Its northerly face is an almost cliff-like slope, probably 2,000 feet in height, which has plainly been produced by recent slumping off of masses of rock. The volume of material thus removed must have been enormous. At the present time blocks or small masses of rock drop off at short intervals and plunge down the slopes with a succession of sharp crashes, and a 586 CLARENCE N. FENNER talus pile of considerable size has been built up by such accumula- tions, but this is of insignificant magnitude in comparison with the total quantity that has been lost to the mountain. Strangely enough, there is little hint as to what has become of this mass of rock. The mantle of ash and pumice that covers the floor of the valley at the foot of the mountain spreads its smooth contours over the whole surface. I think we must conclude that the great rock- avalanche at Falling Mountain was one of the first events accom- panying the recent outbreak of volcanic activity, and that it occurred under such conditions of forcible disruption and violent movement that the material was spread widely over the valley ~ floor. The subsequent deposits of ash and pumice, which here are of very great thickness, smoothed out the irregularities left in the surface of the transported material. The rock-falls that we now observe are probably of the nature of after-effects. Remnants of the fissures that were formed at the time of the original disturbances now afford passages for gases and vapors from below. Along these channels the andesitic wall-rock has been powerfully acted upon and transformed into porous aggregates of new minerals, whereby the rock rapidly loses its cohesive strength. It is not surprising to find that among the new minerals tridymite is promi- nent. The conditions are those under which its formation (as a metastable product) is to be expected. A noticeable effect also is the replacement of many of the pyroxene phenocrysts by aggregates of hematite scales. . These alterations seem to be explicable only on assumptions of rather wide-reaching significance. Apparently the gases that per- meate the rocks and that manifest themselves at the surface by the slowly rising vapor clouds are capable of reacting with the constituent minerals in such a way as to form volatile com- pounds, and the porosity indicates that quantities of material have actually been removed by gaseous transfer. The results of similar processes are visible around the vents of many of the fumaroles on the floor of the valley. Here also there is evidence of the transportation of material in the gaseous medium, and we observe the results of reactions induced by rapidly changing con- ditions of temperature and composition as the gases approach the THE KATMAI REGION, ALASKA 587 outer surface. The most striking result is the deposition of iron compounds around the vents and in the steamy areas—pyrite, hematite, and magnetite—in quantities which, from the observa- tions of Dr. Allen and Dr. Zies, must be very great in total amount. When there is brought before us in such striking fashion evidence of the ability of these volcanic gases to transport material, we are naturally led to a consideration of the various circumstances attending the evolution of such gases and the effects that are likely to be accomplished. One query that arises is as to the results of the continual outpouring of such great volumes of vapor as rise from the neighboring peak,- Mount Martin, and we may ask whether significant changes of composition are not thereby effected in whatever material may lie at the source from which this vapor proceeds, whether it be a body of magma or material of another sort. Unfortunately, insufficient knowledge of the composition of gases rising from the actual throat of a volcano, as well as of their amount and the length of time over which their escape con- tinues, involves the subject in so much uncertainty that definite conclusions as to the quantitative importance of this process are, as yet, hardly warranted. THE NEW VOLCANO NOVARUPTA Near the head of the Valley of Ten Thousand Smokes is the site of Novarupta, a small parasitic vent, which evidently was an exceedingly active volcano during the general eruption, and threw out great quantities of fragmental material, chiefly pumice. Much of this is in much larger masses than those thrown out by Katmai. One such projectile, found about a quarter of a mile away, had a diameter of eight feet. The last act of the vent was to extrude a mass of stiff, viscous glass, which, as it was slowly thrust upward, broke into huge blocks. From a distance this pile of steaming lava-blocks, which is about 800 feet in diameter and 200 feet high, resembles an enormous ash heap. It is surrounded by a circular crater-wall composed of ejected fragments, which is much cut up by actively steaming fissures (see Fig. 13). The material of this is mostly pumice and obsidian, but there are also pieces of shale and sandstone and of dense andesite. The question as to the 588 CLARENCE N. FENNER manner in which this new vent was developed in the fioor of the valley is important. There are undoubted evidences of explosive action, but nothing that may not well be attributed to actions going on after the vent had been opened. What we have to account for here is the formation of a rather small, circular orifice, through which a great amount of pumice was ejected and a small amount of lava was extruded, situated in an area which is much ve Fic. 13—Profile of lower part of Novarupta, and a portion of the inclosing crater wall. Photograph by P. R. Hagelbarger, 1918. fissured but in which other vents of comparable size are lacking. The formation of an orifice of this description is sometimes attrib- uted to the assumed ability of a subterranean body of magma to perforate by explosive action a great thickness of overlying strata and form a cylindrical pipe or conduit (diatreme) for the escape of lava. It is difficult, however, to form a conception of the manner in which such action has been carried out without attributing to the magma properties for which the evidence seems insufficient. It would be necessary to assume not only that an enormous THE KATMAI REGION, ALASKA 589 amount of energy is set free with great suddenness in a narrowly confined space, but that in some manner this force is given a definitely directed tendency upward. The inherent difficulties in this conception are so great that we naturally look for a simpler explanation. We might suppose, as a second possibility, that the underlying magma possessed great expansive powers because of dissolved gases which were struggling to escape, and was thus enabled to effect an upheaval of overlying material, but the natural result of this would be the upturning of huge blocks over a rather wide area, and the escape of pumiceous material from many widely open fissures, accompanied by the ejection of portions of the fractured blocks in large masses. No such evidence is visible around Nova- rupta. The ejection of pumice seems to have been confined mostly to a small opening, and there is no hint in the surface contours of the ash that a widespread chaotic upheaval of strata occurred. Moreover, the largest pieces of ejected sediments found were about the size of one’s fist. We turn, therefore, to a third hypothesis, which is really the simplest of all: that fissuring was first produced, either because of regional strain or because of hydrostatic pressure due to the injection of a sill; and that the magma rose along such fissures in much the fashion that any liquid might do, except that a certain amount of solution was effected, and that near the surface a sudden conversion into pumice resulted in the violent abrasion of the walls. Under this conception Nova- rupta would simply represent a channel along one of the fissures, where chance conditions made escape specially favorable and which therefore tended to enlarge the conduit rapidly and establish more direct connection with the body of magma below. It will appear farther on in this article, as various topics are discussed, that none of the phenomena of the Katmai eruption seem to indicate that these magmas exerted great explosive or expansive powers at depths within the earth, and Novarupta conforms to this idea. Undoubtedly explosions of a violent character occurred here after the magma had reached the surface, but no evidence was found of such explosions prior to its ascent. 590 CLARENCE N. FENNER The banding present in the lava of the dome that now rises above the vent gives very direct evidence regarding the mechanism of its extrusion. This banding is visible at short intervals around the outer circumference of the dome, where masses of rock in place protrude through the general heap of disrupted blocks, and its direction is found to be parallel with the circular outline of the dome. ‘There is also very good evidence of a process of exfoliation of the outer layers as the central core was forced upward.* The character of fracture surfaces of the lava-blocks of the dome is Pos Fic. 14.—‘‘Cornice structure” in Novarupta lava, produced by fracturing and viscous yielding of the hot mass. Photograph by R. F. Griggs, 1919. interesting. It shows that many of the fractures occurred while the glass possessed properties of both brittleness and viscosity, such as are shown by stiff tar. This resulted in effects of the kind shown in Figure 14, where the surfaces formed by intersecting fractures have become wrinkled and fluted. The term “cornice structure’’ suggested itself at once as appropriate for such features. Many bread-crust bombs are found in the vicinity of Nova- rupta. These were ejected from the crater as masses of non- vesicular, plastic glass, and the vesicularity developed during the « Compare Harker, The Natural History of Igneous Rocks (1909), p- 58; Fig. 8. THE KATMAI REGION, ALASKA 591 short interval of time in which the rapidly cooling mass possessed a rigid crust and a plastic interior. The point to be noted is that the magma, rising into the crater from the depths below, did not immediately puff up into pumiceous masses, but accumulated, at times certainly, in pools of non-vesicular lava. The surprising thing is that this condition held in spite of the relief of pressure. On the other hand, we have evidence that under certain conditions very great pressure did not avail to hold the gases in solution, This is furnished by the lava that was later extruded and now forms the dome. In spite of the enormous pressure to which this Fic. 15 Mount Katmai, from the Island Camp. The crater pit extends across nearly the whole space between the two summits. Photograph by R. F. Griggs, 1917. was subjected during extrusion, the contained gases came out of solution and filled the glass with minute vesicles. The explanation of such phenomena as these will be undertaken later. MOUNT KATMAI AND ITS EJECTA Let us consider now some of the features of Mount Katmai, and first the form of the crater as it appears since the eruption. The present appearance of Katmai is shown in Figure 15. Before the eruption the height, as shown by the Coast and Geodetric Survey chart, was 7,500 feet. The top of the mountain has now disappeared and an enormous crater abyss has been formed. 592 CLARENCE N. FENNER Measurements made by Mr. C. F. Maynard, topographer of the 1917 expedition, give the dimensions of this pit as 2 to 23 miles in diameter and 2,000 to 3,700 feet in depth. About one-half of the area at the bottom is covered by a sheet of water of a peculiar, milky, turquoise-blue or green color, and from near the center oi this lake rises a crescentic island. To one standing on the edge of the pit the cliffs appear almost vertical, but their inclination is _ probably not more than 60° to 70° on the average. They seem to be made up entirely of a succession of lava flows. On the western side of the rim, for about one-third of the circumference, an ice wall appears—a survival of beheaded glaciers, and the depression in the southern side of the rim is floored with bowlder deposits of morainal origin. The bottom of the crater, where not covered by the lake, appears from above approximately flat. At the foot of the cliffs are talus deposits, which appear of rather insignificant proportions. At present the activity is very slight. Steam rises slowly from a number of fissures and clefts near the bottom, and the water of the lake is evidently warm, but on August ro snow was lying in many places on the crater floor. The crater of Katmai is a most wonderful and impressive sight, and photographs give but a very inadequate idea of its tremendous proportions (Fig. 16). | A matter of great interest is that of the mechanism by which this huge pit was formed, for this is intimately related to the question of the volcanic processes attending eruptions. Professor Griggs had recognized the importance of solving this problem and had called particular attention to it before our departure for the region. One’s first view would naturally be that the material was blown out bodily in the eruption, but there is good evidence that this is not the whole explanation. A remarkable characteristic of the ejected material is the small dimensions of the fragments. Even on the upper slopes of the mountain there are not many pieces above the diameter of a few inches, and a great proportion of them are much finer. Moreover, almost all the larger pieces are of pumice, and the fragments of older rock have a general maximum size even less than the figures given. Also the propor- tion of these older andesites is rather small, not nearly sufficient THE KATMAI REGION, ALASKA 593 to account for the mass that has disappeared. Their presence is not likely to be overlooked, as their color—dark red to nearly black—renders them conspicuous objects among the accumulations of light-gray pumice. Since this explanation. will not account for the total mass of rock that has disappeared, two other possibilities should be considered: first, solution of the older rock in the new magma; and second, crater subsidence. Respecting solution, it is believed that this process was very active. That this had occurred was suspected several years ago, as specimens of pumice collected by Professor Griggs on former Fic. 16.—The crater pit of Katmai. Topographic measurements indicate a diameter of 2 to 23 miles, and height of cliffs as measured from the level of the lake as 2,000 to 3,700 feet. Photograph by J. D. Sayre, toro. expeditions had been examined microscopically by Professor W. J. McCaughey, of Ohio State University, and the presence of basic phenocrysts in the acid magma and the evidences of instability that they manifested had been noted by him. This has now been confirmed independently and much additional evidence has been secured. Various stages of digestion can be followed until the point is reached where quantities of phenocrysts of hornblende, pyroxene, magnetite, and rather basic feldspar are left undissolved in the glassy matrix. In such instances the original groundmass of the basic rocks has been completely dissolved and the pheno- crysts are corroded. Plainly such phenocrysts are out of place in 504 CLARENCE N. FENNER a rock of the composition of that in which they are found. More- over, pieces of banded and variegated pumice are common, such as those previously described as occurring in the Valley of Ten Thousand Smokes (see page 583), and are attributed to solution of basic rock in the new siliceous magma just prior to ejection. The evidence on these matters will be discussed more fully a little farther on. It appears to show that the new magma, when it rose into the crater, possessed a sufficient degree of superheat to cause it to attack corrosively the basic rock of the crater walls, and, within a brief period, to effect sufficient solution to permit the dispersal throughout its own mass of the basic phenocrysts derived from great quantities of foreign rock-material. The heat requirements seem to demand that in addition to the original store large accessions should have been received, possibly from rising gases.” ‘In any case a new synthetic magma is believed to have been formed in large quantities. The rapidity of destruction of the walls would be attributed to a combination of the shattering effect of explosions and the corrosive action of magma lying in a pool in the crater. The disappearance of much of the rock of the walls may thus be accounted for. Whether all of it may be accounted for by this process and by the ejection of fragments of undissolved rock is not certain. Further evidence will be obtained from analyses (which Dr. Allen has undertaken) of selected material representative of the new magma, little affected by digestion of basic rock; and of other material, representative of the average result attained by the digestion of foreign matter. The alternative hypothesis, that of crater subsidence, is one in regard to which little or no direct evidence has been observed. Apparently rock-slides of considerable importance have occurred at several places in the crater, due to the failure of the vertical = Daly’s discussion of the effect of rising gases (Igneous Rocks and Their Origin, p- 267) is rather misleading. It is true that a bubble of gas, expanding and doing work, loses energy approximately equivalent to the work done, but if the work be expended in producing viscous flow in the surrounding magna, the energy lost by the gas is taken up by the magma, and the system as a whole neither gains nor loses. We may therefore disregard expansion and look upon gas rising from below into a cooler region as a source of heat. THE KATMAI REGION, ALASKA 595 cliffi-walls to support themselves, but this may be quite independent of a general subsidence of the floor of the crater. We know from Martin’s account’ that natives who had apparently fled from the region almost at the beginning of the activities reported that the top of the mountain was gone. We should hardly expect crater subsidence to take place at this early stage. On the whole, it seems that this idea should be applied only if the process of solu- tion, which, in any case, seems to have occurred on a large scale, appears quantitatively inadequate to account tor all the material that has disappeared. At present it seems best to postpone judgment on this until more information is obtained from the analyses. EVIDENCE AS TO THE NATURE OF THE ERUPTIVE PROCESSES A study of the ejecta from Katmai and of the characteristics of the deposits that they form supplies considerable additional information on the eruptive processes. When seen in undisturbed deposits at a distance of, say, eight or ten miles from the crater, the ejected matter forms well-defined strata, such as are shown in Figure 17. The component material is chiefly a light-gray pumice in pieces whose dimensions are three to four inches as an ordinary maximum, and run from this to a very minute size. Mingled with this are specimens of banded pumice, dense obsidian, stony andesites, sedimentary shales, and a sort of volcanic conglomerate. Some of the features of these, and their significance, have been touched upon before but will now be considered in more detail. It was pointed out, in discussing the great sand-flow in the Valley of Ten Thousand Smokes, that the banded and variegated character of some of the pumice indicated a mixture of basic material with the new siliceous magma shortly before extrusion. In such specimens, sharply defined bands adjacent to each other show such differences of composition as are indicated by Dr. Allen’s determination of 74.70 per cent silica in one and 60.40 per cent silicain another. The material thrown out from Katmai crater con- tains similar specimens. The dark bands consist of partly digested basic rock with large quantities of minerals appropriate to andesites. 1G. C. Martin, Nat. Geog. Mag., Vole XXIV (February, 1913), No. 2, p. 147. 596 CLARENCE N. FENNER A lack of homogeneity on a somewhat larger scale is indicated by the fact that pieces of pumice from the same stratum of the ash- fall show considerable variation from one to another in the amount of basic phenocrysts they carry. These features are significant. It seems as if turbulent motion in the lava when it was liquid would have destroyed such inhomogeneities, especially the sharply defined banding; hence, that solution occurred subsequent to the Fic. 17.—The stratified Katmai ash-fall, 8 to 10 miles south of the mountain. Note the sharply defined character of the strata. The heterogeneous material on top is the result of a small landslip since the deposition of the ash and pumice. Photo- graph by B. B. Fulton, tors. rise of the lava and while it was standing in a pool not violently agitated. Evidently the magma did not become inflated at once when pressure was removed in the depths of the earth, but rose as a liquid and stood for a certain period in contact with foreign material upon which it acted corrosively. A careful study has been made to determine the probable source of this basic material. From samples of the ashy strata taken in the field the pumiceous portion, which is greatly preponderant, has been separated. ‘The residue is found to consist principally of THE KATMAI REGION, ALASKA 597 dark, dense material in small fragments. These, to the number of hundreds, have been studied under a binocular magnifier. Their indentification has not been difficult. Sediments excluded, they consist usually of andesites containing small phenocrysts or groups of phenocrysts of plagioclase, hornblende, pyroxene, and mag- netite in a felsitic groundmass. In some, however, the pheno- crysts are absent. Most are dense but some are vesicular. In microscopic section the small feldspars of the groundmass are fre- quently seen to be arranged in flow-lines. It is evident that some of the specimens belong definitely to surface types of rocks and all of them may well be such. The possibility that hypabyssal rocks also may be present cannot be wholly excluded, but no evidence of this origin for any of them has been recognized. In composition they are medium to basic andesites, apparently no different from the rocks that form the walls of the crater pit. Evidence is at hand regarding the absorption of these andesites by the new magma, but before we proceed to consider this matter it is necessary to digress for a moment. As previously indicated, it is supposed that the fragments of andesite found in the ash-fall, or at least a large proportion of them, represent wall-rock that collapsed and became immersed in the pool of lava. One might expect, therefore, that similar pieces would frequently be found as inclusions in the pumice. This does not seem to be the case: on the contrary, their mode of occurrence is nearly always as detached particles. This is a matter that requires examination, and several features have been noted which have a bearing upon the subject. It is found, first, that not only are inclusions of andesites rare in the pumice but likewise inclusions of shale and of all other dense material except the phenocrysts; second, many of the fragments, although not now inclosed in pumice, present evidences of previous immersion, such as films of glass adhering to their surfaces, and corrosion effects; third, although the pumice does not carry inclusions, the obsidians, which must have been derived from the same lava-pool as the pumice, carry great quantities of both shale and andesite. From these facts it seems that the conclusion to be drawn is not that the andesites were never immersed in the magma, but that, in 598 CLARENCE N. FENNER the violent explosions, the frothy, semi-liquid pumice and the dense, rigid andesites reacted differently and were forcibly torn apart. The forms of many of the fragments of andesite are such as to suggest fracture by the explosions. If the forces acting upon them were of such magnitude as to produce fracture, it hardly seems surprising that they became separated from the pumice. Pieces of obsidian, of ordinary maximum dimensions of three to four inches, and of angular shape, were found everywhere in the Katmai ash-fall within the area of coarse ejecta. Many specimens were collected for study, and they furnish interesting information. It is difficult to conceive any origin for them other than the lava pool that gave rise to the pumice; and the presence within them of pieces of pumice (which would inevitably float on the lava) and their general nature suggest that they represent chilled crusts on the surface of the lava. They contain quantities of inclusions of various kinds: sediments, andesites, other obsidian, pumice, and separate crystals. If it be granted that they are fairly representative portions of the lava, the inclusions they con- tain ‘‘frozen in”’ in all stages of disintegration are of great instruct- ive value. Another material present in the ash-fall is evidently closely akin to this conglomeratic obsidian. The matrix is semi-pumiceous to glassy, and the numerous inclusions are of the same sort as those in the obsidian. It is interpreted as a surface scum formed over areas of seething lava. The inclusions in this also have been caught in various stages of disintegration. From the evidence presented by these specimens, the absorption of xenoliths by the magma seems to have taken place in several somewhat different ways. The chemical composition of the irag- ments and their porosity were probably important factors in the matter. In some instances peripheral solution, especially of the groundmass, seems to have been the principal process, or well- defined tongues of lava may cut off portions and allow them to float away. Probably a more usual form of attack is one involving an intimate penetration of the whole mass. Several factors may have been involved in this: first, that portion of the groundmass of the rock that consolidated last of all and forms a binding-material THE KATMAI REGION, ALASKA 599 among the grains may have had a melting-range little above the temperature of the new magma, and the penetration of the latter -and of its vapors was therefore easy; second, an original vesicu- larity may have been present; and third, a porous condition may have been developed during the history of the rock. The third feature is important. It is not hypothetical, but rests upon observation, and is believed to have considerable significance. It has been found that many of the andesitic fragments in the pumice- ous strata have at some former time undergone a process of altera- tion similar to that described for the rocks of Falling Mountain. Quantities of minute, glistening scales of tridymite have been formed, and a replacement of ferromagnesian minerals by hema- tite is observable. This carries several implications: first, these mineral transformations are such as might be expected to result from fumarolic action along fissures in the walls of a crater, but not of the kind that would be looked for at great depths; second, the presence of such fissures and the mineral transformations along them would aid in the collapse of the walls during the activities of the eruption; and third, these altered rocks would be more susceptible to penetration by the magma on their immersion in it. When the process of penetration of magma into porous xenoliths has been thorough, they appear to have become pasty throughout, and what we find is an irregular, lumpy mass, or clot, consisting of basic minerals in a dark groundmass. Under the microscope the phenocrysts are seen to be much corroded. They often contain a great number of inclusions of brown glass, which fairly riddle them, and they look as if they were disintegrating. The groundmass is essentially glassy but contains a multitude of small, irregular frag- ments or splinters of crystals, and much brown dust. The low index of this glass indicates an acid material in spite of the dark color, and it seems doubtful whether the amount of material actually fused or dissolved was large; the process was rather one of intimate penetration by the new magma, resulting in separation and dispersal of the component minerals and a partial breaking up of crystal units. The low silica-content found by analysis (as in the specimen illustrated in Fig. 11) and the dark color are probably due to undissolved phenocrysts and dust. When the 600 CLARENCE N. FENNER penetration by the magma reached a fairly advanced stage before the final inflation occurred, these dark masses as well as the light glass assumed the porous condition; both must have been charged with vapors. : Although, under some circumstances, the bands or schlieren that have arisen from these pasty masses remain sharply distinct for several inches and perhaps much more, it is not unusual, on the other hand, for bands to disappear within a short distance. The facility with which the banding has become obliterated in these cases shows that its sharp definition is not a property that persists in spite of turbulent movements. In the pumice of the early strata of the ash-fall, phenocrysts are almost lacking; in later strata they become exceedingly abundant. Their appearance at this later stage is ascribed to the setting free of phenocrysts from the andesitic wall-rock in the manner described. Their character in the two environments has been carefully com- pared. Those in the andesites have been studied in microscopic sections, and also under a binocular magnifier in such specimens as show surface corrosion, which has left them in partial relief. The phenocrysts of the pumice have, many of them, been set free by explosions and occur loose in the ashy strata. These are easily studied with the binocular magnifier. Others, still inclosed within a matrix of pumice or obsidian, have been examined in thin sec- tions. In the andesites the phenocrysts occur both as isolated crystals within the felsitic groundmass and as aggregates of the kind that Judd has called glomeroporphyritic groups, and have certain characteristics in regard to size, form, and grouping. The component minerals are pyroxene, hornblende, plagioclase, and magnetite. In the pumice we find the same minerals, isolated or in the same sort of groups as before, apparently duplicating in all respects the phenocrysts of the andesites. Let us review briefly the evidence that has been presented on this matter. An immense amount of material has disappeared from the top of the mountain and from the crater walls, and must be accounted for. In the pumiceous strata fragments of andesites are found that have the characteristics of surface-flow rocks, and correspond to what is known of the rocks in the crater walls. The THE KATMAI REGION, ALASKA 601 evidence of fumarolic action that many of them bear is in accord with such a situation. These might account for the rock that has disappeared except that they are quantitatively insufficient. We are left, then, to consider crater subsidence versus incorporation in the new magma. Without trying to decide in this article whether a// of the material may be accounted for by incorporation, the evidence that a large quantity has been taken up in this man- ner has been considered. According to this evidence, numerous specimens of andesite show attack. The processes involve either an intimate penetration and consequent softening of the whole mass, followed by dispersal of the phenocrysts; or the breaking up of the fragments by attack along fissures, simultaneously with solution of the groundmass around the periphery of the fragments and eventual setting free of the phenocrysts. Finally, multitudes ot phenocrysts of the kind that the andesites carried are found to appear in the later strata of the pumice, though the earlier strata are practically free from them. Their instability with respect to their surroundings is indicated by the active disintegration that they are undergoing. The fact that quartz phenocrysts properly belonging to the magma have no association with them is also significant. These facts taken together seem to form strong evidence identifying the phenocrysts of the pumice with those of the former wall-rock, and the disappearance of large quantities of wall-rock is thereby accounted for. Some of the materials in the ash-strata deserve further atten- tion. The obsidians that have been described, when heated in the laboratory, swell up to a frothy white pumice closely resembling the pumice found in the field. When their powder is heated in a closed tube it yields water, hydrogen sulphide, hydrochloric acid or a chloride, and some gas having a fetid organic odor. It is planned to investigate these gases with care. Although many phenocrysts of extraneous origin are found in the pumice, the only phenocrysts that properly belong to the magma are quartz and acid plagioclase. These had probably crystallized out before the magma rose into the crater. ‘The quartz, which is easily recognized, is never associated with the groups of xenocrysts mentioned, but, on the contrary, is found in the purest rhyolitic 602 CLARENCE N. FENNER phases of the pumice. Its presence supplies information regarding the upper limit of temperature within the magma chamber just prior to extrusion. The transition point between quartz and tridymite at atmospheric pressure is 870°C. Probably great pressure will have a perceptible effect in shifting the inversion- point—a thickness of 20,000 feet of rock strata might possibly raise it 1ro0o°—but we can be tals sure that a temperature of less than 1000° prevailed. The distinctly stratified form of the ash-fall, with its indications of a waning and renewal of activity many times repeated, harmo- nizes with the other evidence presented that the melted rock accumu- lated in a pool in the crater rather than that it was discharged as a continuous stream as soon as some hypothetical obstruction, which had previously restrained its escape from the depths, was removed. From the evidence that has been brought together certain deductions may now be made. We see that the magma, as it issued from the depths of the earth, did not at first show a tendency to evolve its gases explosively; that is, did not have an extremely high vapor-pressure; but that this was developed after a short period of standing under the new conditions, and explosive erup- tions ensued. From this we conclude that in the enormous change of conditions consequent upon rapid extrusion internal equilibrium did not keep pace with external changes, and that prior to extrusion such internal combinations prevailed that the tendency of the gases to escape was not extremely great. As an indication of the depth in the conduit at which explosions occurred the relative amounts of the various sorts of foreign material ejected with the pumice is of interest. Pieces of andesites from the crater walls are very common; fragments of shale and sandstone from the sedimentary platform are frequently found, but the quantity is not so great as of the andesitic rocks; and pieces of deep-seated granitoid rocks are almost lacking, though a few specimens were found. This relative abundance seems to show that the violently explosive action was exerted only at the surface or at a moderate depth in the conduit. If the renewal of activity at this vent after a long period of quiet were due to an accumulation of imprisoned forces until they reached a magnitude where they were capable of blasting away THE KATMAI REGION, ALASKA 603 obstructions, the bottom layer of ejecta should be made up largely of material from such a source. As a matter of fact this bottom layer is essentially pumiceous and actually appears to be more nearly free of foreign material and more nearly of rhyolitic composi- tion than any layer above it. The view that the close of a period of activity at a volcanic vent is attended by the formation of a plug of lava which seals up the conduit and that the renewal of activity necessitates the clearing out of such a plug, finds little to support it here. Nor, I think, does a consideration of events in certain other explosive eruptions. lead to views different from those expressed for Katmai. Many instances might be cited in which for months previous to a paroxysmal eruption manifestations have occurred, such as outbursts of gas and ashes, that can hardly be looked upon otherwise than as indicating a quite direct connection between the surface and the subterranean activity. It seems not unusual, too, for lava to appear in the crater and remain com- paratively quiet for a certain period before explosive inflation occurs. The great eruption at Krakatoa in 1883 seems to have followed such a course." At Pelée also there were premonitory symptoms, consisting of an increased evolution of vapors, at times mixed with cinders; later the moderately explosive (though immensely destructive) ejection of the nuées ardentes, accompany- ing the rise of lava in the crater.» A somewhat similar course of events may be found in Koto’s description of the eruption of Sakura-jima.s By what means a volcanic vent can remain sufh- ciently open to permit a free escape of vapors, without allowing magma to issue, and what conditions finally bring this period to a close and cause a body of gas-charged, actively corrosive magma to appear are matters whose explanation presents many difficulties. No theory of volcanism that I have seen appears at all adequate to account for the phenomena. Indeed, some of the fundamental concepts of current theories seem irreconcilable with them. Objection may be raised to the somewhat novel idea that has been presented here, of a state of unstable equilibrium of the tJ. W. Judd, “‘The Eruption of Krakatoa and Subsequent Phenomena,” Report of the Krakatoa Committee of the Royal Society, pp. 11-20. 2 A. Lacroix, La Montagne Pelée et ses Eruptions, pp. 35-30- 3B. Koto, The Great Eruption of Sakura-jima in 1914, pp. 56-82. 604 CLARENCE N. FENNER magma, and the question may be asked as to what combinations are supposed to be entered into between the volatile and non-volatile constitutents that would give the required effects, also as to the nature of the changes in, the physical environment by which the condition of unstable equilibrium is brought about. These are natural queries, and answers would be eminently desirable; never- theless, it is believed that the case rests not upon the ability to answer them but rather upon the plain evidence of the phenomena themselves. It may be of assistance, however, to a comprehension of what is meant by unstable equilibrium of the magma, to con- sider certain familiar phenomena exhibited by obsidians. Any obsidian which, when heated, puffs up into pumice, shows charac- teristics allied to those that I have ascribed to the Katmai magma. The behavior of the obsidian in this respect indicates that it like- wise in its past history underwent changes of condition in which internal equilibrium failed to keep up with external changes. If this were not true it could hardly have retained its dissolved gases but would have evolved them during cooling. In instances of this kind the lack of equilibrium continued even beyond the stage which the Katmai lavas reached, and finally all possibility of evolving gases disappeared because of increasing rigidity, but this result was probably due to factors (such as rate of cooling) which may well be variable. In the case of Katmai and other volcanoes, it seems reasonable to suppose that the magma first experienced very rapid change of conditions during its rise in the conduit, but then remained for a certain period in comparative quiescence, and thus opportunity was given for approximate equilibrium to be reached before a condition of prohibitive rigidity had set in." From evidence of the kind given it appears that examples of unstable equilibrium in magmas, due to sudden changes of environ- ment, are not at all uncommon. Recognition of this fact and of what it connotes may be helpful in directing inquiry into the con- ditions that have brought it about. «Interesting examples of the effect of rate of cooling upon the final product in somewhat analogous systems are furnished by Morey’s experiments on hydrated alkali silicate melts prepared in steel bombs. Rapid quenching gave a rigid, hydrated, unstable glass, while slow cooling caused the expulsion of dissolved water and the formation of a pumice. See G. W. Morey, Jour. Amer. Chem. Soc., Vol. XXXVI (February, 1914), p. 226. THE KATMAI REGION, ALASKA 695 SUMMARY AND CONCLUSIONS A preliminary account has been presented of observations made by the writer as geologist of the expedition sent in 1919 by the National Geographic Society, in co-operation with the Geophysical Laboratory, to the Katmai region. As a result of this work many of the observations made by the director of previous expeditions have been confirmed and supplemented. With regard to one or two others, somewhat different interpretations are given in this article from those of previous publications, but it is believed that on these matters also all concerned are now in agreement. With respect to still other phenomena, which had not been previously described, evidence has been found that affords a basis for extending con- siderably our ideas respecting the processes at work during the eruption. It has been found that the volcanoes of this region, which form a continuation of the Aleutian loop or festoon, are situated in an area of sedimentary rocks remarkable for the absence of folding or obvious faulting. The more recent lavas are basic andesites, contrasting greatly in composition with the highly siliceous rhyolite of the last eruption. In the area of the Valley of Ten Thousand Smokes, it is believed that the injection of a sill or closely similar body of magma into the underlying strata at the beginning of the eruption caused shattering of the rocks above it, and these openings permitted the ascent of magma. The extrusion and inflation of this magma gave rise to a great ash- or sand-flow, analogous in many respects to the nuées ardentes of Pelée and La Soufriére, and led to the formation of the parasitic cone of Novarupta. The fumaroles are thought to be due to the continued evolution of volatile constitu- ents from this body of magma. The development of the new vent of Novarupta is ascribed to the enlargement of a channel along one of the fissures. The later extrusion of the stiff lava forming the dome of Novarupta is found to have been similar in many respects to that of the ‘“‘spine”’ of Pelée. At Falling Mountain the most interesting features are those resulting from fumarolic action. Evidence of a process of solution and transfer of rock material in the gaseous medium was found 606 CLARENCE N. FENNER here, and the results of similar processes around the vents of the fumaroles in the valley were observable. It is suggested that the properties of the evolved gases indicated by this gaseous transfer may at times lead to results of great importance in volcanic processes. A study has been made to determine the manner in which the top of Mount Katmai disappeared and the great crater pit was formed. It seems quite certain that the material was not blown out directly but must be accounted for otherwise. Crater sub- sidence may have been a factor, but it is believed that collapse of the crater walls and incorporation of the material in the new magma were chief features. It is recognized that the latter process demands a large quantity of heat for its accomplishment, and the magma evidently was not at a very high temperature prior to its ascent; therefore accessions of heat seem to be demanded. A considerable problem is thus presented, but it does not seem at all insuperable, and it is believed that the evidences of solution are so strong that they cannot be disregarded. One of the important features of the eruption brings up for consideration a phenomenon to whose significance little attention seems to have been paid hitherto. It is that of a gas-charged magma gradually developing the explosive condition after some interval has elapsed subsequent to its ascent from the depths. The Katmai magma seems to have followed this course, and the phenomenon is apparently not uncommon. This is believed to have great significance and to imply changes of physical environ- ment during its ascent, effected with such rapidity that internal readjustments were not able to keep pace with them. Many of the current theories of volcanism are based upon a fundamentally different conception of the nature and properties of the magma. It is thought that it may be advantageous in many cases to con- sider matters from the new standpoint here suggested. In other matters also, theories that have been proposed and somewhat widely accepted are apparently not in accord with the evidence found here. It has not been possible in this article to discuss these matters exhaustively, and other matters of interest have not been touched upon. Fuller treatment will be presented in articles to follow. July, 1920 ON SOME PHYSICAL PROPERTIES OF ICE MOTONORI MATSUYAMA University of Chicago INTRODUCTORY NOTE BY T. C. CHAMBERLIN While many of the more obvious problems of ice and ice action have been solved in a general way, there remain not a few questions of a more refined sort which require solution before glaciology can rest on a secure foundation. Some of these questions are critically important for they bear radically on interpretations that have already been widely accepted and are currently taught. More extended and more critical field studies are required to solve some of these questions while the solution of others depends on more discriminative and exact laboratory experimentation. All of them call for more searching analyses of the problems themselves, as a source of guidance in field work and in experimentation, as also in the interpretation of results. The glacialists working at Chicago have been trying to do their bit toward the solution of some of these problems and have had under way for some time a series of attacks along several lines in both field and laboratory. This paper presents the preliminary results of a careful series of laboratory determi- nations carried out by Professor Motonori Matsuyama, of the Department of Geophysics of the Kyoto Imperial University, Japan, who has been spending the year at Chicago. When McConnell, followed by Miigge, announced that ice crystals are minutely laminated in planes normal to their optical axes and that movement along these planes was notably easier than in other directions, it was felt by many that these disclosures offered a happy solution of the anomaly of glacial movement which seemed to be a quasi-fluidal flow in a body obviously rigid. But later critical studies raised serious doubts as to the actual participation of the gliding planes in ordinary glacial motion, and so the subject came to demand more refined examination. Up to date, no one, so far as I know, has determined what is the measure of the resistance to motion along these planes compared with the stresses actually brought to bear upon them in ordinary glacial motion. Nor has it been shown whether the relation of these gliding planes to one another is of the elastic or the viscous order. But even if these properties were known, there would still remain the radical question whether glacial motion actually takes place by means of movements along these planes—or in any other way within the constituent crystals—or whether it is essentially a motion between the constituent crystals. There are here 607 608 MOTONORI MATSUYAMA therefore two quite distinct problems. The investigations of Professor Mai- suyama relate to the second of these. The method of Matsuyama is, so far as I know, unique in that he deals with bundles of crystals which have a common known orientation. He thus brings movement along the gliding planes into experimental competition with movement between the crystals. All are familiar with experiments upon the yield of glacier ice where the mass under test was formed of many crystals of diverse orientation, but this very diversity of orientation stood in the way oi a strict interpretation of the results. Matsuyama, however, so selected his prisms and cylinders as to force an alternative between combined movement on gliding planes and movement between the crystals. His method and his achievement are therefore notable. It is further to be observed that in his experimentation he used the tor- sional method, following Michelson, in which, the errors arising from the stretching or compression of the prisms or cylinders are avoided, since the cross-section remained constant throughout the trial. But as check upon the results of torsion, Matsuyama added the method of bending in which stretching on one face and compression on the other were involved. In the interpretation of the results of this method he called in the resources of the petrographic microscope with the discriminating results given in the text. INTRODUCTION The motion of an ice sheet along the mountain slopes and over a large area of the Continent may be caused by more or less differ- ent forces. Besides the external forces, it is also important to know what is the behavior of the ice itself in such motion. Numer- ous works have been published on these problems, among which those of McConnell and Miigge are famous and have been referred to by many authors. According to them an ice crystal can be sheared more easily in the direction parallel to the basal plane than in any other direction. The elaborate works of the members of the Cornell University geological stafft added important contri- butions to the same line. Their idea is that the gliding planes of ice crystals arranged parallel to their basal sections control the behavior of an ice mass as the main factor. Deeley” calculated the viscosity of ice from Main’s experiment and found its value to be 6X10” c.g.s. at o° C., while his own observations on Swiss glaciers gave 125 X10” C.g.S. ™R.S. Tarr and J. L. Rich, Zeits. f. Glets., Vol. VI (1912), pp. 225-49; R. S. Tarr and O. D. von Engeln, Zeits. f. Glets., Vol. TX (1915), pp. 81-139; O. D. von Engeln, Amer. Jour. Sci., Ser. 4, Vol. XL (1915), pp. 449-73. 2R. M. Deeley, Geol. Mag., New Ser. 5, Vol. IX (1912), pp. 265-60. SOME PHYSICAL PROPERTIES OF ICE 609 Various investigations related to glacier problems were planned and worked on for many years in this department by Professors T.C. Chamberlin and R. T. Chamberlin. The present work is a part of that series and gives the result of preliminary investigations on the elastic properties of ice. Since torsional force applied to a circular rod is the only type of strain, as Professor Michelson’ says, in which the cross-section remains constant, the main part of the present study consists of observations on torsional deformation. Later some attempts were made to determine the Young’s modulus and also to observe what would happen in bending. The preliminary work was begun early in the autumn of 19109, but owing to a delay in securing certain necessary apparatus, it was not until about the beginning of February, 1920, that the work really began in earnest. About the end of March it became so warm that it was no longer possible to continue the work on ice. Because the time available for this cold work was so short, the results are to be considered only as preliminary; yet they were so suggestive that the present writer considered it worth while to publish them even with these limitations. During the research, Professor R. T. Chamberlin took constant interest and gave important suggestions for which the writer is very much obliged. LIMIT OF ELASTICITY According to Professor Michelson? the deformation of an elastic body passes through four different stages as it goes on, and in the first portion the deformation is characterized by being approximately proportional to the stress. The limit of this part is well defined in some materials while in others it is not so dis- tinctly marked. The following description will show that ice also has a rather well-defined limit of elasticity in some cases at least. Method.—In the present experimentation a circular rod of ice was held horizontally with its one end fixed to a rigid stand and the other end attached to a circular disk of radius 7.45 cm. movable t A. A. Michelson, Jour. Geol., Vol. XXV (1917), Pp. 495. 2 Ibid. 3A. E. H. Love, The Mathematical Theory of Elasticity, p. 112. 610 MOTONORI MATSUYAMA around a knife-edge at its center which rested on a horizontal plane. The whole equipment was kept in a thermos box, the temperature inside of which could be kept low and constant by using freezing mixture. The torsional stress was applied by putting weights on the scale pan hanging from the limb of the circular disk through the bottom of the box. The amount of torsion was observed by means of a telescope through a window in one side of the box. In the first set of experiments, the test piece was cut out from artificial ice and consisted of parallel crystals with their optic axes transverse to the longer dimension of the bar. The first bar was 13.15 cm. long and 1.73 cm. in mean diameter. The deformation was found to depend upon the rate of increase of force, especially near and beyond the elastic limit. -In the first experiment the force was increased by adding two pieces of weights, each weighing 4.40gm., every five minutes. The total time duration for this observation was 3.5 hours and meanwhile the temperature was constant at —7°0C. The observed relation between the weight and deformation is shown in Figure 1. From this curve one can easily see that a rather marked change of yielding is recognizable at the point B, up to which the strain is more or less proportional to the stress. This feature is none but the characteristic of an elastic curve’ and we may say that ice behaves like an elastic solid. Later it was learned that the point B was not so well defined in some cases as in the present one. It is for further study to see when this point will be sharply shown and when not. MODULUS OF RIGIDITY Rigidity is the resistance of a material to shearing force and is given by the ratio of the amount of shear to the applied force. If a circular rod of length / and radius a is twisted through an angle ® by an external couple PL, the modulus of rigidity n will be given’? by PL=}an a t Given in every book on elasticity. 2 Poynting and Thomson, Properties of Matter, p. 79. SOME PHYSICAL PROPERTIES OF ICE 611 It was found that it took about 10 minutes to reach the new equilibrium position when an ice bar is twisted by a certain stress within the elastic limit. The deformation in the foregoing obser- vation in the part AB, therefore, must be slightly modified for this effect. Careful observations were made, increasing the load by the same amount at intervals of 20 minutes. As the result it was found that the mean deviation was 5‘2 per 26 gm. on the 32°) ane eos ao ones eee oes SO 0S000 060c0 50055 paste eae esheets feabab lata oh ch Coe Weight in gm. BGeea | re [A ae Torsion in degrees Fic. 1.—Torsion curve of an ice bar, 13.15 cm. long and 1.73 cm. in diameter, with crystals transverse to its longer dimension. Arm of the torsional couple 7.45 cm. Temperature —7°.oC. scale pan, the temperature being kept at —7°5C. From this and other known values of the constants, we obtain N=1.9X 10° c.g.s. The value of 7 was calculated from data in several experiments intended primarily for other purposes. The following table gives the values of 2 thus obtained for ice bars with optic axes of their component crystals transverse to them. Temp. (C.) : < ue India rubber....... 1.6 X10? 5 > fad | be iS} |_| | Eo 2 65H | So =” PENSEEEEEE ° . w | aN bo 50 y | 8 4o PT INS ra RI @ 3 pa o Ay to (e) Torsion in degrees Fic. 9.—Recovery curve for the rod with crystals transverse to its longer dimen- sion. Crosses for constant force and dots for increasing force. 100 > So) oO > fo} o Io = 60 [eae ao) Ec 5 ~NCIo 8D zine S 40 8 2 E oO Lea a 20 E-—- = fo) 2 4 6 8 ae) 12 I4 16 18 20 22 Torsion in degrees Fic. to.—Recovery curve for the rod with crystals parallel to its axis. Crosses for constant force and dots for increasing force. RECOVERY FROM DEFORMATION While observing the deformations by constant force as well as by increasing force, the amount of return toward the original state SOME PHYSICAL PROPERTIES OF ICE 623 after removal of the force was observed in each case. Beside the amount of torsion, the recovery will depend upon various other conditions, like the temperature, mode of applying the force, and its duration. These values are given in the following tables: ConsTANT Force. CrystTALts TRANSVERSE Temperature Weight Duration Deformation Recovery Percentage —9°o0C 22 gm ohr5™ Onnu Ones 100 (054 aimee 44 ii ©) ) (3) @ 13 100 =0).Csiave one 66 On @) Oo 14 71 =O: HFoovsoved 88 2O Oo 32 O 24 75 —=(HaOono tera IIo Be oO 36 oO 26 72 =(O).C)s 6 ane oe 132 345 T2315 © 4I 63 = [3 .didan aoe 154 6 20 2 50 uy @ 41 Qe cist tes ss 176 IO 39 Be 2 Tee? 40 SO cain cas 229 37 20 9 44 ii ats) 13 ConsTANT Force. CrysTALs PARALLEL Temperature Weight Duration Deformation Recovery Percentage =O72Cs. 22.6 5 gm. ohs5™ °° 44’ oy fy 100 5/0 Ole eaves fe 3s 44 I 45 © 10 o 10 100 Oasis oc 88 I5 10 I 38 ih it 77 iC to acne ee 132 us ©) 2 48 Te 30 a Oro 220 29 30 7 12 2 | ae) SST atistale slopes 220 I2 40 5 24 o 48 | 15 SNR bT ye) stasis (ais 308 48 0 i wn Te an(0) 12 = Are Ate als 05,5 | 440 TO I3 38 I 56 14 =F Vomocacog 660 18 oO 20 36 Q ste) INCREASING Force. CrystaLts TRANSVERSE Temperature Increased Rate Duration |Final Weight | Deformation | Recovery Percentage = 425 Cues. .44 gm./min. 13420" 352 gm. a sey! 1 Bet! 19 AWN TI. Slav .88 5 10 264 3 20 I Io 35 => Pa Obdamore 1.76 a 222 PD © 54 44 Ow Ais osc 1.76 ) 1s 660 8 58 2 46 31 =) Glee een DD 2 QO) 440 Ones I 46 29 ee eee 4.4 2 10 572 7] 389} ie | 27 a OB eNe Sie ots 8.8 iz & 572 5 28 i ~ip- | 31 S(t eee 8.8 Tas 572 550 1 SY). || 25 ORO ag. ces D2ROl. © 30 660 5 go | © Ss | 18 = ener 44.0 O15 660 Ao Meee | 30 INCREASING ForcE. CrysTALs PARALLEL Temperature Increased Rate Duration | Final Weight | Deformation! Recovery | Percentage (aloes .88 gm./min. GEE Washo xeaaal, Onsi7/ On251 =| 44 = eSheee 1.76 4 10 440 6 30 2 @) |) 31 GQHOurs css 44.0 O15 660 1 NY ° 36 | 41 624 MOTONORI MATSUYAMA The amount of recovery in percentage of the deformation is shown in Figures 9 and 10. For small forces the recovery is com- plete. As the amount of force which has been used increases the percentage of recovery decreases rapidly until the formation amounts to about four degrees. After this its decrease becomes slower and it approaches gradually to the zero line. Thus it is seen that the transition from the stage of nearly perfect elasticity to that of partial elasticity takes place very slowly in the cases both when the torsion is parallel to the optic axis of the constituent crystals and when it is parallel to the basal planes. BENDING EXPERIMENTS The observations on torsion of ice just presented have suggested that an aggregate of parallel crystals of ice is more easily deformed by a shearing force parallel to the axes of the constituent crystals than parallel to their basal sections. We have also seen that Young’s modulus, or proportionate resistance to elongation which is generally determined by the method of bending, is greatest when the deforming force bending the test bar acts parallel to the basal sections of the constituent parallel crystals. The next step was to test the matter further by observing the behavior of bars of differ- ently orientated crystals when bent beyond the limit of elasticity. It was desirable to try the experiment with bars of different sizes compared to the constituent crystals. But the temperature of laboratory was not favorable for the use of larger bars with larger deforming force. Bars of moderate size with rectangular cross-sections were therefore prepared with different orientations of crystals. Each bar was supported at both ends by knife-edges and a weight of roo gm. was hung from the middle point. The bending was measured by the lowering of this central portion. In Figure 11 some of the results of these observations are given, the data for the bars being given in the following table: Bar Optic Axes Span Breadth Thickness Temperature Ca eaters Horizontal 9.0 cm. .9o cm. .60 cm. —5°8C.to —2%0 C. Chee ee Parallel g.0 .9O 60 —5.8 to —3.0 Dinos eke Vertical g.0 .60 .58 =F 1k) —O.2 Ceasar Parallel 9.0 .60 58 90) 1K) OoO SOME PHYSICAL PROPERTIES OF ICE 625 The first specimen bent much more easily than the second and broke at the point A in the figure. The same applies to the third and fourth specimens. In another case, three test pieces, each o.80cm. wide and o.50 cm. thick, were supported with an 8.0 cm. span between the supports. The lowering of the central part by 100 gm. in the first eight hours was 2.68 cm., 0.2 cm., and 1.22 cm. respectively for the different orientations a, b, and c. A part of the weight 2 | 10 o 8 5 i 2) a= [| eo: _ o = io) = a 4 2 ° ° 2 4 6 8 ate) 12 14 16 Time in hours Fic. 11.—Relative ease of bending of ice bars with different orientations of crystals. hanging from the second bar was found supported from beneath by accident so that its bending is not comparable with others. The temperature during this observation ranged from —5°o hOn-— 257. C. It is clear from these observations that ice bars with their constituent crystals perpendicular to the length, bend and break more easily than the bar with crystals parallel to the length. In the case of observation for the bars 6 and c, the temperature became so high that the bar suffered from pressure melting at the 626 MOTONORI MATSUYAMA knife-edges, thus preventing freemovement. On account of this, the result may not be much relied upon and consequently the relative ease of bending of the bars a and 6 is not determined. The other observation stated above also failed to give any idea about this. As to the nature of deformation of ice comprising of an aggregate of crystals, many authors have claimed that it depends upon the behavior of a single crystal. It is stated that in a case of bending of an ice bar consisting of several crystals, most of the bending had taken place in one of the crystals lying with its crystal axis nearly horizontal and approximately parallel to the length of the bar. This is understood to mean that in such a case movement along the gliding planes of an ice crystal parallel to its basal plane is more effective than movements along the contact surfaces of adjacent crystals. If this theory is applicable to the present case the third bar in which the gliding planes were transverse to the long dimen- sion should bend most easily. But quite to the contrary, bar of orientation c, instead of bending most readily, suffered the least bending of any of the three types of orientation. As to the mechanism of bending within the limit of perfect elasticity it is generally understood by physicists that bending of a bar is caused by shortening of the concave side and elongation of the convex. When bending becomes larger than this, it is difficult to solve the case as a simple mechanical problem. It is probable in such a case that the portion near the point of applica- tion of force is subjected to bending, while the other parts are elongated with some degree of sliding at the knife-edges. The mechanism at the bending-point will depend upon the structure of the material. If it is deformable more easily in one direction than in another the problem becomes complicated. The resulting deformation in such a case will be the combination of that effect with the result of shortening and elongation phenomena. The experimental fact above described that bars of type c bend less easily than the others suggests that contact surfaces between adjacent crystals play greater réle than the so-called gliding planes parallel to the basal plane. R.S. Tarr and J. L. Rich, Zezts. f. Glets., Vol. VI (1911), p. 236. SOME PHYSICAL PROPERTIES OF ICE 627 MICROSCOPIC EVIDENCES The theory that an ice crystal is composed of thin laminae parallel to its basal plane gives rise to the conclusion that when the crystal is bent by force perpendicular to its optic axis, it will show simultaneous extinction at the bent and unbent portions under the petrographic microscope.t. Tarr and Rich? have described the case in which this optical property of the bent bar was not changed, as well as when it was changed. In the latter case the original optic axis was either parallel or perpendicular to the length of the bar, and if we assume the latter as the case, it is in contradiction to the idea that movement along the gliding planes controlled the bending. The present writer examined the bent parts of an ice bar and the results were very suggestive, showing facts which seem not to agree with former ideas. A bar of ice with crystals parallel to its length was bent under certain stress. When the bent portion was thinned down and examined under the microscope with the Nicols crossed, it clearly showed an extinction strip across the bar, which moved along the bent portion as the stage was turned. Wanting to be sure about this, the writer asked Professor R. T. Chamberlin to see it and he recognized the same fact with cer- tainty. The test piece in this case consisted of one main crystal with small portions of other crystals on either side. The same fact was observed in one more case but with the oncoming of warm weather these two were the only trials which it was possible to undertake in the present investigation. ‘Examinations without crossed Nicols revealed another impor- tant fact. The test specimen consisted of parallel crystals whose optical axes were horizontally transverse to the length. The bent | portion was examined under microscope so as to see the side of the bar, i.e., the basal planes of the crystals. In the field of the microscope, it was found that faint but distinct straight lines nearly parallel to each other had developed on the sections of the crystals. The boundaries of the crystals were zigzag and some- times the straight lines developed in the crystals were observed to =R.S. Tarr and O. D. von Engeln, Zeits. f. Glets., Vol. [X (1915), p. 111. 2R.S. Tarr and J. L. Rich, Zeits. fMGlets., Vol. VI (1911-12), p. 235. 628 MOTONORI MATSUYAMA start from the angular points of these zigzag boundaries.. The direction of the parallel straight lines was not the same in different crystals. Sometimes apparently two systems of these straight lines developed in one crystal, nearly perpendicular to each other. When the unbent portion of the same bar was examined in the same way, the crystals were found to be bounded by very smooth boundaries and no straight lines in their sections were visible. In another case, the same crystals were identified before and after bending took place. The same contrast of the disturbed and undisturbed crystals was observable in this case. At the time of these observations, the equipment for petro- graphic photography was not available for the writer. He was obliged to content himself with very careful sketches of these phenomena as they appeared to him. These are shown in Figures 12and 13. Since these figures represent the sections nearly parallel to the base, the straight lines in the sections must be considered to show the development of a system or two of parallel planes parallel to the optic axis. Uniform extinction was generally observed throughout each individual crystal, but in some crystals portions divided by the straight lines showed slight difference in extinction. Bearing upon the question of straight lines in the section, Tarr and Rich? describe one case in which phenomenon of the sort were observed. They regarded it as notable, however, that this was the only one of their experiments in which bending took place by shearing. Their experiment was about the bending of a single crystal of glacier ice whose optic axis was parallel to the supporting edge. In the present investigation, the same phenomena was observed whenever the constituent crystals were perpendicular to the bending plane. It has already been stated that deformation either by torsional, or by bending, stress took place least easily when the force was applied parallel to the basal section of the constituent crystals. This was thought to suggest that in the deformation of an aggregate of parallel crystals, the contact sur- faces between adjacent crystals probably have played an important «R. S. Tarr and J. L. Rich, Zeits. f. Glets., Vol. VI (1911-12), p. 243. SOME PHVSICAL PROPERTIES OF ICE 629 Fic. 12.—Microscopic appearances of ice crystals after bending. a and b: unbent portions. c,d, and e: bent portions. a b c Fic. 13.—Microscopic appearances of ice crystals before and after bending. a, b, and c: before bending. d, e, and f: after bending. 630 MOTONORI MATSUYAMA part if the idea of McConnell and Miigge is correct. The develop- ment of the planes parallel to the optic axis suggests that an ice crystal has a greater tendency toward deformation parallel to the optic axis than perpendicular to it. To what degree this property of an ice crystal and the contact surfaces are concerned in the deformation of ice is to be decided by further study. CONCLUSIONS In some specimens of ice a sharply defined elastic limit was noted, though in other cases it was not so clearly shown. The modulus of rigidity of ice, when the crystals are perpen- dicular to the axis of the test piece, is very small compared to that of metals, and is about 2X10° c.g.s. There is a slight indication that it is greater when the shearing is parallel to the base of the constituent crystals than when it is perpendicular. The Young’s modulus is also very small compared to that of metals. It is largest when the crystals are parallel to the length of the test piece, and has the numerical value about 20 X 10° C.g.s. Elastic fatigue was marked after repeated torsion. On account of the fact that it was often necessary to use certain bars in suc- cessive experiments during which they suffered from different amounts of fatigue, it was difficult to compare the results bearing on ice bars with crystals parallel and perpendicular to the length of the test piece. Still there were some indications that beyond the limit of elasticity the former orientation was stronger than the latter against torsion. In the case of bending experiments, this was clearly shown. The torsional deformations both by constant and by increasing forces were observed and the result is shown by curves, though no mathematical conclusions were made. The recovery curves showed that the observation was approaching the stage where no recovery would take place after removal of the force. When an ice bar with crystals parallel to its length was bent, the bent portion showed the change of optical character, the extinc- tion swinging around the curve. In each crystal when the bent specimen consisted of parallel crystals horizontally across it, parallel straight lines were observed. SOME PHYSICAL PROPERTIES OF ICE 631 These facts seem to show that gliding planes parallel to the base of each crystal are not the controlling factor in the deformation of ice and probably are not even an important factor. But instead, adjustments along the contact surfaces of adjacent crystals and perhaps the development of planes of weakness in the constituent crystals parallel to their long axis seem more effective in the process of deformation. A TEST OF THE FELDSPAR METHOD FOR THE DETERMINATION OF THE ORIGIN OF. METAMORPHIC ROCKS CHARLES GORDON CARLSON The University of Wisconsin 1. Purpose of paper.—That feldspars may serve as indicators of the original character of gneisses and schists is dependent upon the narrow range of composition possessed by the plagioclase feld- spars of igneous rocks. Thus more than one kind of plagioclase feldspar is rarely found in an igneous rock except in certain zonal intergrowths or in some porphyries where the feldspars forming the phenocrysts may be of slightly different composition from those of the groundmass. In sediments, except in rare cases where they are derived from rocks having feldspars with a narrow range of composition, the limited feldspar composition found in igneous rocks is not to be expected. Usually sediments are derived from many sources and consequently mixtures of all kinds of feldspars are possible. It would seem reasonable then to believe that gneisses and schists with a narrow range of feldspar composition are probably igneous in origin, whereas metamorphic rocks with several varieties of feldspar are very likely of sedimentary origin. This belief, how- ever, rests on the fundamental assumption that the feldspar range typical of sediments does not radically change during anamorphism of these sediments. It is readily seen that if such a change does take place it vitiates any conclusions which might be reached. Similarly, if in the anamorphism of igneous rocks a radical change in the original feldspar composition results, this also would militate against the efficacy of the feldspar method. That feldspars undergo alteration in various stages of the meta- morphic cycle is generally recognized. It is not known that this 632 THE FELDSPAR METHOD 633 alteration tends to produce feldspars of varied composition from one particular feldspar, nor conversely to change a wide feldspar range into a narrow one. The purpose of the work of this thesis was to determine the efficacy and validity of the hypothesis as above stated, namely, that metamorphic rocks having a narrow range of feldspar composi- tion are probably igneous in origin, whereas those having a wide range of feldspar composition are more likely of sedimentary origin. To test the validity of this hypothesis it was first neces- sary to get some idea as to the abundance of feldspars in various sediments and also to determine the range in composition of these feldspars. This involved a study of sediments both in the uncon- solidated and consolidated form. It was then further necessary to study. metamorphic rocks of known sedimentary and igneous origin in order to note whether the feldspar composition was such as would have characterized the original sedimentary or igneous equivalent. The methods used in this study and the results obtained are presented in this paper. The writer wishes to acknowledge his indebtedness to Dr. Edward Steidtmann,; of the University of Wisconsin, for suggesting the fundamental idea upon which the feldspar method is based, and to Professors A. N. Winchell and C. K. Leith for suggestions and criticisms. 2. Methods used to determine feldspars.—In the determination of the feldspars two distinct methods were used depending upon the character of the material to be examined. Where thin sections were available and the rock was fairly coarse grained the Fouque method was found very serviceable. When thin sections were not available and the material was so fine grained as not to be adapted to the Fouque method, the material was studied in powdered form and the feldspars deter- mined by immersion in a series of liquids of known index. With the liquids either the Becke or inclined illumination method can be used. The determination of feldspars from rock powders in this manner is especially valuable in cases where the feldspars are partly altered, where the rock is fine grained, when the feldspar content is low, and for all unconsolidated sediments. 634 CHARLES GORDON CARLSON 3. The materials studied.—In getting material together for study the attempt was made to make this selection one which would most thoroughly test the feldspar method. The mineralogic composition of unconsolidated and consolidated sediments as well as of metamorphic rocks of known origin was therefore determined. In order that the sediments might represent the breaking down of as many rock formations as possible, they were chosen so as to include a wide geographic and stratigraphic, distribution. The aim was also to avoid limiting the material studied to any one particular realm of deposition. Beach sands as well as sands of glacial, eolion, and locustrine origin were therefore chosen. The consolidated sediments examined included arkoses, graywackes, tuffaceous sandstones, and shales. Since the purpose of studying the metamorphic rocks was to determine whether anamorphism causes any changes in the feldspar composition of the original rock, the gneiss and schists were selected which showed different kinds and degrees of change. 4. Tabulation of results —The table on page 636 shows the results of the feldspar determinations for the various kinds of material studied. 5. The relative abundance of feldspars in sedimenis.—The data available are not sufficient to warrant a dogmatic statement as to whether certain feldspars are more abundant in sediments than others. ‘The studies by the writer of a large number of sediments of different origin, as well as of wide geographic and stratigraphic distribution, suggest very strongly, however, that certain feld- spars are very common in sediments, whereas others are quite rare. Orthoclase, microcline, and the acid plagioclases are much more frequently met with in sediments than the basic feld- spars. Microcline seems to be more common than any of the others, so that a careful study of sands which appear to be entirely composed of quartz usually reveals a few grains of this feldspar. By referring to the accompanying diagram (Fig. 1) the relative abundance of the various plagioclase feldspars is strikingly brought out. This abundance of the feldspars mentioned indicates either that they are especially common in the rocks from which they were derived or that the basic plagioclases suffer much more rapid Albjfe Ogoclase Ongoclase Andesine Basie Bylonile Vatevial Albite Oligoclase Andesine x Labpadorile Labr Beach Lae es De ar a ae ric eacenae | ae / GClecia/ heen Lake Sana en ee eel zB alpaca eRe me ales Ac! a rents | R and a Ne| Send (ee as Te eed eR = ae . on Ny eas Meer Co) ESS aes ena ey S |e | Lh 8 | sen as Paes ET S$ | Dove An ee eS ea ey Sand Ey area Ea Fe ee | SEDIITENT S 9 > > Y g SARAE a “ SS Ss] s N € NS a iN R Q > ° Q s 5 >» nS N : Ee a : a S aoe ag ieee a ee [Se ee a a eT S Sa a Sandsione ema ae reo) TE pe PERS jj jt | Esereaatoees aera sf a a ee a> Lh S8 sao A ~ NN 8% & NN Q 8@ H NS Roan 88 | 720 as | Ree) So] Cress 4 Ree.) oe Se aT Oe See| ladon ee Ble be Resa [a] ‘ ie ae Fic. 1.—Diagram showing the range in feldspar composition for the various materials studied. The numbers refer to those on the accompanying tables where additional data are given. The solid black lines opposite the material indicated in the left-hand column show the range in feldspar composition for that material. Note the wide range of feldspars in the material of sedimentary origin as compared to that which is igneous in origin. CHARLES GORDON CARLSON 636 wu Liv" Uv} JaSIvOD [eIIO}eWI Jo yUIO sod €g ‘soprsed jerourm pue yor quao sod z ‘aedspjey pur ‘z}1enb queso sod + ‘ayeuoqied jue sod v6 ‘1apmod=uorzisoduios = jerautyl azIs “WU YZI° JO puvs Jo yng “s[enptser AAvay yUad Jad S ‘tedspyay pue ‘zj1enb yuas sad ¢r ‘ayeuoqie 9 quad Jed 2$—uorjisoduos jerouryy sjenp -Isaq AAvay pue redspyoy ‘z}1enb quad Jad / ‘sajoied oyeys yueo sod g ‘a}euoqried quad rad S$g=pues Sty} Jo uorjtsoduros jeorsoyesouryy aula -Olotm ‘ouIsopue ‘azIq Ty aSe[D0U} -10 pure ‘auiposorur ‘aut -sapue ‘aserpsosijo ‘oqIq;y qoutes ‘uooriz ‘oyyede ‘ausyys -ledAy ‘oyisne yueo rod S$ ‘1eds -ppey ques rod +r ‘z,1enb yueo 10d sv ‘ayeuoqies yuad rod gf =purs SIy} Jo uorjtsodutod [eorsoye1oUrpy pues jo yuao sad $ ynoqe sur10y redspye,q SYILUIOY SUIPOIDIUL “ase;DOYI10 ‘OyIOpeiqe] dIseq ‘our -sapue ‘aseposijo ‘ozIq;Ty purysy ys0on -UY UOT}BUIIO} 93309 YQ Ul YIoMm Woy purysy soonuy ‘sqUIOg x1019v'] pur 98005) UdaMyoq Are -njso JO YNoW Woy 9u0}s purysy Ysoonuy dUI[IOIO -IW pue ‘osepooyy10 ‘our -sopue ‘asepoosi[o ‘aqIq,y dUTPOIOIU d}Opesqey d1seq ‘a}110p -eiqe, ‘oulsapue ‘oseyo -OSI[0 ‘ayIq(e asepOosI[Q asepoyy -10 pue ‘ouT[IOJOTUT ‘OUT -sopue ‘asepo0sI[o o}Iq,TV punoy siedspjo,q urs -U0dSIM ‘JUIOg YINOS UISUODST/\. “10}2]PPUNL purjsy suva “TINS “eurporeD YInog AY[eIOT NIOTYO NMONX AO SWOIOU OIHCYONVLAWN GNV SLNAWIGHS 1O NOILISOdNWOO UVdSATad OL SV GANIVLEO VLVG ONIMOHS ATIVL -OUl] WO} paar -op Ajesie’y “pues j°°** 9 [ero -BUl [VLe[s poyIoM -o1 suIvJUOD “pues | **° s pues Ss [ebe[s oulos sul 5) -uIe}U0D ynq 9U0}S gz -QUII] WOIf PdATIop = AjeSiey ‘pues yprag |-*° °°: | ey a dp) a B uelIiquieg g) pues ouneyy | £ ot pues dye] [Oa io) pues yorog ZEN) [eloye “ON ~~ ise) Ne) THE FELDSPAR METHOD ‘anbnog pue dyx0gq 1OJ puvys , pas poy, papeey uwnNjoo 94} JapuN vf pur _ S19}}9] OTT, x peztuyory Ayqeiepisuod st redspya.q IBROISSPLIT -OISSBAN jo ‘sollag yIRMON Jo aed st yoy ose sopijied yooy ~“Aperyo reds -ppey pue zz1enb jo pasoduros yoy dUISOPUL ISB[IOSI[O uISUOD [elj}seqi0} Jo Suteq se pazordioqur ‘SOLOS YIRPMIN' 94} WOIf st Yor -o1q ‘juepunqe siedspyey oseporse[g jyuepuNnqe sUTpO.IOIPL Aavay gt “redspjoy queo sod 11 ‘gyeuoqies yuoo sad blz ‘z,1enb A \oudye ‘esepostyo ‘ayqry |-sIM “uIog UuoZUTTD asoyly joo SI auryo -O1NTW «=pue ssPR[D0Y}IO elueA[ASUUd (Quojspurs UMOIG) A |foyqye oseposto ‘azIqTy UMO}S|OUIUUN FT OSOSLV ia |bamunenn FI 9UT[IOII yuesoid -IUI pue ssR[I0y}10 ‘our -Sopue osepoosijo ‘asvio urluoinyzy A |-Os1o ‘ozIq7e aseposI[O UISUODST ‘ASTIN FT ayoemAein |°*°*"° €1 UISIIO gulpororm ‘9}1L10p DISSPLI J, -eiqe] dIseq ‘ouIsopue orssean [° q \‘ouqe ‘eseposijo ‘oyI1qTy Aasiof MON OSOMIY [ooo ZI ISBJIOYIAO ‘9SBPIOSI][O 9UI UvIIquIV)-d1g J |-sapue ‘oseposijo ‘ayIq7y | UrTSUOoSTA ‘nesne ayoemAein, [°° * II asepoyqO ‘auIsapue uRlLIquUie)-d1g A jeqqwe oserposto ‘ayIq~y | UlsuOoSTAA ‘nesne IICMABID) [°° + °° * Or QUI[IOINIW PUR ISPID uPRlIquie) -O4}10 ‘ouIsapue asejId q |-o31[0 ‘asepoostjo ‘ayIq;y | UlsUODST\ ‘nesne auo}spuvs dISOyTy [°° 6 ase[D0yj10 pure ourposstw ‘ouIsep BIULOFI[eD) q |-ue ‘aseposyo ‘oyqry |‘yIeVq 23e5 Usploy pues sung |°°°°*° 8 “s[eNpIsor duT[IOIN -IW pue asepOYIIO ‘941 d}IMIOJOP &}OIUG -IOpeIge] diseq ‘oyIIOpel | YIM polfize14s190} UT uPlIiquieg q |-qe] ‘outsopue ‘asepos1[Q |UIsUODSI ‘U0J[PPIIT pues oumeyy | L quad god og :uottsodwos jeroulyy S]}UIWIIPIS po}eprposuod 638 CHARLES GORDON CARLSON ueliquies uerluoInyy -pjayj JO swIos {yUepUNqe dUTPOIOITT urluoIny, + ‘“JUepUNqe ssePoyIC SISv[IO -1sejd Jo JUepuNgqe jseva| oyLIOpeiqe’T juepunge jou sseporseylg juepunge jou aseporserg a} LIOULO[SUOD pouleis-suy SI yoy ‘jURpuNqe st oseposelg SyIVUIIy ayIopesqgey S1seq ‘our dUIPPOIOTU ‘aq110p ‘guisopue ‘ozIq,y IUTIOIOIUT PUL ISPII -oyyo ‘ayopeiqey ‘out q |-eaqe| aula -O10TU ~=pue 9SR]D0N}I0 ‘QUOpeiqe, oIseq ‘out q |-sepue ‘aseposijo ‘ozIq;y ISe[IOYIAO pue oulportu ‘o}110p q | -eaqe, ‘outsopue ‘ozIGTy q |-sepue ‘oseposijo ‘ozIqTy q |-Sepue ‘asepostjo ‘olIqTy arrysdure py aura -O1NIW =pue 9SP[IOY}10 4 |‘outsepue ‘ozIqye ase[DOSI[O ISP[IOYJIO pure oUl[D J j-or1omt ‘outsapue ‘oqIqry ase[0yy -10 ‘o}WOpeiqey] dIseq ‘guisopue ‘ase[d0sI[0 A | ‘ovqre eseposio ‘ouqry puno,y sredspja,q MAN “1a}UID UOIeID ayoemARin [+ *° €z WeSIYIIAT “Woy -ysnoyy Jo ysveyyNos guoyspurs o1soyry |°° °° zz RJOSOUULP[ “93v}10g asoyre poysepy [°° iz O1eJUQC “9114SIq WeqoD “Woy sjiAeqd asoyy |: oz uesIyoI, Uoyjs0U ase URMBUOOMIY ‘Ayuno0d) uesou0juQ aUO}SpUeS DISOYIY [°° 61 uIS -uoosIM ‘Avg 07U090 ayerouiojsuog | -*: gt ueIuOINn;y o1reyUQ “WYeqod ayetowojsuo, |°* °°: Lt euRyuoy, = ‘“T[BVYOIy asomry | - gt Ayyeo07T [e978 TA ‘ON SJUSUIIPIS poyeprjosuod panuryjuoo—NTOTXO NMONM AO SMO0O0e OIHCYOWVLAW GNV SLNAWIGAS AO NOILISOdWOOD UVdSATAA OL SV CANIVLEO VLVG DNIMOHS ATAVL THE FELDSPAR METHOD ‘onbnog pur ayxdegq IO} purjs ,.pas poy, Papvay UUNjoOD 94} JapuN YF pure _ S19}}0, DY Ty juepuNq? sUT[IOIOIIT quepunqe duTpIOIOIPT ase uRIUOINy ouIsapur [TV | purpArey “19}S9y9]] sslous apus,quiopyy | °° 9¢ yuep -unqe aSvJI0Y}AIO pure ouUr]d umispgq ‘pouuep ArAydiod -O1NP + “azIq(e aaoqe [[V |-A~y ‘osny, ap sued ssiour) |° °° ce oulsapur [LV BIsIOIL) UIaY}ION ssiou3 uroy [°° °° re ynot} ATUO o}IGTy |-oouUOD ‘UOJSeUMOY,[, aytueIs plosstouny |°* °°" ee s}josnyo ATWO azIqTy |-esseyy ‘SIOATY 9oI J, o}IURIS plossiour) | °°" °° ze oul jNoTWeUUOD, 9} VIIWO[SU0D -sopue ‘ase[s0sI[0 ‘azIG Vy ule} YINosS asoyre poyeg foo: 1 our -sapue ‘aseposijo ‘aytqry |ANO YOR MaN Ivan SSIOUS O][[AUOID) | of aSepIOYIAO ‘QUIS BURIUOW, -sapue ‘aseposijo ‘ayIqTy |‘JolysIq UIOYysuURy ystyos zyreng) | - Ge aSEPOYIIO puv duUI[D -O10TIU «‘oUISepuR aSPID s}jesnyo -o81]0 ‘ayIq7e aseposyO | -esseyy “WeysulAy, SAOuH) pr2e 7°" QZ auryo -O13TU ‘asepOyyIO “ouT s}josnyo -sapue o}Iq(e ‘aseposQ | -essepy “AOA [IAL | Sslous oyerato0jsuoy |° °° * Le asepoy} -10 ‘auTpor1sTu ‘az 10p oy eqeaer@) -eiqey, ‘eulsopue ‘ozIGTy “SUIUIEYSTUIO [, o}Izjaenb s1soyry | gz suisopue ‘oyiqry |purfArepy ‘orournyeg ssious o1ownyeg f°” cz 9} LIOp Ce eo eeeay, Ve -BIQE] ‘QUISApUR ‘ISePOSITO BIULOF[VD ‘O[SaT, auo}spuRS snosdByN], SyOOYW Snosusy dydiowejzpy syooy ArejusuIpas s1ydiowejzaPL 640 CHARLES GORDON CARLSON decomposition. ‘The latter seems the more reasonable conclusion since many of the sediments studied have had their origin in areas of basic igneous rocks. At Keweenaw Point for example the Keweenawan sediments show a very small amount of the basic feldspars as compared to the acid varieties and yet the sediments have been largely derived from rocks of a decidedly basic character and from rocks in which basic feldspars are known to be very common. It is also generally recognized that the calcic feldspars are more readily decomposed than the more alkaline varieties. Iddings states that The alkalcic feldspars are not attacked by hydrochloric acid. The more calcic feldspars are decomposed by the acid in proportion to their content of calcium. Thus oligoclase and andesine are not attacked, labradorite is slightly acted upon, bytownite and anorthite are decomposed with the separation of gelatinous silica. In the rocks the more calcic feldspars are more readily decomposed than the more alkalcic feldspars in general. Feldspars are much more common in sediments than has generally been supposed. A large number of “sandstones” and ‘“‘quartz”’ sands were in many cases found to have a considerable percentage of feldspar. Sands with a 5 per cent content of feldspar are not at all uncommon, while certain glacial and marine beach sands may contain feldspar up to 25 per cent. 6. Feldspar range of rocks studied.—It was desired to determine, by the work pursued in connection with this thesis, just what range in feldspar composition can be expected in sediments, and further to ascertain whether, during anamorphism, there is any change in the feldspar composition of the original igneous or sedi- mentary equivalent. ‘The results obtained show that almost any combination of the various feldspars can be found in sedimentary rocks. Of the twenty-four samples studied, these samples includ- ing unconsolidated and consolidated sediments, twenty-three showed a range in feldspar composition from albite to andesine. Labradorite was found in eleven of the samples, while anorthite, due undoubtedly largely to its ready solubility as well as compara- tive rarity was not noted in any of the samples studied. As was expected glacial and marine beach sands show a very large range tJ. P. Iddings, Rock Minerals, p. 204. THE FELDSPAR METHOD 641 in feldspar composition. Studies of metamorphic rocks of known igneous and sedimentary origin showed that the former retained their limited feldspar composition, whereas. the metamorphic- sedimentary rocks included feldspar combinations such as would characterize the original sedimentary rock. The conclusion, as based upon the work done, is that there is no decided change, during anamorphism, of the feldspar composition possessed by the original unmetamorphosed material. 7. The usefulness of the feldspar method as compared with the present criteria used in the determination of the origin of metamorphic vocks.—The present criteria which are used to determine the igneous or sedimentary origin of metamorphic rocks are dependent upon field relations, together with chemical and mineralévical com- position. Field evidence consists chiefly in tracing metamorphic rocks into the less altered igneous or sedimentary equivalents. Thus a basalt has often been observed to grade into a chlorite or micaceous schist. Similarly banded gneisses are often associated with and grade into granites. Chemical evidence suggestive of a sedimentary origin consists, according to Bastin" “of a dominance of magnesia over lime, potash over soda, excess of alumina and high silica. If the chemical composition is essentially that of an igneous rock this fact favors igneous origin.” Mineralogical evidence favoring a sedimentary origin consists of a high content of quartz as does also an abundant development of aluminum silicate minerals. The presence of graphite probably denotes a sedimentary rather than an igneous origin.2 Rounded grains of such minerals as garnet, sphene, and especially zircon have been taken as evidence of sedimentary origin. These minerals are especially resistant to weathering and will remain after the other minerals have been completely altered. The plagioclase feldspar method is an addition to our mineral- ogical criteria. The results obtained prove the feldspar method to be a valid and reliable method for the determination of the * Edson S. Bastin, ‘Chemical Composition as a Criterion in Identifying Meta- morphosed Sediments,” Jour. Geol., XVII (1909), p. 472. 2 J. D. Trueman, ‘The Value of Certain Criteria for the Determination of the Origin of the Foliated Crystalline Rocks,” Jour. Geol., XX (1912), pp. 228-58, 300-15. 642 CHARLES GORDON CARLSON metamorphic rocks to which it is applicable, and this means any rock containing recognizable feldspar constituents. The studies show that metamorphic rocks in general, except where they have suffered alteration due to ordinary weathering or hydrothermal alteration, contain such constituents. Where hydrothermal altera- tion has been effective, as in the proximity of the intrusive por- phyries of the west, some other criteria must generally be resorted to. Even here, however, the alteration is not likely to have pro- ceeded far from the main intrusive, so that by following a forma- tion into its unweathered portion, recognizable feldspars may often be found. The feldspar method is to be preferred to the heavy residual or ‘‘zircon”’ criterion. The theory upon which the heavy residual method is based is undoubtedly a valid one, yet the studies of a large number of sediments show that any inter- pretations as to the origin of metamorphic rocks which are based upon its use, cannot be but uncertain. In the examination of marine beach sands from South Carolina and Anticosti Island, crystals of zircon and titanite were found which retained perfectly their crystal outline. Dr. W. H. Twenhofel reports similar results from a study of coral beach sands from the Hawaiian Islands. Such a sediment after conversion to a metamorphic rock would, on the basis of the zircon method, have been interpreted as suggestive of igneous origin. It must be borne in mind, however, that of all the criteria at present available for the determination of the origin of schists and gneisses, the use of field relations, where possible, is by far the most conclusive. Chemical and mineralogical criteria must therefore be subordinated to it. On the basis of practical usefulness and reliability the feldspar criterion should supply a . valuable addition to our present laboratory methods. SUMMARIES OF PRE-CAMBRIAN LITERATURE OF NORTH AMERICA EDWARD STEIDTMANN University of Wisconsin II. ONTARIO In the region northeast of Lake Huron, the pre-Cambrian rocks according to Collins and others show one conspicuous unconformity. The rocks beneath this unconformity comprise a series of quartzites and other clastic sediments, the Timiskaming series, etc., intruded by granitic rocks. Unconformably beneath these sediments is an older series, the Keewatin, including basic flows, some acid extru- Sives, iron formations, dolomites, etc. The Keewatin is intruded by the Laurentian granites and gneisses. Above the conspicuous unconformity are two series of slightly metamorphosed dominantly clastic sediments separated by an inconspicuous unconformity. The lower one, the Bruce series, locally contains tillites. The upper series is generally known as the Cobalt series. At Killarney on the north shore of Lake Huron, Collins has found that the Bruce and possibly the Cobalt series are intruded by the Killarney granite and in this locality they assume many of the characteristics of the older series, the Timis- kaming. The youngest pre-Cambrian rocks are Keweenawan, basic dikes and sills. Northwest of Lake Superior, Lawson has restudied the Rainy Lake and Steeprock Lake districts. Greenstones and other rocks typical of the Keewatin are widely exposed in this region. Beneath them are acid schists called Coutchiching by Lawson. Uncon- formably above the Keewatin in the vicinity of Rainy Lake are a series of conglomerates and slates called the Seine series by Lawson. In the Steeprock Lake district, the Steeprock Lake series lies unconformably between the Keewatin and Seine series. The Steeprock Lake series, besides clastic sediments, comprises fossil- bearing dolomites. 643 644 EDWARD STEIDTMANN The youngest rocks of the region are basic dikes classed as Keweenawan. Baker’ classifies the pre-Cambrian rocks of the Kingston area in southeastern Ontario as follows: Great unconformity Keweenawan—Trap, diabase, and gabbro intrusives Intrusive contact Algoman—Coarse-grained granite and syenite intrusives with later pegmatites Intrusive contact Laurentian—Gray to pink, medium to fine-grained, granitic gneisses Intrusive contact Grenville—White, coarsely crystalline limestone with quartzite and rusty weathering gneisses. Dark green to black gneisses—thoroughly impregnated with minute dikes of Laurentian granite, now also changed to gneiss. As reported by E. L. Bruce,? the succession in the Cripple Creek Gold district located about twenty-five miles southwest of Porcupine, Ontario, is: Glacial and Recent Peat, unsorted and more or less sorted sands and clays Unconformity Post-Laurentian Diabase dikes Igneous contact Laurentian Gray granite—reddish gneissoid granite Igneous contact Keewatin Greenstones, schists, diabase, and iron formation The Kirkland Lake and Swastika} gold areas are located in the Timiskaming district, fifty miles north of Cobalt. The pre- Cambrian rocks are classified as follows: Later dikes—Diabase Intrusive contact 1M. B. Baker, ‘‘The Geology of Kingston (Ontario) and Vicinity,” Ontario Bur. Mines, 25th Ann. Rept., Vol. XXV, Part 3 (1916), pp. 1-36, 19 figs., map. 2E. L. Bruce, geologist, and W. R. Rogers, topographer, “‘Cripple Creek Gold Area, Ontario Bur. Mines, Vol. XXI (1912), Part I, pp. 256-65, 9 figs. 3 A. G. Burrows and P. E. Hopkins, ‘‘The Kirkland Lake and Swastika Gold Areas, Ontario Bur. Mines, 23d Ann. Rept., Vol. XXIII, Part II (1914), pp. 1-30. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 645 Cobalt series—Nearly flat-lying conglomerate with bowlders of granite and syenite Unconformity Post Timiskaming intrusives—Granite, syenite, feldspar, porphyry, lampro- phyre Intrusive contact Timiskaming series—Quartzite, graywacke, conglomerate with schistose derivatives. The conglomerates contain a variety of pebbles derived from the Keewatin Keewatin—Greenstone (basalt andesite) diabase, quartz porphyry, feldspar porphyry, iron formation, dolomite Burrows’ maps the Matachewan Gold area on Montreal River in latitude 48. Below is the table of formations: Animikean-Cobalt series—Conglomerate, quartzite, graywacke, slate Unconformity Algoman—Granite, syenite, and thin acid intrusives Intrusive contact Laurentian—Granite and gneiss Keewatin—Greenstones, iron formation, some quartzite, conglomerate, etc. Burrows’ classifies the pre-Cambrian rocks of the Porcupine Gold area of Ontario as follows: Keweenawan—Quartz diabase, olivine diabase Intrusive contact Algoman—Granite porphyry, feldspar porphyry Intrusive contact Pre-Algoman—Lamprophyre, serpentine quartz porphyry Intrusive contact Timiskaming series—A series of schistose conglomerates, interbanded slate and graywacke, quartzite “‘carbonate”’ rock Unconformity Keewatin—A couple of largely schistose basic to acid volcanics, agglomerates, ash rocks, iron formation, rusty weathering, “carbonate,”’’ diabase, serpentine, etc. The gold occurs in quartz veins cutting the Keewatin and Timiskaming series and the pre-Keweenawan intrusives. They are believed to be related genetically to the Algoman intrusives. * A. G. Burrows, ‘‘The Matachewan Gold Area,” Ontario Bur. Mines, Ann. Rept., Vol. XXVII (1918), Part I, pp. 215-40, maps and illustrations. ? A. G. Burrows and P. E. Hopkins, ‘‘The Porcupine Gold Area” (Third Report), Ontario Bur. Mines, Ann. Rept., Vol. XXIV, Part III (1915), pp. 1-57, 44 figs. inclusive, maps; see also ibid. (Second Report), Ontario Bur. Mines 21st Ann. Rept., Vol. XXI (1912), pp. 205-49, 37 figs. 646 EDWARD STEIDTMANN The Whiskey Lake area includes two unsubdivided townships, Nos. 137 and 138, in the third and fourth tier of townships north of Lake Huron. The pre-Cambrian rocks of the area are provi- sionally classified as: . Middle Huronian—Conglomerate and quartzite Unconformity Lower Huronian—Conglomerate, quartzite, slate, and limestone Great unconformity Sudbury series—Slate and probably quartzite, part of the greenstone Unconformity Keewatin—Most of the greenstone and green schist Laurentian—Granite and syenite Coleman?’ classifies the pre-Cambrian succession along the north shore of Lake Huron as follows: Keweenawan—Basic volcanic eruptives and basic sills. Subordinate coarse, usually red sediments, probably indicating warm, dry climate Animikie—Black slates, volcanic tuff, bowlder conglomerate Huronian—Arkose and quartzite, shallow lake or sea deposits indicating cool climate. Bowlder conglomerate or tillite formed under glacial conditions Sudburian—Pillow basic lava flows. Coarse sediments—conglomerates, bowlder beds, arkoses, quartzites derived mainly from the dis- integration of granites Grenville—Quartzites, schist, and impure calcareous sediments whose relation to the Keewatin is uncertain. Intrusion of granites Keewatin—Basic eruptions and jaspititic iron formations Coleman classifies the pre-Cambrian rocks of the region north of Lake Huron extending from Point Mamainse to Wanapitie as follows: Keweenawan Discordance Post-Laurentian { Animikie Discordance Upper Huronian Great discordance Sudbury series Pre-Laurentian Great discordance Keewatin—Probably equal to the Grenville series « A. P. Coleman, ‘‘The Whiskey Lake Area,” Ontario Bur. Mines, 22d Ann. Reft., Vol. XXII (1913), Part I, pp. 146-54, 5 figs. 2 A. P. Coleman, ‘‘The Pre-Cambrian Rocks North of Lake Huron with Special | Reference to the Sudbury Series,” Ontario Bur. Mines, Ann. Rept., Vol. XXIII (1914), Part I, pp. 204-36, map, 18 figs. 3A. P. Coleman, ‘‘The Sudbury Series and Its Bearing on Pre-Cambrian Classi- fication,’’ Congrés Géologique International XII. Session 1914. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 647 The major divisions are based on the position of the various series with reference to a conspicuous unconformity and to certain granite batholiths. He recommends that this section be adopted as a standard for the Lake Superior region. In 1915 Coleman’ classified the rocks of the Canadian Shield to the northeast of Lake Huron as foliows: : Keweenawan (Mamainse and nickel eruptive) Discordance Late Proterozoic ; Upper Huronian Small discordance > Typical Huronian Lower Huronian Great discordance (Laurentian granite and gneiss) Early: Proterozoic Eruptive contact Sudburian—Timiskaming, Pontiac, etc. Great discordance ( (Granite eruptive through lower series) Archaeozoic Eruptive contact Keewatin and Grenville The important points of this classification are the recognition of two major unconformities in the succession and the naming of certain granites and gneisses which are intrusive into rocks younger than the Keewatin, as Laurentian. In his discussion following the classification, he substitutes the term Animikie for Upper Huronian. The term Sudburian for a series is recent in the general discussions of the stratigraphy of the Canadian Shield. This series, Coleman states, is typically developed in the Sudbury district, where it consists chiefly of quartzites, slates, and con- glomerates, without limestones or dolomites, and with almost no carbon. ‘These rocks are severely folded, but not intensely altered excepting near intrusives. Their deposition was followed by the extrusion of lava Sudburite, the effusive equivalent of Norite. Other probably Sudburian areas include portions of the region to the northeast of the Wahnapitae River, and the Timiskaming, Gowganda, Larder Lake districts, the area of the Pontiac series of Quebec, the Doré formation of the east shore of Lake Superior, and certain rocks of Heron Bay on the north shore of Lake Superior, tA. P. Coleman, “‘The Proterozoic of the Canadian Shield and Its Problems,” Problems of American Geology (1915), pp. 81-161. 648 EDWARD STEIDTMANN the Nipigon area, the Onaman Iron Range, and the Seine River in the Rainy Lake district. The Sudbury series he regards as being partly a delta deposit laid down in a moist cool climate, but finds it strange that carbon is lacking. In the interval between Sud- burian and Huronian time, the area was folded and eroded to a surface very much like that of the present Canadian Shield. For the nature of the Lower Huronian, he refers to Logan’s type section on the northeast coast of Lake Huron. ‘Tillites are a characteristic constituent of the Huronian, but in addition it con- tains stratified deposits. Other Lower Huronian areas are found in the Larder Lake, Chibougama, and Steep Rock Lake districts. The Lower Huronian rocks have a marked unconformity at their base, but are in general less severely folded than the Sudbury series. At the start the climate of the period appears to have been cool and glacial. The existence of animals is suggested by the occurrence of limestone. The Animikie he characterizes as a period of great submergence during which great quantities of iron compounds and black slates were deposited. The Keweenawan of the Canadian Shield rests upon the eroded Animikie. It includes three series, of which the two lower are chiefly sedimentary, while the upper is largely volcanic. The sediments consist largely of sandstones and conglomerates, charac- terized by red color and absence of carbon. The volcanics are chiefly basic flows, but possibly include some felsites and por-_ phyries. Dikes are common, but as yet no definite volcanic vents have been found. Other areas of the Canadian Shield probably containing Keweenawan are the Nastapoka and Manitaunick Islands, Central Labrador, and the south side of Hudson Straits, the regions of Lake Athabasca, Great Slave Lake, and the area between the east side of Great Bear Lake along the Copper Mine River northward to the Arctic Ocean. Deposition during Kewee- nawan time, according to Coleman, was chiefly on the land in a warm dry climate. He speculates as to the source of great quan- tities of lavas and relates the development of the Lake Nipigon, Sudbury, and Lake Superior basins to the collapse of the surface resulting from the extrusion of the lavas. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 649 The base of the succession exposed in the Gowganda‘ dis- trict consists of Keewatin greenstones mostly of igneous origin associated with some iron formation. They are intruded by batho- liths of Laurentian granite. Overlying the granites and green- stones with well-marked unconformity are Huronian sediments which are separated into two members by a faint unconformity. The lower group from 500 to 1,000 feet thick consists of con- glomerates, arkose, graywacke, and slates, showing poor assort- ment, variable bedding, and till-like character in the coarse phases. Locally, the beds are associated with rhyolitic extrusions. The upper is a single quartzite formation 600 or more feet thick, ran- ging from arkose to pure, well-bedded quartzite. Intruded into the preceding are Keweenawan diabase sills and dikes. Selected areas between the original Huronian and the Cobalt and Sudbury districts were examined by Collins’ with the view of correlation. The Bruce, Blind River, Whiskey Lake, Espanola, and Round Lake areas were selected, the widest gap between them being about 28 miles. Collins recognizes two major stratigraphic divisions, the pre- Huronian and the Huronian. They are separated by the most conspicuous unconformity of the region, characterized by a strong basal conglomerate, great differences in structure, metamorphism, igneous intrusions, and general lithologic character of the two groups. The pre-Huronian consists of basic schists and gneisses mostly of igneous origin, granite batholiths of more than one period of intrusion and highly metamorphosed slates and quartzites. The pre-Huronian has not been completely subdivided into strati- graphic units and its various members have not been traced and correlated over the entire region. The Huronian is separated into two units by an unconformity far less pronounced than the one at the base of the Huronian. Individual beds of both divisions have been traced successfully from district to district. The lower division, called the Bruce =W. H. Collins, ‘The Geology of Gowganda Mining Division,” Canada Geol. Surv. Mem. No. 33 (1913), 121 pp., 4 pls., 5 figs. 2W. H. Collins, “The Huronian Formations of Timiskaming Region, Canada,” Canada Geol. Surv. Mus. Bull. No. 8 (1914), 27 pp., 2 maps, 1 fig., 1 pl. 650 EDWARD STEIDTMANN series, consists of a thin basal conglomerate, white quartzites with interbedded, well-sorted conglomerates, an impure siliceous limestone, and some graywacke, whose maximum thickness is more than three thousand feet. The upper division, or Cobalt series, includes tillites, quartzites, graywackes, a few thin im- pure limestone beds and grades upward into pure quartzites. It has in part the characteristics of glacial till associated with stream and quiet-water deposits contemporaneous with glaciation. Locally it shows minor unconformities. The local terms, Bruce and Cobalt series, rather than Lower Huronian and Upper Huron- ian, are applied to these divisions because their full equivalence to these units in the original Huronian is regarded as doubtful. Collins' advocates that a local classification of the pre-Cambrian rocks of the Timiskaming region be adopted and that their correla- tion with other districts be postponed until they are better known. He emphasizes the importance of the unconformity at the base of the Cobalt series as major plane of division. The various series are classified by him as pre-Huronian and Huronian. His classi- fication follows: { Diabase | Sudbury norite Intrusive contact Whitewater series Huronian Lorrain series Local unconformit Cobalt series Great unconformity Batholithic granite intrusive Intrusive contact Sudbury, Timiskaming, Fabre series | Unconformity Granite intrusives Keewatin group Keweenawan Pre-Huronian Collins? reports that hitherto unknown granites intrude the Bruce and probably the Cobalt series along the coast of Lake Huron. t W. H. Collins, ‘‘A Classification of the Pre-Cambrian Formations in the Region East of Lake Superior,’ Congrés Géologique International XII. Session 1914, PP- 399~4097- 2W. H. Collins, “The Age of the Killarney Granite (Ontario),” Canada Geol. Surv. Mus. Bull. No. 22 (1916), 12 pp., 1 pl., 1 fig. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 651 Collins' reports on the Onaping map area about fifty miles north of Sudbury. Following his former practice, he divides the pre-Cambrian rocks into Huronian and pre-Huronian. His table of pre-Cambrian formations follows: Huronian—Keweenawan Olivine diabase Basic intrusives ? Quartz diabase Quartz norite and intermediate varieties Intrusive contact Cobalt series—Upper white quartzite Banded cherty quartzite Lorrain quartzite Gowganda formation: Conglomerate Graywacke Limestone Great unconformity Pre-Huronian Batholithic intrusives—Granite gneiss and its differentiates Schist complex—Altered volcanic and intrusive rocks, iron forma- tion, and other sediments P. E. Hopkins’ reports the succession in McArthur township of the Porcupine Gold Area as Late intrusives—Diabase Timiskaming ?—Slates Laurentian—Granites intrusive into Keewatin Keewatin—Greenstone, serpentine, hornblende schists, porphyries, carbon- ates, and chert magnetite iron formation Hopkins’ reports on the Beatty-Munro Gold area in the Larder Lake mining division of Ontario, latitude 48° 30’, longitude 80° 15’. The rocks are all pre-Cambrian and are classified by Hopkins as follows: Post-Timiskaming intrusives—Feldspar porphyry dikes Intrusive contact Diabase dikes and stocklike masses Intrusive contact 1 W. H. Collins, ‘‘Onaping Map Area (Ontario),’’ Canada Geol. Surv. Mem., No. 95 (1917), 157 pp., 11 pls., 8 figs., 2 maps. 2 P. E. Hopkins, “‘Notes on McArthur Township,” Ontario Bur. Mines, 21st Ann. Rept., Vol. XXI (1912), Part I, pp. 278-80, 2 figs. 3 P. E. Hopkins, ‘‘The Beatty-Munro Gold Area (Ontario), Ontario Bur. Mines, Ann. Rept., Vol. XXIV (1915), Part I, pp. 171-84, 9 figs., 1 map. 652 EDWARD STEIDTMANN Timiskaming series—slate, graywacke, quartzite conglomerate, and schistose derivatives Igneous—Feldspar porphyry—telation to Timiskaming uncertain— intrudes Keewatin Intrusive contact Keewatin—Amygdaloidal and ellipsoidal basalt, diabase, serpentine, iron formation, and breccia, with metamorphosed equivalents The gold occurs as free gold and as tellurides in quartz veins cutting Keewatin and Timiskaming rocks. ‘The veins contain high- temperature minerals, viz., pyrrhotite and tourmaline. Kindle and Burling’ conclude that the escarpment of pre- Cambrian rocks which overlooks the plain of Paleozoic sediments north of the Ottawa and St. Lawrence rivers is due to normal faulting, the sediments being on the downthrow side. The facts which indicate this are: (a) the presence of Paleozoic outliers rest- ing on a hummocky surface of pre-Cambrian rock north of the escarpment, the corresponding Paleozoic beds south of the escarp- ment being about seven hundred feet lower, (b) the extreme regularity of the escarpment, (c) the absence of Paleozoic re-entrants along the escarpment, (d) the lack of clastic material from the lime- stone adjacent to the pre-Cambrian rocks of the escarpment, (e) the dissimilarity of the escarpment features with other nearby pre- Cambrian borders where normal erosion has even yielded an escarp- ment of Paleozoic rocks, (/) the escarpment is at the northern border of a zone in which subsidence or normal faulting is characteristic. Knight? finds the following succession of pre-Cambrian rocks in the Thessalon area on the north shore of Lake Huron to the west of Killarney. ‘This is the original Huronian area of Logan. Diabase dikes intersecting Nipissing Keweenawan—Diabase Intrusive contact Nipissing diabase, similar to that at Cobalt and Gowganda— shows local gradations into pink micro pegmatite. Thessalon greenstone, a fine-grained basal sometimes amygdaloidal Intrusive contact tE. M. Kindle and L. D. Burling, ‘Structural Relations of the Pre-Cambrian and Paleozoic Rocks North of the Ottawa and St. Lawrence Valleys,” Canada Geol. Surv. Mus. Bull. No. 18 (1915), 23 pp., 2 pls., 6 figs. 2C. W. Knight, ‘‘The North Shore of Lake Huron,” Ontario Bur. Mines, Ann. Rept., Vol. XXIV (1015), Part I, pp. 216-41, 13 figs. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 653 Animikean—z. Pink quartzite and arkose with thin beds of jasper con- glomerate similar to Lorrain series at Cobalt 2. Slatelike graywacke—beautifully and thinly bedded 3. Conglomerate, graywacke, slatelike graywacke, quartzite, arkose Great unconformity Algoman—Granite, massive, and at times gneissoid Knight" and others report on the Abitibi-Night Hawk Gold Area southeast of Cochrane on the Canadian National Railway. The succession includes Keewatin rocks consisting of basic pillow lavas, rhyolites, basalt, diabase, hornblende, and chlorite schists, which are overlain by slate graywacke, quartzite, conglomerate, and iron formation. These Keewatin rocks are intruded by diabase and gabbro, peridotite and pyroxenite, granite and other acid rocks, quartz diabase and olivine diabase dikes. In 1911* A. C. Lawson restudied the Rainy Lake area which he had reported on in 1887. In 1887, Lawson reported that the pre-Cambrian rocks of this region were all Archean and that the succession from the bottom upward is as follows: A series of clastic sediments metamorphosed to mica quartz schists and paragneisses called the Coutchiching. This series is conformably overlain by the Keewatin, consisting dominantly of amygdaloidal and ellipsodial greenstone lava flows, chloritic schists, and other basic rocks of a similar nature. The Coutchi- ching and Keewatin were intruded by batholithic masses of granite and granite gneisses and allied acid igneous rocks which caused the doming up of the rock into which they were injected. The restudy of the Rainy Lake area by Lawson in 1911 was occasioned by the fact that the United States Geological Survey and the International Committee of 1898 did not accept Lawson’s - conclusion that there existed a Coutchiching series of rocks strati- graphically below the Keewatin. This dissent from the opinion of Lawson was based on field work by Van Hise in various parts tC. W. Knight, A. G. Burrows, P. E. Hopkins, and A. L. Parsons, ‘“ Abitibi- Night Hawk Gold Area,” Ontario Bur. Mines, 28th Ann. Rept. (1919), 84 pp., maps, and illustrations. 2‘*The Archean Geology of Rainy Lake,” restudied by A. C. Lawson. Canada Geol. Surv. Mem. No. 40 (1913), 111 pp., geological map in pocket, 9 pls., 1 fig. 654 EDWARD STEIDTMANN of the Rainy Lake area, and in consequence of the examination of the Coutchiching series on the east end of Shoal Lake and along parts of the Seine River by the International Committee of 1898. The International Committee found that the so-called Coutchi- ching of Lawson on the east end of Shoal Lake consisted of con- glomerates and other clastic sediments which unconformably overlie the Keewatin. They then concluded that all of the rocks mapped by Lawson as Coutchiching are not below the Keewatin. In consequence of Lawson’s restudy of 1911, he persists in classifying the rocks of the Rainy Lake area as Archean. He holds to this classification because he regards it as historically correct, having been, he claims, the usage of Logan in his map of the north shore of Lake Huron, and furthermore, he believes that the erosion interval which intervenes between the rocks of the Animikie series and those which precede it is the most conspicu- ous in the pre-Cambrian rocks of the Lake Superior region. He believes that the rocks on the far side of this erosion interval show greater metamorphism and more intense folding and a larger number of intrusions than those on the near side of this interval. On re-examining the Coutchiching rocks on the east side of Shoal Lake, he finds that the conclusions of the International Committee are correct for this particular locality. He finds no evidence, however, to change his original conclusion regarding the Coutchiching which is wrapped around domes of intrusive granite and which dips under the Keewatin at a low angle in the region of Rice Bay and around Bear’s Passage. Lawson’s classification of the rocks of the Rainy Lake district follows. Keweenawan—Diabase dikes { Granite, porphyritic, and syenite gneisses, and a basic facies of | syenite ( Lamprophyric rocks Huronian Quartzite and slate, and schists (Seine series) | Conglomerate Laurentian—Granite and granite gneiss Anorthite Algoman Hornblende gabbro Limestone (one seam) Greenstone, greenstone schists, felsite, sericite schist, ash beds, agglomerate, siliceous slates and schists, chert, mica schist Coutchiching mica schist, paragneiss, and phyllite Archean Keewatin | PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 655 Lawson‘ classifies the pre-Cambrian rocks of Steeprock Lake, Ontario, as follows: Algonkian Keweenawan Erosion interval Animikie Eparchean interval ( Granite gneiss, intrusive in the Seine series Irruptive contact Seine series Acute deformation and erosion interval Steeprock series Erosion interval Granite gneiss, intrusive in the Keewatin Irruptive contact Keewatin Coutchiching Archean The Steeprock series comprise interbedded sediments and irruptive rocks: dark-gray slate, agglomerate, greenstones and green schists, conglomerates, and limestone. Van Hise and Leith have correlated them as Lower Huronian. Lawson describes certain radial calcareous and siliceous fossil structures of the limestones. The rays of these fossils extend to a roughly circular limit in section normal to the axis of the organism. In oblique sections, the border is usually elliptical. In some cases, they are cornucopia-shaped, the rays sometimes showing conical or elliptical septa. The rays vary from one to fifteen inches in length. The paper is supplemented by descriptive notes on the fossils by W. D. Walcott. Miller and Knight? discuss the metallogenetic epochs of Ontario. Most of the metal production comes from Keweenawan rocks and consists chiefly of silver, nickel, and copper in the order named. Next in importance are the Algoman gold-bearing granite intru- sions which have been productive at Porcupine and many other places. The Keewatin has furnished a small tonnage of iron ore. tA. C. Lawson, “‘The Geology of Steeprock Lake, Ontario,” Canada Dept. of Mines, Mem. No. 28 (1912), 23 pp., 2 pls. 2W. G. Miller and C. W. Knight, “‘Metallogenetic Epochs in the Pre-Cambrian of Ontario,” Ontario Bur. Mines, Ann. Rept., Vol. XXIV (1915), Part I, pp. 244-48, map; Roy. Soc. Canada Trans., 3d Ser., Vol. IX (December, 1915), Section IV, PP. 241-40, 1 fig. (map). 656 EDWARD STEIDTMANN Erosion has destroyed much ore. Some ore has been preserved by folding and faulting. Miller and Knight’s™ classification of the pre-Cambrian of Ontario follows. Keweenawan Unconformity Animikean Includes rocks called Animikie heretofore, also Logan’s type section, and the Cobalt and Ramsay Lake series. Minor unconformities occur within the Animikean Great unconformity (Algoman granite and gneiss Igneous contact) Laurentian of some authors, the Lorrain granite of Cobalt, and the Killarney granite of Lake Huron Timiskamian Includes sedimentaries of various localities heretofore called Huronian—also the Sudbury series of Coleman Great unconformity Same order as that at base of Animikie (Laurentian granite and gneiss) Igneous contact { Grenville (sedimentary) agi | Keewatin (igneous) The authors differ from Collins and Coleman in that the latter recognize a twofold division of the Animikean group. Other differ- ences are largely a matter of names and emphasis on the relative importance of various features. Lawson, Coleman, and Collins emphasize the unconformity at the base of the Animikean and recognize two major groups. Miller and Knight stand alone in concluding that the Grenville of southeastern Ontario is in part interlayered, but largely above the Keewatin. Other authors are either less confident or express doubt as to the position of the Grenville. Parsons? describes the productive iron deposits of the Michipi- coten district. 1 W. G. Miller and C. W. Knight, ‘‘Revision of Pre-Cambrian Classification in Ontario’, Jour. Geol., Vol. XXIII (1915), pp. 585-99. 2A. L. Parsons, ‘‘The Productive Area of the Michipicoten Iron Ranges (Ontario), Ontario Bur. Mines, Ann. Rept., Vol. XXIV (1915), Part I, pp. 185-213, 22 figs., 3 pls., maps. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 6 57 Quirke* maps the Espafiola area representing the eastward extension of the original Huronian. His classification of the pre- Cambrian follows. Great unconformity _ Huronian Keweenawan—Diabase injection Igneous contact Cobalt series Gowganda formation—Massive slate Conglomerate members: £2.) .15),). eee ee 400 feet KS AN ACKe SAL CHn 2) RSS 2.2 \%s 2g dian ae PDS oe ee eC 400 feet Beddedsconelomeratencs.).s- sts 52 eae ee 650 feet Unconformity Bruce series DCUMC MIRGUAGLZILe rn sts clei ania sf edhe ea eee eee 8,000 ( ?) feet Espanola group HES ATO ag MMACSLOMEM Pe tte es wise opais ae Aahoe eee 25 feet ISAO ARP TAVWACK Ee: 4 Si Sect tidak ats he kt Res Be ae 280 feet BSCE MEINE SCORE Arye Nees te Salo! ev AMA 2A, aes Ee 150 feet Slight unconformity WINGSISSA PI CGUATEZACE so. <.2hcls Simo. s hs sche cbeeoe's show ols Cee 4000 feet Great unconformity Pre-Huronian Granite intrusions Igneous contact Basic intrusions Igneous contact Schistified sediments Large-scale maps were made by Stansfield? of certain mica, apatite deposits in the townships of Hull and Buckingham of Ottawa Valley. The pre-Cambrian rocks underlying this area comprise: Igneous intrusives c) Trap dykes b) Gabbro and pegmatites of the mineral deposits a) Older pegmatite veins Grenville series Ottawa gneiss The Ottawa gneiss includes granite and syenite gneiss, crystal- line limestones, quartzites, garnet-gneisses, sillimanite gneisses, tTerence T. Quirke, “‘Espafiola District, Ontario,” Canada Geol. Surv. Mem. No. 102 (1917), 75 pp. 6 pls., 8 figs., map. 2 John Stansfield, “‘Certain Mica, Graphite, and Apatite Deposits of the Ottawa Valley,” etc., Canada Geol. Surv. Summ. Rept. 1911 (1912), pp. 280-85. 658 EDWARD STEIDTMANN and certain unidentified gneisses have been listed as the con- stituents of the Grenville series in this area. The Cobalt series' comprises a basal conglomerate, resting on a nearly level surface and an assemblage of arkose, quartzite, graywacke, and argillite. Gradations are found in both vertical. and horizontal directions. The finer-grained bedded varieties are assigned to a lacustrine origin. The heterogeneous, angular conglomerates with ‘‘soled,” and occasionally striated pebbles are believed to be glacial. __ The Larder Lake’ district located on the boundary between Ontario and Quebec, about thirty miles north of Lake Timis- kaming shows the following succession of rocks, according to Morley E. Wilson: Pleistocene and recent Gravel, sand, clay, and till Huronian Conglomerate Graywacke Arkose Conglomerate Igneous contact Diabase, gabbro, syenite porphyry; the first two probably time equiva- lents of similar rocks in the Cobalt district Laurentian Granite, gneiss, granodiorite, pegmatite, aplite Unconformity Keewatin—Greenstones and green- Pontiac schists composed of biotite stone schists largely of effusive origin, and quartz. Relation to Kee- Larder slate and dolomite quartz por- watin unknown phyry, rhyolite and aplite intrusive into the preceding Igneous contact Wilson? argues against widespread correlations of pre-Cambrian rocks and urges that for the present local names should be given to series and formations. t Morley E. Wilson, ‘‘The Cobalt Series: Its Character and Origin,” Jour. Geol., Vol. XXI (February-March, 1913), pp. 121-41, 3 figs. 2 Morley E. Wilson, ‘Geology and Economic Resources, Larder Lake District, Ontario,” Canada Geol. Surv. Mem. No. 17 (1912), 62 pp., 11 pls., 5 figs., 2 maps. 3 Morley E. Wilson, “‘Sub-Provincial Limitations of Pre~-Cambrian Nomenclature in the Saint Lawrence Basin” (Abstract), Budl. Geol. Soc. Am., Vol. XXIX (1918), PP. 90-91. [To be continued] REVIEWS Oil Investigations in 1917 and 1918. Bulletin 49. Illinois Geologi- calesunvey,, 1909) Pp: 144; The volume consists of five papers bearing on the oil and gas of Illinois. The first, ‘‘ Petroleum in Illinois in ror7 and 1918,” by N. O. Barret, contains statistics of the economic phase of the oil industry. The salient facts, as summarized in the report, are (1) that in ro1r7 Illinois fell in rank from fourth to fifth among oil-producing states, due to the actual decline that same year in Illinois production, and to the enormous increase in oil production in Kansas in that year; (2) that in t918 with further decrease of production in Illinois and a notable increase in production in Louisiana, Illinois fell to sixth place as far as quantity was concerned; (3) that evidence of the high grade of Illinois oil is found in the fact that Illinois ranked fourth and fifth in value of product in 1917 and 1918, when it ranked fifth and sixth in quantity of oil produced. The three following papers, “Brown County” and ‘‘Goodhope and Laharpe Quadrangles,” by Merle L. Noble, and ‘Parts of Pike and Adams Counties,” by Horace N. Coryell, are reports dealing with the geology of the areas mentioned with particular reference to oil and gas possibilities. All the reports have good structural maps. In the areas described there are four possible oil and gas horizons, (1) the Pottsville conglomerate, (2) the Niagaran dolomite, (3) the Hoing sand (Silurian, just below the Niagaran), (4) the Maquoketa shale and Galena Platts- ville limestone or dolomite (Ordovician). The Hoing sand has furnished the best showing of oil, but prospecting in it is hazardous, owing to the discontinuous and lenticular nature of the sand. The other formations have furnished only slight showings of oil and gas in this territory. In Warren County gas occurs in small quantities in the glacial drift; the gas probably was derived from the decomposition of vegetable matter buried in the drift, and no large amounts are to be expected. The fifth paper, “‘Experiments in Water Control in the Flat Rock Pool, Crawford County,” by F. B. Tough, S. H. Williston, and T. E. Savage (in co-operation with the U.S. Bureau of Mines), is a statement of investigation and work done in regard to corrective work in water control in oil wells. Various aspects of the problem are discussed, including the use of mud fluid and cement in water control. R.A. J. 659 RECENT PUBLICATIONS —Grover, N. C. Surface Water Supply of the United States, 1916. Part XII. North Pacific Drainage Basins. A. Pacific Basins in Washington and Upper Columbia River Basin. [U.S. Geological Survey, Water-Supply Paper 442. (Prepared in co-operation with the states of Washington, Montana, and Idaho.) Washington, 1or19.]| —GroveER, N. C., et al. Surface Water Supply of the United States, 1917. Part IV. St. Lawrence River Basin. [U.S. Geological Survey, Water- Supply Paper 454. (Prepared in co-operation with the states of Minne- sota, Wisconsin, New York and Vermont.) Washington, 1919.| ———. Surface Water Supply of the United States, 1916. Part VI. Mis- souri River Basin. [U.S. Geological Survey, Water-Supply Paper 436. (Prepared in co-operation with the states of Colorado, Montana and Wyoming.) Washington, 1919.] Surface Water Supply of the United States, 1917. Part VIII. Western Gulf of Mexico Basins. U.S. Geological Survey, Water-Supply Paper 458. (Prepared in co-operation with the state of Texas.) Wash- ington, 1919.] Surface Water Supply of the United States, 1916, Part X. The Great Basin. [U.S. Geological Survey, Water-Supply Paper 440. (Prepared in co-operation with the states of Utah, Nevada, California, Oregon, and Wyoming.) Washington, 1919.] Surface Water Supply of the United States, 1916. Part XII. North Pacific Drainage Basins. B. Snake River Basin. [U.S. Geological Survey, Water-Supply Paper 443. (Prepared in co-operation with the states of Oregon, Nevada, and Washington.) Washington, 19109.] —Grover, N. C., and Hoyt, W. G. Surface Water Supply of the United States, 1917. Part V. Hudson Bay and Upper Mississippi River Basins. [U.S. Geological Survey, Water-Supply Paper 455. (Prepared in co-operation with the states of Minnesota, Wisconsin, Iowa, and Illinois.) Washington, 1o19.| —HAarkeEr, ALFRED. Petrology for Students. An Introduction to the Study of Rocks under the Microscope. sth ed., revised. [Cambridge: Uni- versity Press, 1919.] —HARNSBERGER, T. K. The Geology and Coal Resources of the Coal-bearing Portion of Tazewell County, Virginia. [Virginia Geological Survey, Bulletin No. 19. (Prepared in co-operation with the U.S. Geological Survey.) Charlottesville, 1919.] 660 RECENT PUBLICATIONS 661 —Hawaiian Volcano Observatory, Monthly Bulletin of the. Vols. VI and VII. [Honolulu, ro19.] —HEIkes, V. C. Gold, Silver, Copper, Lead, and Zinc in Arizona in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918. Part I, Sec. 15, pp. 329-68. Washington, 1920.] Gold, Silver, Copper, Lead, and Zinc in Nevada in 1918. Mines Report. [U.S. Geological Survey, Mineral Resources of the United States, 1918. Part I, Sec. 12, pp. 217-64. Washington, 1920.] —HENDERSON, C. W. Gold, Silver, Copper, Lead, and Zinc in New Mexico and Texas in 1917. [U.S. Geological Survey, Mineral Resources of the United States, 1917, Part I, Sec. 24. Mines Report. Washington, 1919.] Gold, Silver, Copper, Lead, and Zinc in New Mexico and Texas in 1918. Mines Report. [U.S. Geological Survey, Mineral Resources of. the United States, 1918. Part I, Sec. 14, pp. 303-28. Washington, 1920.] Gold, Silver, Copper, and Lead in South Dakota and Wyoming in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part I, Sec. 8. Washington, 1919.] —Hewett, D. F. Manganese and Manganiferous Ores in 1917. [U.S. Geological Survey, Mineral Resources of the United States, 1917, Part I, Sec. 23. Washington, ro19.| —Hii1, J. M. Arsenic, Bismuth, Selenium, and Tellurium in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part I, Sec. 9, pp. 193-99. Washington, 1919.] Gold, Silver, Copper, Lead and Zinc in the Fastern States in 1918. Mines Report. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part I, Sec. 11, pp. 211-15. Washington, 1920.] Platinum and Allied Metals in 1018. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part I, Sec. 10, pp. 201-9. Washington, 1oro.| —HILLEBRAND, W.F. The Analysis of Silicate and Carbonate Rocks. [U.S. Geological Survey, Bulletin 700. Washington, 1919.] —HoweE, J. L., anp Hortz, H.C. Bibliography of the Metals of the Platinum Group—Platinum, Palladium, Iridium, Rhodium, Osmium, Ruthenium, 1748-1917. [U.S. Geological Survey, Bulletin 694. Washington, 1919.] —HvENE, FRIEDRICH VON. Beitrige zur Kenntnis der Ichthyosaurier im deutschen Muschelkalk. [Separat-Abdruck aus Palaeontographica, Beitrige zur Naturgeschichte der Vorzeit. Herausgegeben von J. F. Pompeckj in Tiibingen. LXII. Band. [Stuttgart: E. Schweizer- bartsche Verlagsbuchhandlung. 1916.| Beitrage zur Kenntnis einiger Saurischier der schwabischen Trias. [Separat-Abdruck aus dem Neuen Jahrbuch fiir Mineralogie, Geologie und Paldontologie. Jahrg. 1915. Band 1. (S. 1-27 und Taf. I-VII.)] Coelurosaurier-Reste aus dem unteren Muschelkalk. [Separat- Abdruck aus dem Centralblatt f. Min. etc. Jahrg. 1914.] 662 RECENT PUBLICATIONS —HUENE, FRIEDRICH VON. Eine interessante Wirbeltierfauna im Buntsand- stein des Schwarzwaldes. [Separat-Abdruck aus dem _ Centralblatt f. Mineralogie, etc. Jahrg. 1917. No. 4.] ; Neue Beschreibung von Ctenosaurus aus dem Gé6éttinger Bunt- sandstein. [Separat-Abdruck aus dem Centralblatt f. Min. etc. Jahrg. 1914. No. 16.] —TIllinois Geological Survey. Bulletin No. 40. Oil Investigations in 1917 and 1918. [Urbana, ro19.]| —Inter-America. [English: Vol. III, No. 2. New York: Doubleday, Page & Co. December, 19109.] —Jack, R. L. The Phosphate Deposits of South Australia. [Geological Survey of South Australia, Bulletin No. 7. Adelaide, r919.] —Jrans, J.H. Problems of Cosmogony and Stellar Dynamics. [Cambridge: University Press. 1910.] —Jones, E. L. A Reconnaissance of the Pine Creek District, Idaho. [U.S. Geological Survey, Bulletin 710-A. Washington, 1919.] —Jones, E. L., Jr. Deposits of Manganese Ore in New Mexico. [U.S. Geological Survey, Bulletin 710-B. Washington, 1919.] Deposits of Manganese Ore in Southeastern California. [U.S. Geological Survey, Bulletin 710-E. Washington, 1919.] —Jonss, E. L., Jr., aND Ransome, F. L. Deposits of Manganese Ore in Arizona. [U.S. Geological Survey, Bulletin 710-D. Washington, 1920.] —Karz, F. J. Silica in tor8. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part II, Sec. 17, pp. 379-84. Washington, 1919.| —Knowtton, F. H. A Catalogue of the Mesozoic and Cenozoic Plants of North America. [U.S. Geological Survey, Bulletin 696. Washington, ‘T9109. ] —Leonarp, A. G. Possibilities of Oil and Gas in North Dakota. [North Dakota Geological Survey, Bulletin 1. Grand Forks, 1920.] —LesHer, C. E. Coal in 1917. Part B. Distribution and Consumption. Te wEss Geological Survey, Mineral Resources of the United States, 1917, Part II, Sec. 35, pp. 1203-59. Washington, 1919.] —Lovucutin, G. F.,and Coons, A.T. Slatein1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part II, Sec. 11. Washing- ton, 1919.] —McCasxkevy, H. D., and Burcuarp, E. F. Our Mineral Supplies. [U-S. Geological Survey, Bulletin 666. Washington, 1o19.| —Matort, C. A. The “American Bottoms” Region of Eastern Greene County, Indiana—A Type Unit in Southern Indiana Physiography. [Indiana University Studies, Vol. VI, March, 1919. Study No. 4o. (Price, 25 cents.) Bloomington, 1919.] —Manninc, V. H. Experiment Stations of the Bureau of Mines. [U.S. Bureau of Mines, Bulletin 175. Washington, 1919.] RECENT PUBLICATIONS 663 —Martin, G. C. Gold, Silver, Copper, and Lead in Alaska in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part I, Sec. 6. Mines Report. Washington, 1o19.| —Martin, G. C., and others. Mineral Resources of Alaska. Report on Progress of Investigations in 1917. [U.S. Geological Survey, Bulletin 692. Washington, ror19.| —Maryland Geological Survey. Report on the Physical Features of Anne Arundel County. With topographical, geological, and agricultural soil maps. [Baltimore, 1916.| —Morcan, P. G. The Limestone and Phosphate Resources of New Zealand (considered principally in relation to agriculture). Part I. Limestone. [New Zealand Department of Mines, Geological Survey Branch, Bulletin No. 22 (New Series). Wellington, r1o19.] —NIckKLEs, J. M. Bibliography of North American Geology for 1918, with Subject Index. [U.S. Geological Survey, Bulletin 698. Washington, 1919.| —Norrnrop, J. D. Petroleum in 1017. [U.S. Geological Survey, Mineral Resources of the United States, 1917, Part II, Sec. 31. Washington, ro19.] —QOspon,C.C. Peatin1g18. [U.S. Geological Survey, Mineral Resources of the United States, r918, Part II, Sec. 15, pp. 331-56. Washington, 1919.] Peat in the Dismal Swamp, Virginia and North Carolina. [U.S. Geological Survey, Bulletin 711-C. Washington, 1919.] —PARDEE, J. T., AND JonES, E. L., Jk. Deposits of Manganese Ore in Nevada. [U.S. Geological Survey, Bulletin 710-F. Washington, 1920.] —RansomeE, F.L. Quicksilverin 1918. With a Supplementary Bibliography by Isabel P. Evans. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part I, Sec. 7. Washington, ror9.] —Resources of Tennessee. Vol. IX (1o19), nos. 1, 2, 4. —RicHarpson, W.D. The Singing Sands of Lake Michigan. [Science, N.S., Vol. L, No. 1300, pp. 493-95. Lancaster, Pa., Nov. 28, 1919.] —Rocers, G. S. The Sunset-Midway Oil Field, California. Part II. Geochemical Relations of the Oil, Gas, and Water. [U.S. Geological Survey, Professional Paper 117. Washington, 1919.] —SaveEr, C. O. The Geography of the Ozark Highland of Missouri. [The Geographic Society of Chicago, Bulletin No. 7. The University of Chicago Press. Chicago, 1920. (Price $3.00 net; postpaid, $3.12).] —SCHWENNESEN, A. T. Geology and water Resources of the Gila and San Carlos Valleys in the San Carlos Indian Reservation, Arizona. [U-S. Geological Survey, Water-Supply Paper 450-A. Washington, rgro9.] —SEaArs, J. D. Deposits of Manganese Ore in Costa Rica and Panama. [U.S. Geological Survey, Bulletin 710-C. Washington, 1919.] —Srwarp, A. C. Fossil Plants: A Text-book for Students of Botany and Geology. Vol. IV, Ginkgoales, Coniferales, Gnetales. [Cambridge: University Press. 1919.] 664 RECENT PUBLICATIONS —Siupson, E. S. Sources of Industrial Potash in Western Australia. With Two Appendixes. 1. Examination of Western Australia Seaweeds for Potash and Iodine, by I. H. Boas; 2. Alunite Deposits at Kanowna, by T. Blatchford. [Western Australia Geological Survey, Bulletin No. 77. Perth, 1g19.] —Smitu, G.O. Fortieth Annual Report of the Director of the United States Geological Survey to the Secretary of the Interior, for the Fiscal Year Ended June 30, 1919. [U.S. Geological Survey. Washington, r19109.] —SmitH, P. S. Sulphur and Pyrites in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part II, Sec. 16, pp. 357-77. Washington, 1919.| —STONE, R. W. Gypsum in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part II, Sec. 12. Washington, roro.| Sand and Gravel in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918. Part II, Sec. 13. Washington, 1919. | —Tuompson, D. G. Ground Water in Lanfair Valley, California. [U.S. Geological Survey, Water-Supply Paper 450-B. Washington, 1920.] —Tuomson, J. A. Secrets of Animal Life. New York: Henry Holt & Co. IQIQ. —WapsworTH, J. M. Removal of the Lighter Hydrocarbons from Petroleum by Continuous Distillation, with especial reference to plants in California. U.S. Bureau of Mines Bulletin 162. Petroleum Technology 45. Wash- ington, 19109. —WALTHER, J. Allgemeine Palaeontologie. Geologische Fragen in biolog- ischer Betrachtung. 1. Teil: Die Fossilen als Einschliisse der Gesteine. Berlin: Gebriider Borntraeger, W 35. Schdneberger Ufer 12a. 1919. Preisea 120 Visigehe —Warp, L. K. Annual Report of the Director of Mines and Government Geologist of South Australia for 1918. Adelaide, 19109. —Washington Academy of Sciences, Journal of the. Vol. 9. [Easton, Pa.: 211 Church St.] : —WELIs, R. C. Sodium and Sodium Compounds in 1918. [U.S. Geological Survey, Mineral Resources of the United States, 1918, Part II, Sec. 8. Washington, 1919.| —Western Australia Geological Survey. Annual Progress Report of the Geological Survey for the Year 1917. [Perth, 1918.] —wWisconsin Academy of Sciences, Arts and Letters, Transactions of. Vol. XIX, Parts I and II. [Madison, 1918.] ~ CONTRIBUTIONS FROM WALKER MUSEUM ~ ANEW SERIES—VOLUME II ae No. 1. The Structure and Relationships of Diplocaulus. | _~ With plates. By Herman Douthitt. 28 cents postpaid. No. 2. Labidosaurus Cope, a Lower Permian Cotylosaur aR from Texas. By Samuel W. Williston. 28 cents postpaid. i te _ No.3. The Phylogeny and Classification of ae be - By Samuel W. Williston. 28 cents postpaid. Bs. ‘phe No. 4. (2) The Evolution of Vertebrae ; (2) The Osteology . at 5 e . ae of Some American Permian Vertebrates. III. By Samuel | _ W. Williston. 52 cents postpaid. ae x a tee TITLES IN VOLUME I 1 No. x. The Vertebrates from the Permian Bone Structure of the Fore Foot of Dimetrodon. a Bed of Vermilion County, Illinois. With plates. By E. C. Case. 27 cents postpaid. f ByE.C.Case. 53 cents postpaid. — No. 7. Notes on Some Carboniferous Cochlio- 1 §©>© No. 2. Prodromites, A New Ammonite Genus donts. With descriptions of seven new species. # «fromthe Lower Carboniferous. Withplates. By With plates. By E. B. Branson. 27 cents post- | James Perrin Smith and Stuart Weller. 27 cents paid. | _—s~postpaid. No. 8. The Osteology of Some -American Per- 1 No. 3. Palaeontological Nain With plates. mian Vertebrates. By Samuel W. Williston. 29 | By E,C. Case. 27 cents postpaid. i geen ae oe i | a - age QO. (O22 Gn e Osteology of Some erican | coal oi eae Pa Se ie BOE G Permian Vertebrates. II. (2) Synopsis of the Narain 27 cents postpaid ae pea plea peat By — | ea 5 syne i amue . Williston. 29 cents postpaid. Ws oe oe coer eae Ory aetas Eee No. to. (x) Atactocrinus, a New Crinoid Genus ey sess ae piatesee CUAREWVCHET |) 27, Cents from the Richmond of Illinois. (2) Description | —- Postpaid. of a Ste. Genevieve Limestone Fauna from Monroe No. 6. (t) The Osteology of the Skull of the County, Illinois. By Stuart Weller. 28 cents Pelycosaurian Coons, Dimetrodon. (2) On the ‘postpaid. ign Vol. I, 274 pages, cloth; $2.50 (bostpard $2. 70) Brn UNIVERSITY OF CHICAGO PRESS _ CHICAGO Ren BSE 2 ILLINOIS FINE INKS 4x2 ADHESIVES For those who KNOW Drawing Inks Eternal Writing Ink Engrossing Ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste Office Paste Vegetable Glue, Etc. Are the Finest and Best Inks and Adhesives Emancipate yourself from the use of corrosive and Today, as always, the typewriter user < | who wishes to reach the lowest ‘cos nes level of typing must ae to Mite Rem- || ington. i The Self-Statting Rewubeae for" cor- | respondence, the Key-Set Remington || for form work, and the Remington " Accounting Machine (Wahl Mechan-— | u ism) are the latest achievements i in | ill-smelling inks and adhesives and adopt the clerical labor He Higgins Inks and Adhesives. They willbea revelation to you, they are so sweet, clean, well _ put up, and withal so efficient. 3 At Dealers Generally ih i ‘Remington Toews Company. Gncorneronee CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, London 374 Broadway 271 Ninth Street Brooklyn, N.Y. | Ege York rk | American Permian Vertebrates 4 By SAMUEL W. WILLISTON. a 23) Vole notes and descriptions of new or little-known amphibians. and e a Sa reptiles from the Permian deposits of Texas and New Mexico. | The material upon which these studies are based was for the | most part collected during recent years by field parties from the University , | of Chicago. The book is offered as a contribution to knowledge on the | subject of ancient reptiles and amphibians, with such summaries and P| definitions based chiefly on American forms as our present knowledge per- sl mits. The work is illustrated by the author. — a ei | "i 2 HIS work comprises a series of monographic canine swith “biietee : | S vit 146 pages and 38 plates, 8vo, cloth; $2.50, pa oS Ba = THE ea Si Le waigg ween PRESS CHICAGO - ILL . NUMBER 8 “THE. EDITED By ; "THOMAS C. CHAMBERLIN AND ROLLIN D. SALISB te ae mi With the Active Collaboration of : roLLER, Invertebrate Paleontology ALBERT. JOHANNSEN, Petrology Podieg ae ROLLIN T. CHAMBERLIN, Dynamic Geology ASSOCIATE EDITORS | JOHN C. BRANNER, Leland Stanford Junior University - RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. WILLIAM-H. HOBBS, University of Michigan — FRANK D. ADAMS, McGill University r CHARLES K. LEITH, University of Wisconsin WALLACE W. ATWOOD, Harvard University WILLIAM H. EMMONS, University of Minnesota } ARTHUR L. DAY, Carnegie Institution El iors D DEGEER, Sweden WORT! ‘NOVEMBER-DECEMBER 1920 SM AND THE FORMATIVE PROCESSES. XIII. THE BEARINGS OF E AND RATE OF INFALL OF PLANETESIMALS ON THE MOLTEN OR A us S es a - 3 T. C. CHAMBERLIN 665 TOCENE PENEPLAIN IN THE COASTAL PLAIN- - Hernan F. CLELAND 702 pecsl INTERPRETATION OF JOINTS - - - Watrer H. BucHeR 707 Hf omy - - - - - - - i piese W. Guonne 731 IES OF PRE- CAMBRIAN LITERATURE OF NORTH AMERICA EpWArp STEIDTMANN 743 - = - - - - - = - - arts - 752 Bee TE Make eri) in, alot oe Lt ee et amg CHICAGO, ILLINOIS, U.S.A. Rotini 1 THE CAMBRIDGE UNIVERSITY PRESS, Lonpon THE MARUZEN-KABUSHIKI- KAISHA, Tokyo, Osaka, Kyoto, Fukuoka, seal fi pea THE MISSION BOOK COMPANY, SHANGHAI . Mo rceueraerciice 1920 THE JOURNAL OF GEO ae EDITED By _ THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY. With the Active Collaboration of STUART WELLER ALBERT JOHANNSEN Invertebrate Paleontology ; Petrology Sere T. CHAMBERLIN Dynamic Geology March 15, May 1, June 15, ce I, Genteaber 15, November 1, December rs. price is $4.00 per year; the price of single copies is 65 cents. Orders for service of less cae a alf Y year will be charged at the single-copy rate. | Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawai Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. { Postage is charged extra as follo For Canada, 30 cents on annual subscriptions (total $4.30), on single copies 4 cents (total 69 cent Fa for all other countries in the Postal Union, 53 cents.on annual subscriptions (total $4. 53), on sing] copies 11 cents (total 76 cents). {| Patrons are requested to make all remittances payable - The Uni versity of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, ‘Lodo: E. C. 4. Yessy subscriptions, including postage, 21s. each; single copies, including postage, 35. 6d.each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 9.00 » each; single copies, as postage, Yen 1.50 each. aa For China: The Mission Book Company, 13 North Szechuen Road, Shanghai. Yearly ‘sub- scriptions $4. 00; single copies 65 cents, or their equivalents in Chinese money. Postage extra, i if. ) mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies 11 cents. _ eae Tae Claims for missing numbers should be made within the month following the regular month of publi- cation. The publishers expect to supply missing numbers free only when losses have been = transit, and when the reserve stock will permit. ae Business correspondence should be addressed to The University of Caees Press, Chicago, Ill. oo Communications for the editors and manuscripts should be addressed to the Editors of THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, ll. i i , Contributors are requested to write scientific and proper names with particular care and i in citations to follow the form shown in the pages of the Journal. Separates, if desired, must be ordered in advance of publication. Separates of cae articles will be printed as ordered, of which 50 (without covers) will be furnished gratis, the actual cost of the remainder Bs (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon is the amount of work in remaking the pages into forms, presswork, paper, binding, etc. _ Separates | con- taining half-tones may be expected to cost somewhat more, the increase depending. upan the number 4 of cuts and the amount of work required upon them. — 4 From time to time supplements to the Journal will be issued. One on The ian of the e Feldspars, by H. L. Alling (about 100 pages) will appear shortly, and will be sent without ase to | subscribers. ta Se es age oe { ee ak i Entered as second-class matter, March 20, 1893, at the Post-office at Chicago, tl. under the Act of March 3 1879 i: Acceptance for mailing at specu rate of postage bypyided for in Section 1103, Act of October ce 1017 july 15. roT8 ; Among articles to appear in early numbers of the Journal of Geology are the following: Summaries of Pre-Cambrian Literature of North America. Papers V-VII. By Epwarp STEIDTMANN. Features of a Body of Anorthosite-Gabbro in Northern New York. By Wiitam J. MILLER. A Glacial Gravel Seam in Limestone at Ripon, Wisconsin. By {F. T. THWAITES. The Physical Chemistry of the Crystallization and Magmatic Differ- entiation of Igneous Rocks. By J. H. L. Voct. A New Form of Diplocaulus. By M. G. MEnt. Volcanic Earthquakes. By CHARLES DavISsON. The Mechanical Interpretation of Joints. PartII. By WALTER H. BUCHER. Types of Rocky Mountain Structure in Southeastern Idaho. By GEORGE RoGERS MANSFIELD. The Stratigraphic and Faunal Relationships of the Meganos Group, ' Middle Eocene of California. By Bruce L. CLARK. Cycles of Erosion in the Piedmont Province of Pennsylvania. By j F. BAScoM. Russell Fork Fault of Southwest Virginia. By CueEstTER K. WENTWORTH. Strand Markings in the (Pennsylvanian) Sandstones of Osage County, Oklahoma. By. SWNEY PowERs. Diastrophism and the Formative Processes. XIV. Megadiastro- _—.. phism. By T. C. CHAMBERLIN and R. T. CHAMBERLIN. Goode’s Base Maps and Graphs A series of outline maps for all classes of work in applied sciences and the various fields ofresearch. Prepared by J. Paul Goode, Professor of Geography, University of Chicago. These maps are adapted for use in every grade from the university to the common school: in geography, including commercial or economic geography, in physiography, geology, botany, zodlogy, anthropology and ethnology, sociology, economics, politics, and history. The maps have been prepared by being first drawn on a large scale, to insure accuracy of detail, and then greatly reduced in the engraving. In the quality of the drawing they are superior to most maps used in books and magazines. All maps are perforated for insertion in a standard notebook cover with page 8 X 104 inches. The larger sizes fold once or twice into this cover. A new classification has been arranged, but it should be noted that the maps themselves have not been altered. New maps and graphs are shown in blackface type. SERIES I SERIES It SERIES II SERIES IV Maps I Io1 yee eee The World: on Mercator’s projection. torHc Lard 301Hc The World (continents): Homalographic projection. Re 201HcE..... The World (continents, Eastern half): Homalo- graphic projection. 2zo1HcW_..... The World (continents, Western half): Homalo- graphic projection. 4 to1Ho Lee ome The World (oceans): Homalographic projection. I01S “ee ye The World (continents): Sinusoidal projection. ro1P Me Bits The World in Polar Hemispheres: Lambert’s azimuthal, equal-area projection. 201PN Ree The World (North polar hemisphere): Lambert’s projection. 201PS ed The World (South polar hemisphere): Lambert’s projection. 201PB Reet Same map as 1o1P showing Isobars for January and July. 2 102 202 eee North America: on Lambert’s azimuthal projection. 3 103 203 Sica South America: Sanson’s projection. 4 104 204 304 Europe: conic projection. 5 105 205 atte Asia: Lambert’s equal-area projection 6 106 206 ond Africa: Sanson’s projection. 7 107 Wee Bake Australasia: Mercator’s projection. 9 109 eras 300 United States of America: conic projection. ia 2090E aa United States of America (Eastern half): conic projection. 209W Aa United States of America (Western half): conic projection. Io IIo Sas ae United States of America: state-outlines only: conic projection. ; III ts says Canada: conic projection. an 112 es eth Mexico: conic projection. 14 114 iss Ma The British Isles: conic projection. ais Peet, 215 sal Europe, Northwestern: conic projection. 16 116 216 Kae Europe, Western and Southern: conic projection. 7) 117 see oe France: conic projection. 18 118 Woe are The Spanish Peninsula: conic projection. IQ IQ ver Loot Italy: conic projection. 20 ar Best Site Central Europe: conic projection. 21 121 Ase Bape The German Empire: conic projection. 24 124 eG ah The Levant: conic projection. 132 232 Ree United States of America, by counties with names: conic projection. ue 133 a: Gre Chicago and vicinity: conic projection. 51 I51 aa ieee North Dakota: conic projection. GRAPHS 80 as re Pes Climatic Chart (yearly). ae 18r Bees eye Monthly Weather Chart. THE UNIVERSITY OF CHICAGO PRESS, CHICAGO, ILLINOIS VOLUME XXVIII NuMBER 8 THE POR NAL OF GEOLOGY NOVEMBER-DECEMBER 1920 DIASTROPHISM AND THE FORMATIVE PROCESSES XIII. THE BEARINGS OF THE SIZE AND RATE OF INFALL OF PLANETESIMALS ON THE MOLTEN OR SOLID STATE OF THE EARTH T. C. CHAMBERLIN University of Chicago In the last article of this series' it was found (1) that the solar gases, as they were expelled to form the planetary systems, were so mixed that they were unfitted to form solid bodies such as the terrestrial planets, the planetoids, and the satellites, until after they had been sifted by a selective process, (2) that the sifting process introduced such a serious departure from familiar modes of gaseous condensation as to require reinterpretation, (3) that the process of concentration was also complicated by inherited motions, (4) that it was still further conditioned by the formation of pre- cipitates and precipitate aggregates, and (5) that the planetary cores, while in process of formation, were subjected to viselike squeezing, more intense below than above, followed by partial relaxation, so that selective extrusion attended the closing pro- cesses, involving the ascent of the lighter mobile matter and the compression and reorganization of the rest, thus contributing «“Diastrophism and the Formative Processes. XII. ‘The Physical States of the Planetary Nuclei during Their Formative Stages,” Jour. Geol., Vol. XXVIII (1920), pp. 473-504. 665 666. T. C. CHAMBERLIN — toward high density, rigidity, and elasticity in the central parts. It was further found that the shapes of the planetary cores were influenced from the very outset by the gyratory system of circula- tion that attended their formation, and that they thus failed to take on strictly spherical forms, so that they were likely to yield unsymmetrically to the heavy masses later built upon them by planetesimal growth. Even the primitive circulation thus had its influence on the diastrophism that developed much later. Let us now consider the planetesimal growth. This involves (1) a study of the nature of the planetesimals at the start, (2) the conditions that affected the mode and extent of their growth, and (3) the modes and rates of their infall and the effects of these on the molten or solid state of the earth, as also on its content of explosive gases. THE NATURE OF THE PLANETESIMALS AT THE START The way in which the planetesimals are supposed to have arisen has been made clear in previous articles, but it will facilitate our present study to note that they took their starts from two main sources: (1) solar molecules driven into orbits by the original solar expulsion, and (2) molecules thrown out into orbits from the nuclei later by molecular interaction. There were other sources of planetesimals, but they may be neglected here. In both classes the planetesimals started as molecules chiefly. To some extent they may have been newly formed precipitates from the solar gases, or precipitate aggregates formed by the union of the fresh pre- cipitates. Such precipitates are thought to form in the sun’s photosphere now. They would be likely to have been formed by the expansion of the solar gases just after these emerged from solar pressure. The essential point here is that, whether molecules, precipitates, or precipitate aggregates, they were minute. Whether they afterward grew to notable sizes depended on the conditions that controlled their later history. Chief among the controlling influences were the dynamic properties given the planetesimals by their expulsion, and the gravitative stresses that controlled the field into which they were driven. It is to be kept ever in mind that they were bodies projected into swift independent flight, each DIASTROPHISM AND THE FORMATIVE PROCESSES 667 in its own path under control of its own inertia and the gravitative stresses of its environment. The planetesimals shot out from the nuclei had the simpler history, and are easiest followed to gain typical pictures of planetesimal behavior as a basis for estimates of their modes and rates of infall. Let us picture the earth nucleus as pursuing a nearly circular orbit about the sun while certain of its outer molecules were escaping from it in various directions by reason of exceptional velocities given them by cumulative successions of rebounds from favorable collisions. It is easy to fall into the error of supposing that these molecules, thus escaping in different directions, would take orbits quite discordant with that of the nucleus and thus pass into the meteoritic rather than the planetesimal class. As constituents of the nucleus, they already had motions relative to the sun, and of course carried these with them they went into orbits of their own, except in so far as these motions were reduced or increased by their ejection from the nucleus. The velocity of the nucleus in its orbit should have been of the order of eighteen miles per second, and that of all the molecules of the nucleus about the same, some a little more, some a little less, by reason of their participation in rotation, et cet. It was the new and additional velocity which the escaping molecule had been given, measured at the border of the sphere of control of the nucleus, which determined its orbit after it had escaped. Only in extremely rare cases would molecular interaction give to an escaping molecule a speed greater than the parabolic velocity respecting the nucleus, that is a velocity sufficient to carry the molecule to infinity so far as the restraining attraction of the nucleus was concerned. ‘The parabolic velocity of even the full-grown earth at the border of its sphere of control is 1.75 miles per second, so that we leave a large margin of safety if we assume that molecules almost never were shot away from the border of the sphere cf control of the nucleus at more than two miles per second. Now, as the nucleus was moving at eighteen miles per second relative to the sun, a molecule shot directly back- ward would still have a velocity of sixteen miles per second rela- tive to the sun, and a molecule shot directly forward would have a velocity of twenty miles per second relative to the sun, while 668 T. C. CHAMBERLIN those shot out at lesser velocities, or sidewise at various angles, would have intermediate velocities. All would therefore be moving in the same general direction as the nucleus and their orbits would still be similar. The molecules in these new orbits would therefore be planetesimals, because they would revolve about the sun in orbits similar to those of the planets. The kinetic theory of gases requires us to suppose that molecules of the lighter gases escaped from the outer border of the earth- nucleus with some degree of frequency while it was hot and diffuse, and that such molecules have continued to escape from the outer border of the earth’s atmosphere ever since, but much less fre- quently. And so of all other planets that have atmospheres, and of the sun as well. Practically all the molecules that thus escaped into orbits still remained within the sphere of control of the sun and were liable in time to be picked up again, so that this whole system of escape and recovery constitutes a mode of exchange of atmospheric material between the domains of the sun and the planets. It contributes to the maintenance and equilibrium of our atmosphere as elsewhere set forth. Let us now look to the gathering in of the planetesimals in this typical case, for that is the vital point here. Under the laws of mechanics the planetesimals shot forth in this way would come back to the virtual points of their escape at the end of each revolu- tion in their new orbits, unless they were diverted by some inter- vening influence. From this it is easy to jump to the conclusion that they would all soon be picked up again by the nucleus, but not so, in general. They were nearly all thrown into larger or smaller orbits and that determined the time at which they should get back to the point of their origin. In most cases this was either earlier or later than the return of the nucleus and hence recapture was avoided. Figure 1 and the accompanying periodic data, pre- pared by Dr. MacMillan, make this very clear. From the point A at the top of the figure let one molecule be shot forward at a speed Io per cent greater than that of the nucleus and let another mole- cule be shot backward so that its velocity shall be 10 per cent less than that of the nucleus. The first molecule will take the outer * The Origin of the Earth (1916), pp. 13-17. DIASTROPHISM AND THE FORMATIVE PROCESSES 669 orbit, the second, the inner orbit. The molecule in the inner orbit will get back to A in .77 of the time required by the nucleus to reach this point, while the molecule in the outer orbit will require 1.424 times that period, i.e., if it takes the nucleus three hundred and sixty-five days to complete its orbit, the planestesimal that was shot backward and took the inner orbit would reach A eighty-four days ahead of the nucleus, while the molecule that was shot ahead and took the larger orbit would return to A one hundred A Fic. 1.—O.N. represents the orbit of the nucleus; O.0.P., the orbit of the outer planetesimal; O.I.P., the orbit of the inner planetesimal. Periodic times, earth=1; inner planetesimal=o.77; outer planetesimal=1.424 (MacMillan). and fifty-five days after the nucleus had passed. There was therefore no immediate danger of collision and recapture in either case. Only by waiting for a concurrence of the schedules, which would be liable to be thwarted by perturbations, or by a pro- tracted series of orbital shiftings, or by changes of orbital form or dimensions, could this be brought about. Were these planetesimal molecules likely to unite with one another in the course of their flights and so grow to larger sizes? The figure illustrates this point also. It is very obvious that the two planetesimals specified have no opportunity at all to unite 670 T. C. CHAMBERLIN © with one another until they return to the vicinity of A. But one of them would reach A two hundred and thirty-nine days earlier than the other. Even in this specially favorable case when they had a common node there was no immediate opportunity for union. Of course cases of less divergence could be chosen in which there was a nearer coincidence of orbits and of time schedules, and if there were many orbits there would be some real crossings farther from the nucleus, but those chosen illustrate the prevalent fact that even though such bodies have similar orbits and sometimes actual crossings, they may yet remain independent for long periods. Their mutual attractions would, in general, aid in bringing this about ultimately, but instead of this they might be brought into co-ordinate orbits like those of the earth and moon and revolve together in harmony indefinitely. At best the process was likely to be a very slow one. The picture of molecules drawn directly together, as in the case of static bodies or of gases, 1s very com- monly substituted for the real case, and is very misleading. When all possible cases are considered, as well as the multitude of planet- esimals, there are enough chances of collision and coalescence, especially with the nuclei, to make the process of ingathering effect- ive in the course of long periods, but in its very nature it cannot be a speedy process. When planetesimal molecules or even precipitate aggregates collide, rebound would be more likely to follow than coalescence, unless they were electrically charged. Coalescence almost inevitably follows collision with nuclei but not encounters between planetesimals. This simple example of the evolution and behavior of planet- esimals illustrates the mechanism by which they are maintained and the contingencies of their capture or their mutual coalescence, where the conditions are exceptionally favorable. For the case most important in the formation of the earth, we must turn to the solar molecules which were driven directly into orbits by the original propulsion from the sun under the stimulus and attraction of the co-operating body. These were subject to the law of return to the points of their origin, but they were greatly diverted by the pull of the co-operating body and so largely lost all such systematic relations to a given nucleus as those that made the previous case DIASTROPHISM AND THE FORMATIVE PROCESSES 671 so simple and instructive. The orbits in this case were distributed through greater space and more irregularly, and hence their coalescence with one another and their capture by the planetary nuclei, as a rule, required greater changes in the forms, dimensions, and attitudes of their orbits. We will turn to concrete specifica- tions and numerical values presently. THE SIZE OF PLANETESIMALS NOT IMPORTANT IN RESPECT TO MELTING EFFECTS Lest we stress the growth of planetesimals too much, it is prudent to observe at once that the sizes of the planetesimals were not matters of vital moment so far as the total energy-effect of their infall was concerned, for whatever was gained by concen- tration of mass was lost by less frequent infalls. On the whole, less energy available for conversion into heat was carried into the earth by the united planetesimals than by the same mass ununited, for in coming together, energy of motion was converted into heat and this was dissipated at the point of union in open space; the combined mass carried so much less energy into the earth-core. However, whenever combination took place, there was relatively less resistance and heating of the atmosphere in plunging through it, and so relatively more heating of the surface of the earth. As we shall see a little later, however, the chief effect at the earth’s surface was lateral dispersion and an elastic or explosive reaction, resulting in a great scattering of débris with little obvious melting. None the less, we shall consider the melting effects of large planet- esimals as well as small ones. THE LIGHT SHED ON SIZE BY EXISTING PLANETESIMALS It has been shown in previous papers" that the union of mole- cules and the growth of small aggregates could take place with more or less facility up to a certain order of size, but that beyond such order conditions unfavorable to further growth arose and increased relatively, so that indefinite growth was probably limited, as a general rule. It appeared that chemical, electrical, and t Article X, this Journal, Vol. XXVIII, No. 2 (February-March, 1920), Pp. 140-44. 672 T. C. CHAMBERLIN cohesive attractions functioned effectively in the early stages; but that fragmentation, abrasion, and exfoliation came into increasing effectiveness as larger sizes were attained. Theoretically, then, growth from the minute state at which the planetesimals started, took place presumably up to limited sizes with relative facility, beyond which the presumption of much larger growth was adverse, except under unusual conditions. Theory, however, does not define at all closely where the balance between the opposing agencies was to be found, and so we turn to naturalistic evidence which is more decisive. 1. The zodiacal planetesimals.—It is an accepted view that the zodiacal light is due to the reflection of solar light from minute solid or liquid particles distributed in a lenslike form about the sun. The central plane of the lens is essentially coincident with the common plane of the planetary system. The outer border of the lens reaches to some undetermined distance beyond the earth. Under favorable conditions, it is possible to trace the counter- glow (Gegenschein), on the side of the earth opposite the sun, into continuity with the zodiacal light on the sunward side. It is not improbable that the edge of the lens is extremely attenuated and extends indefinitely outward in the plane of the planets. The form and extent of the lens in the planetary plane make it scarcely less than certain that the particles are sustained by orbital dynamics, and that the orbits are of the planetary type and that hence the particles are planetesimals. This warrants us in turning to them for light on the sizes and masses of planetesimals. Their testimony is obvious for they are certainly quite small. Though they envelop the earth and must in many cases be quite near, the individual planetesimals are too small for detection. Although they are certainly very numerous, their joint mass is not known to affect the motions of any body. They are inter- preted either as remnants of the original planetesimal system, or as more recent products due to the projection of solar matter so close to the planets that it is drawn forward by them into elliptical orbits about the sun. If the first view is correct, or in so far as it is correct, the planetesimals are exceptionally old and should DIASTROPHISM AND THE FORMATIVE PROCESSES 673 have reached the fullest growth to which they are ordinarily subject. If they are of more recent origin, they merely bear testimony to the common size to which planetesimals of the younger order attain. But as they are so obviously minute their testimony in either case is weighty. 2. The satellitesimals of Saturn’s rings.—Satellitesimals are merely special forms of planetesimals. It is convenient to distin- guish between them in certain cases, while in other cases the generic term planestesimal is most satisfactory. In the Saturnian case, they are notable for their very close association with one another and for their definite borders. These hint at a special origin, per- haps the disintegration of a satellite by the differential attraction of Saturn, since they lie within its Roche limit. At any rate, the closeness of the individual satellitesimals to one another gives them rather pointed bearing on the question of growth to large sizes, for their nearness to one another should favor this, if mutual attraction has any appreciable effect. According to Bell’s studies of their albedo, they are chiefly very minute particles.t Only rarely is there evidence of masses reaching as much as a meter in diameter. Trituration as a consequence of their mutual collisions is probably the dominant size-controlling agency in this case. 3. Precipitate aggregates formed from condensing gases. The chondrules.—When the gases or vapors of stony and metallic sub- stances mixed with lighter gases were expelled from the sun into the vacuum of interplanetary space, they must probably have been greatly expanded and cooled and the stony and metallic sub- stances thrown down as precipitates at successive stages according as the appropriate temperatures were reached. As each gaseous substance was diffused through the others, the precipitates could at first have been little more than molecular in size, but by subse- quent interaction, in the fashion of Brownian particles, the first precipitates were brought into contact with one another and in their fresh, hot, viscous states should have united into larger aggregates rather freely. In so far as they solidified, they naturally t Louis Bell, “‘The Physical Interpretation of Albedo, II, Saturn’s Rings,”’ Astro- phys. Jour., Vol. L (July, 1919), pp. 1-22. C7 T. C. CHAMBERLIN — took the form of concretions or of crystals. I have ventured to suggest that this may be the mode in which chondrules—little organized bodies that enter into the formation of 90 per cent of known meteorites—were formed. Later I shall suggest that the formation of the common small meteors of the sky may have taken place in practically the same way as planetesimals, i.e., by the progressive aggregation of precipitates from the stony and metallic ingredients of solar gases shot into interplanetary space and there cooled, the distinction between the two being their orbital characters and planetary relations. If these suggestions are in the line of truth, the chondrules give very specific evidence on the usual sizes of planetesimals, for they range from the size of a walnut down to fine dustlike particles. 4. The negative evidence——Concurrent with these concrete sources of evidence, supported by theory, is the significant fact that no bodies of a distinctly larger or planetoidal order of magni- tude are seen to revolve in the region of the earth’s or bit or within it. We have found reasons for suspecting that the normal dynamic stresses in this inner solar region have always been too great to permit the collective aggregation of precipitate clouds of so little mass. At any rate, the negative testimony of observation stands against any view that postulates an abundance of bodies of plan- etoidal size in the region of the earth or their effective participation in the formative processes of the earth or the moon. Let us then assume that the chondrules are our best natural- istic guide in respect to the normal sizes of planetesimals. For definiteness in the computations that follow, let us assume as a convenient representative weight, one-fiftieth of a pound, say one- third of an ounce, or about 9 grams. In testing the probability, or otherwise, that planetesimal infalls would produce a molten state of the growing earth-core, it will make no essential difference whether the planetesimals were somewhat larger or somewhat smaller than this size which is adopted merely for definiteness and convenience. After inspection of the melting effects in a case thus made as nearly normal as we conveniently can, we will try to test the effects of supposedly larger planetesimals. DIASTROPHISM AND THE FORMATIVE PROCESSES 675 THE TIME OVER- WHICH THE INGATHERING OF PLANETESIMALS WAS SPREAD _ It is obvious that the time over which the infall of planet- esimals was spread is an essential factor in determining whether the heat of their infall would melt the earth surface or not. And so if there are any naturalistic evidences bearing on this point, they should be brought under consideration at once so that they may serve as guides or tests where assumptions have to be made in trying to deduce the period of infall from the mechanics of the case. Successful study of earth history has been found to rest much more largely on naturalistic considerations than on deduc- tion, especially when the premises involve so much that is assump- tive. An approach along naturalistic lines may be found in biologic evolution combined with geologic chronology. 1. The intimations of biologic evolution.—It seems to be the con- sensus of opinion among those best fitted to judge that the portion of life-evolution that has taken place since the faunas and floras of the early Paleozoic offered a fair criterion for judgment, is of the order of one-tenth of the total life-evolution, or some such pro- portion. This proportion will therefore be made the basis of the time-scale used in the following discussion. It will be easy to modify the results of the computations to suit any other proportion that may be thought nearer the reality. I do not think that any other proportion which is tenable will change the general tenor of the conclusions, so far as these bear on the melting effects of planetesimal infall. Two geologic time-scales are now in use, an older one built on estimates of the present rates of geological progress, and a newer one built on radioactive processes. For myself I regard the latter as much the more trustworthy. ‘The former seems to me to need radical correction (1) for the exceptional speed of present denuda- tion due to the stripping of very large portions of the surface of its native protection, and (2) for the exceptional speed induced by the present high relief of the surface brought about by recent diastrophism.t But let us use both scales. Those who prefer t See Article VIII of this series, ‘‘The Quantitative Element in Circumcontinental Growth,” this Journal, Vol. XXII (1914), pp. 516-26. 676 T. C. CHAMBERLIN the old scale will doubtless concede that the proportions of the radioactive scale may be used safely as a means of extending the older scale over the Proterozoic and Archean eras, where its own criteria are not available. Using the radioactive scale, the begin- ning of the Paleozoic may be placed, in round figures, at 410° years ago, the beginning of the Proterozoic at 12 X10° years, and the oldest portion of the Archean that has been determined in respect to age, at 16 X10° years. Using the old scale, the beginning of the Paleozoic may be placed at 108 years and—using the radio- active scale for proportionate extension—the beginning of the - Proterozoic at 310° years, and the earliest determined Archean at 4X10° years. To fill out the total period of life-evolution on the radioactive scale, an allowance of 24X10° years previous to the earliest determined Archean must be made, making the total life- period 4X10? years. To similarly fill out the total period of life- evolution on the old scale, 610° years is to be allowed previous to the earliest determined Archean, and 10 X 10° years for the whole life-period. In thus using the proportions of biologic evolution as an indica- tion of the period over which the growing stage of the earth was spread, it is to be noticed that, to avoid making the assigned period of planetesimal infall unfairly long by including too much of the tailing-out stage, I shall consider all planetesimals that fell during the last 400,000,000 years (radioactive scale) of the Archean era, and all that have fallen since, 16 X10° years in all, as though they had fallen within the computed period. On the other hand, all that fell during the nuclear stage, i.e., before a definite earth-core was formed, are necessarily excluded from the estimated period of life-evolution, since the conditions were incompatible with life. No doubt the infall during the distinctly nebulous portion of the nuclear stage may have been more rapid than during the biologic stage but that does not concern us in considering the period of- biologic activity preceding the earliest age-determined Archean. It is the special merit of the planetesimal hypothesis that it takes due account of biological requirements, as generously interpreted as the leaders in biological inquiry demand. The biological DIASTROPHISM AND THE FORMATIVE PROCESSES 677 evidences are regarded as among the most cogent that bear upon the duration of the early history of the earth. Qualified and defined as thus specified, the period of effective planetesimal infall subsequent to the nuclear stage is made to range from 600,000,000 years on the old geological scale, to 2,400, 000,000 years on what seems to me to be the more probable radio- active scale. These assignments of time may impress some readers as very long, but the question is to be asked anew, are they longer than the biological evidence requires? We shall soon inquire whether they are any longer than the mechanics of the case warrant. But before passing on, it is to be noted that the interpretation of biological evolution should no longer suffer from duress due to supposed limitations of time, such as were vigorously urged during the last half of the last century by advocates of the contractional theory of the sun’s heat and of other physical tenets which were really less well grounded than the biological and geological inter- pretations. This alien stress is now not only lifted, but a new theoretical urgency of precisely opposite import has taken its place, a seemingly imperative need to find a source of heat for the maintenance of the stars of such potency as will enable them to serve their indicated functions in the protracted history of star clusters and our stellar galaxy. For this, a stellar longevity of the order of ten billion years, or some such great period, seems to be required. Short of trespassing on some such time allowance as that, biology and geology cannot be said to be necessarily restricted for lack of solar endurance. The seeming demand of biological and geological evidences for a total earth age of three or four billion years need not be thought extravagant or unreasonable, if either class of evidence is found to really require it. 2. The intimations of the planetesimal mechanism.—Let us now turn to the planetesimal mechanism to see what may be its most probable time requirements. Neglecting planetesimals of high and unusual orbital range, a fair and at the same time conservative working approximation to the extent of that portion of the planet- esimal field which was tributary to the earth, may be made by 678 T. C. CHAMBERLIN taking the width of the tract now occupied by the planetoids as its breadth, and for its depth the limits of the earth’s dominant attraction in competition with that of Mars on the outside, and that of Venus on the inside. These give roundly 55 X10° miles in breadth and 58 Xz10° in depth. They define the cross-section of the planetesimal ring which curved around the sun with the path of the earth-core near its center. The actual field was much larger than this, but the planetesimals outside these limits are neglected to compensate for any lateral thinning inside. The area of the cross-section was therefore roundly 3 X10"5 square miles and its curved length 292 X10° miles. For a working case of the medium order, let the mass of the earth-core, at the beginning of the specified period of planetesimal infall, be taken as one-third of the final earth-mass, leaving two-thirds of the earth-mass in the form of planetesimals to be gathered in. It will be seen that this propor- tion makes the mass of the planetesimals large and favors effective infall. It is taken merely as a fair working basis without any intention of implying an opinion as to the ratio of the nuclear to the planetesimal portions which actually obtained; that may best be reserved for further study. ‘Taking the masses and dimensions of Mars and Venus as guides, in accordance with our comparative studies (Article X), the earth-core should have had a diameter of about 6,000 miles. Its disk would then have an area of 28 X 10° square miles, roundly. This is the fleeting target which the widely scattered planetesimals must hit, if they were to take part in the earth-building, or to change the simile, this is the area of the sweeper that must gather in the planetesimals from their vast field to build its one-third mass up to a three-thirds mass. 1. As rigorous treatment is impracticable, modes of approxima- tion are our only recourse; and so, as a simple and purely artificial first approach, suited to give a realistic impression of the immensity of the field that must be swept, let us suppose that the planetesimals stand still while the earth-disk sweeps through it at its normal speed, changing its path in such an effective way as to clean up an entirely new swath at each revolution. Even by this impossibly speedy method, 100,000,000 years, roundly, would be required. DIASTROPHISM AND THE FORMATIVE PROCESSES 679 2. To make a first approach of a natural kind that can be treated mathematically, Dr. MacMillan has suggested that the planetesimals might be treated as though they were particles of gas which would close in upon the track of the earth-core as it revolved through the center of the tract, though the dynamics of gases are radically different from those of planetesimals, and cor- rections must be made accordingly. To gather in all the planet- esimals under these conditions would take an indefinite period; to gather in go per cent would require somewhat over 260,000,000 years. Keeping in mind that this is not the real case, but merely one that can easily be treated, it is worth while to note that one- fifteenth of the earth-mass would fall in after 260,000,000 years had passed and that nine-fifteenths would be systematically dis- tributed over this period with infall greatest at the start in due proportion, but it does not give warrant for excessive concentra- tion in the early stages. If that is assumed, it makes the more certain a non-melting rate in the later stages and the infall during these would furnish the outer shell of the earth to a depth beyond the reach of most problems of immediate geologic interest. The vital point, however, is that in this substitute case, like the real one, the laws of mechanics require a distribution of infall over long periods. 3. The next step toward the real case is the substitution of heterogeneously revolving particles for the previous gaseous par- ticles. A gaseous organization is a failing structure in the sense that when any inner portion of it is removed the rest collapses sufficiently to fill the space. In an orbital organization no such collapse takes place, each remaining body is sustained in its orbit by its own moving force. This makes a radical difference in the rate of ingathering by a body like the earth-core in the case in hand. Those planetesimals whose paths had actual crossings with that of the earth-core would be picked up, if not disturbed by per- turbation, whenever their time schedules became coincident at the crossing, but not before, normally. Those planetesimals—by far the greater number—which had no such actual crossings at the start, would circle through their independent orbits indefinitely, 680 T. C. CHAMBERLIN if they were not thrown out by collision, which would be rare in such vast space, or perturbed by other bodies, among which the earth-core would be the most influential in most cases. But such perturbations work very slowly, and their effects on the orbits involved are not easily visualized by any except experts in orbital dynamics. It is easy, however, to see that the case is far different from the direct collapse of gaseous particles and that it must occupy much greater time. In the lack of any rigorous determination of just how much longer the ingathering process would take, we may merely note that if it be taken as no more than two or three times longer, the total period would at least equal the biological require- ments given above on the older geological scale. But such bodies in heterogeneous orbits belong to the meteor- itic type, and would not arise normally from the dynamic influences postulated by the planetesimal hypothesis, nor would their aggre- gation give rise to planets in concurrent revolution, for lack of the requisite moment of momentum, unless it were assigned them by some supplementary hypothesis such as revolution of the whole assemblage. As in the preceding case this assumption only serves as a step toward the real case. 4. The distinctive feature of the postulated planetesimals was that they were moving in the same general direction as the collecting body and at the same general rate of speed. ‘The process of collection was therefore confined to overtakes and to convergencies of orbits. The differences between this and the preceding case may be com- pared to the different degrees of danger of collision between auto- mobiles when, in one case, they are running in a common direction, on the right side of the road, under fairly well regulated speeds, and, in the other case, running wildly at random in both direc- tion and speed. So planetesimals, circling about the sun in more or less concurrent orbits, only collide and coalesce in so far as they deviate from concurrence with the rest of the system or are per- turbed in their independent orbits and drawn into coalescence by overtakes or convergencies. In so far as their orbits were con- current, the moment of momentum of the combined mass was nearly as high as the sum of the individual moments of momenta, and so, if at any stage the orbits became adjusted to one another, they DIASTROPHISM AND THE FORMATIVE PROCESSES 681 might revolve in harmony indefinitely, as do the earth and moon. It was this concurrency of movement that made the evolution of a planetary system highly endowed with moment of momentum a possibility. Therein lies the soul of the planetesimal theory. But evolution under these conditions requires great lapses of time. But lest this be overstressed, it is to be noted that the sub- parallelism of orbits and the subequality of speeds gave greater effect to the mutual attractions of the earth-core and the planet- esimals, and so tended either to bring them together or else into harmoniously adjusted orbits, such as those of the earth and its satellite. Compared with the much more familiar gaseous and meteoritic types, the fundamental tendency of a planetesimal system is not so much direct concentration as concurrent revolution, though, in so far as the nuclei are competent, they gather in the smaller bodies. It seems clear, therefore, that the time required for collecting the planetesimals would be some multiple of that assigned in the preceding case. It is not clear just how large it would be, but if taken at three or four, the total time requirement would equal the maximum estimate of the biological requirement. In the nature of the case, it should not be less, for life-evolution could not proceed until a solid core was formed and the rate of infalling planetesimals permitted a congenial temperature. At any rate, however large may be the latitude for different numerical estimates of the total time and rate of planetesimal infall, it is altogether clear that a precipitate ingathering is incompatible with the mechanics of the planetesimal system. THE RATES OF PLANETESIMAL INFALL a) The infall of normal planetesimals.—We have already found reasons for thinking that the planetesimals were usually small, as their name implies, and have chosen one-fiftieth of a pound as a working figure. We have also chosen one-third of the total mass of the earth as the amount of material already in the earth-core and two-thirds as the amount still in the form of planetesimals at the beginning of the specified period of infall. The mass of the present earth is 6X10 tons. There would then be 4X10” planetesimals of the specified mass to be gathered 682 T. C. CHAMBERLIN into the core to complete the growth of the earth. The earth-core, taken at 6,000 miles in diameter, would have a surface area of 3 X10'S square feet. As there were 4X10” planetesimals in all, 13X10 planetesimals must fall upon each foot of earth-core surface, on the average, to build the body up to its present mass. | Now, if we take the total period of infall, as given above on the radio-geo-biologic scale, at 2.4X10® years, a planetesimal one-fiftieth of a pound in weight, falling upon each square foot every 6.7 days, or a little less than once a week, would have com- pleted the growth of the earth in the time specified. It will be agreed, I think, that this does not remotely approach a rate suff- cient to melt the earth surface. If there is any doubt as to the dissipation of energy following the impact of a falling body, see later discussions. If we take as the period of infall the biological requirements as estimated on the older geologic time-scale, 6 X 108 years, a planet- esimal falling upon a square foot once in about forty hours would build the earth up to its present mass in the time estimated. This again, I think it will be agreed, is not near the melting-rate for the general surface. If we make the time of infall equal to the highest of our range of estimates from the mechanics of the case, 3X10? years, an average fall of a planetesimal on each square foot once in a little over eight days would suffice, or if we take the minimum of the estimates, 18 10°, a planetesimal once in about five days would answer, in either case far from a general melting-rate. If we fall back upon the untenable assumption that the planet- esimals distributed themselves after the manner of gaseous par- ticles—made merely as a first step in approach—and take the computed 26 X 107 years as the total time, the average rate of infall upon each square foot would be about one planetesimal in seven- teen hours. Even this does not seem to be a rate that would threaten the melting of the earth, and yet it is much more rapid than is permitted by the mechanics of the real case under the basal assumptions made. DIASTROPHISM AND THE FORMATIVE PROCESSES 683 Let us now reverse the mode of inquiry by trying to approximate a rate of infall that would cause the melting of the earth surface, and then compare results with those reached in the preceding ways. If the mass of the earth-core equaled one-third that of the present earth, an atmosphere of sufficient depth to protect its sur- face from the direct impact of planetesimals of the specified mass would have surrounded it. The melting of the earth must then have hung upon the competency of the infall to so heat the upper atmosphere as to melt the earth surface some miles below. About half the heat acquired by the thin upper air would have been quite promptly radiated outward and the melting left to the other half. The effect of the air on meteorites plunging into it is suggestive in this connection. As soon as a film of meteorite-substance becomes viscous enough to yield to the high pressure of the air condensed on the meteorite’s front by its high speed, the film is driven backward and dissipated along the meteor’s path forming the “‘streak”’ of the “shooting star.” Only a very small part is melted at any one instant, or left in any one spot. Even this minute part only reaches the first stages of the molten state and hence is very quickly cooled again to the solid state. To apply this to planetesimals, it is to be noted that the mean velocity of meteorities is probably four or five times that of normal planet- esimals, and their moving energies sixteen to twenty-five times as great in proportion to mass. The working picture, then, in the case in hand, is that of a little mass, one-fiftieth of a pound, mak- ing a similar but feebler streak of quickly heated, quickly cooled matter, down the center of a column of air one square foot in cross-section. This must take place in such close succession as to melt one square foot of the earth surface at the bottom of the atmosphere in spite of outward radiation. To really complete the picture, it is necessary to add that the lower atmosphere would soon be filled with the dust of the dissipated planetesimals and the melting of the surface would have to be effected through this screen. It seems clear that to effect general melting the upper atmosphere must be heated throughout to the melting-point of average rock-substance, and kept at that temperature in spite of 684 T. C. CHAMBERLIN convection and radiation. As radiation increases with the fourth power of the temperature, it would be very effective as the red- hot stage was approached. As the case is beyond the reach of experiment or rigorous com- putation, specific estimates of rate can be little more than matters of judgment. Let us therefore resort to the serial method, which sometimes leads to a decisive conclusion even when definite quan- titative values are unavailable.t Let each reader fix upon such rate of infall as seems to him competent to produce a molten state of the earth surface under the given conditions. Let us then see how such a rate fits into the range of rates which the mechanics of the case permits. Too great a discrepancy may be about as decisive as if the precise rates were known. ins working test is the final arbiter. If one’s assumption is that a planetesimal plunged into the upper end of each square-foot air-column once every second, the column would be built up to the present surface in 4,119 years. It will be recalled that our first, but wholly arbitrary and excep- tionally speedy mode of sweeping up the planetesimal field required 100,000,000 years, and the most speedy natural method 260,000,000, and that both of these hypothetical cases required less time than the real case. Tf one planetesimal fell upon each square foot once per minute, the total time would still be only 247,140 years. The competency of such a rate to melt the earth surface would, I think, at least be open to question. If the rate were one planetesimal per hour, the total period would be 14,828,400 years, which is about one-seventeenth of the time of ingathering required on even the gaseous assumption. Moreover this rate would give a cooling period to every column of air more than 3,000 times as long as the glowing period, estimated from the mean duration of ‘‘shooting stars.” At one planetesimal per day per square foot, the total time would be 355,881,600 years. I think it will be agreed that this rate of infall would fall far below a liquefying rate, and yet even «“’The Methods of the Earth Sciences,’ Pop. Sci. Mo. (November, 1904), pp. 70 and 71 (““The Method of Multiple Series’). DIASTROPHISM AND THE FORMATIVE PROCESSES 685 so fast a rate of infall as this does not seem to be warranted by the mechanics of the case. Apparently the only line of escape from the import of such a serial trial lies in postulating that the rate of infall in the earliest stages was sufficiently more rapid than the mean rate to effect melting in such early period. A declining rate of infall is, of course, to be presumed, and has been taken into account. The rate used in the computations is the mean rate for the specified accession when assumed to be distributed over only the period which followed the formation of the earth-core and preceded the earliest time- determined Archean, 16X10® years ago. The accessions before that period were reckoned as part of the mass of the earth-core, and the accessions since were thrown into the specified period to avoid counting the long tailing-out period of 1,600,000,000 years (radioactive scale). The period thus made the basis of compu- tation represents an intermediate stage of infall and was given the benefit constructively of all subsequent infall. We excluded such infall as was contemporaneous with the evolution of the nucleus from its nebulous state until a definite earth-core was formed, because it necessarily preceded life-evolution, and because it is not separable from nebulous condensation and the other nuclear conditions. In connection with the irregularities of the original outburst, there may have arisen some incalculable rates of infall. These would doubtless have made themselves felt chiefly in the nuclear stages. Our endeavor was to include in the computations only the systematic ingathering into which the action settled as a secular process. The physical state of the nucleus during its evolution from a nebulous state into an earth-core has been left an open question, reserved for further consideration. Meanwhile, a molten state during that period has been treated as one of the alternatives, and as a not improbable one. ‘The infall of planet- esimals during that stage may probably have been an important factor in determining the state which actually prevailed. But all that is held to antedate the growth of the outer part of the earth. This embraces about all that has yet been brought under study in geological and biological inquiries. To reach a satisfactory basis for these inquiries is the soul of the present issue. The state of 686 T. C. CHAMBERLIN the core does not radically affect most geological and petrological problems. The infall of supposedly large planetesimals.—In the foregoing tests it has been assumed that planetesimals normally grew to about the same order of size as the chondrules, and that the dis- ruptions and abrasions they suffered after reaching this size kept them down to about the order of the little masses that form “shoot- ing stars.’ Let us now consider the melting effects likely to follow if the planetesimals had grown to very much larger sizes. To keep as close to the actual as practicable, let us base our first study on the phenomena of Coon Butte, or Meteor Crater, Arizona, interpreted as the work of a gigantic meteorite, or cluster of mete- orites or, if you please, the nucleus of a comet, accepting as con- clusive, in the main, the disclosures of the drillings, shafts, and _ trenches of Barringer and Tilghman. ‘Then, let us base our second study on the craters of the moon, on the assumption—made solely for the sake of the study and without acceptance—that they were formed by the impacts of still larger bodies. Case I. The testimony of Coon Butte or Meteor Crater.—There is no reason to think that the celestial mass whose plunge into the earth formed Coon Butte was a planetesimal, because, among other reasons, it came from the northward, an unlikely direction for a planetesimal and because its indicated velocity was probably too high. The work done by it, however, is very instructive respecting the physical effects of such a falling mass under natural conditions. The essential phenomena are a circular rim of upturned strata, covered thickly by outthrown débris, 130 to 160 feet above the surrounding plain, inclosing a crater nearly 4,000 feet in diameter and 440 feet deep, measured from the original surface of the hori- zontal sandstone and limestone from which the crater was formed to the top of the present partial filling. Crushed rock, mingled « The following are among the more important papers on the subject: A. E. Fotte, Amer. Jour. of Sci., Vol. XLII (1891), p. 413; also Proc. Amer. Assoc. Adv. Sci., Vol. XL (1892), pp. 279-83; G. K. Gilbert, 13th Ann. Rept., U.S. Geol. Surv., Part I (1892), p. 98; rgth Ann. Rept., Part I, (1893), p. 187; Geol. Soc. of Wash. (President’s Address), March, 1896; Science (N.S.), Vol. III (1896), pp. 1-13; O. A. Derby, ‘Constituents of the Canyon Diablo Meteorite,” Amer. Jour. of Sci., Vol. XLIX DIASTROPHISM AND THE FORMATIVE PROCESSES 687 with meteoritic matter, lies below the floor of the crater to a depth of about 660 feet. Below this, disrupted rock seems to grade into undisturbed sandstone at points between 1,100 and 1,200 feet beneath the general plain. Rock masses and clastic material, coarse and fine, were thrown from the pit and strewn over the adjacent plain for distances of one to two miles on all sides, while meteoritic matter, distributed subconcentrically, reaches out to an extreme distance of 53 miles. The rim and pit, while subsymmetri- cal, have sufficient asymmetry to indicate an intall from a northerly direction, perhaps N. NW. to S. SE. ‘The chief mechanical effects were the formation of the crater by the breaking up of perhaps 8X 108 tons of rock, and the hurling out of perhaps half of it, the turning up to high angles of the previously horizontal limestone and sandstone beds of the crater-border, the crushing of large quantities of sandstone to silicious rock flour, and the develop- ment of some schistosity in connection with it. The chief thermal effects were the partial metamorphism of some of the rock flour and the development of incipient fusion in other portions of it, some of this portion becoming vesicular. The crushing and heating were obviously the direct effects of the impact, the upturning of the rim and projection of the débris as obviously the effects of the attending lateral thrust and the quasi-explosive reaction that followed. The energy involved in the mechanical effects must be sub- tracted from the total energy of the impact before the heating effects can be theoretically deduced. The very large sum total of these mechanical effects shows how great would be the error of computing the energy of infall in terms of heat and using that as (Feb., 1895), pp. to1-10; D. M. Barringer and B. C. Tilghman, “First Mention of the Discovery that the Crater Is an Impact Crater and Not a Crater Produced by a Steam Explosion” (President’s Statement), Proc. of Acad. Nat. Sci. (Philadelphia, Dec. 5, 1905); D. M. Barringer, “Coon Mountain and Its Crater,” Proc. Acad. Nat. Sci. (Philadelphia, Dec., 1905), pp. 861-86 (issued March 1, 1906); B. C. Tilghman, “Coon Butte, Arizona,” ibid., pp. 887-914; J. W. Mallet, Amer. Jour. of Sci., Vol. XXI (May, 1906), pp. 347-55; J. C. Branner, Science, Vol. XXIV (Sept. 21, 1906), pp. 370-71; H. L. Fairchild, at Tenth Session of the International Geological Congress, in Mexico, September 14, 1906, Compte Rendu, X Session, Congrés Géol. Inter. (Mexico, 1906), p. 147; O. C. Farrington, “Analysis of Siderite Oxides or Iron Shale,” Amer. Jour. of Sci., Vol. XXII (Oct., 1906), pp. 303-9. 688 T. C. CHAMBERLIN a measure of the melting effects. This would be a tempting line of attack but is quite inadmissible because the mechanical effects alone call for more energy than can be reasonably assigned to the meteoritic material found. The only safe recourse is the direct evidence. The heating effects implied by the direct evidence are singularly small compared with the mechanical effects. To a considerable, but not closely determined, extent, the crushed sandstone shows incipient schistosity with partial metamor- phism, obviously a compressive effect, the heat of which did not rise to the grade of fusion. To a considerably smaller extent, if I interpret the descriptions correctly, the crushed sandstone shows the early stages of fusion, while some of this portion has become inflated and pumaceous, but no appreciable masses were left in the state of glass or other completely fused product. If fully melted matter was formed at all, it was probably dispersed by the explosive reaction. It seems quite clear that the portion which became vesicular did not become fully fused and fluent, for, in part at least, the bedding lines were not wholly obliterated. These portions seem, however, to have been rendered distinctly viscous and susceptible of inflation. This must probably have taken place during the resilience which followed the compression. The internal gases could scarcely have puffed the viscous rock while the intense pressure of the impact was on. If, on the other hand, they had remained viscous until the pressure from the falling back of the exploded débris was brought to bear, they would have collapsed, at least in all deeply buried portions. Apparently they had cooled in their inflated state while the pressure was off. It seems, therefore, that there was practically no liquid rock left when the explosive reaction was over. This is a matter of radical importance in its bearings on the question of producing a holo-liquid earth by such impacts. It shows that a very high proportion of the energy of impact was converted into another mechanical form, not into heat. There is no question about the greatness of the energy of impact; the mechanical work involved in the formation of the crater and of its rim, as also in the crushing and scattering of the débris, demon- strate that. And yet there is no evidence that this violent impact left even the smallest pool of lava. The significant feature of the DIASTROPHISM AND THE FORMATIVE PROCESSES 689 case lies in its clear evidence that the energy of impact was chiefly transformed into lateral thrust and resilience of quasi-explosive type. Confessedly the most outstanding problem left is to find a source of energy adequate to the mechanical effects so impressively forced on attention. The case still remains something of a puzzle on that account. Meteoric matter has been found so widely dis- seminated through the débris, both within and without the crater, that the origin of the crater is no longer in doubt, but yet the amount of meteoric matter thus far brought into evidence seems clearly too small to be adequate. The suggestion of Barringer that the infalling mass was a cluster of meteorites or a comet’s head is plausible in itseli—and the orbits of comets are such as to make a bump into the earth a recognized contingency—but these suggestions give little help in the matter of adequacy. A larger mass than has been found seems to be required to satisfy the | effects realized. For such computations as I have made, a siderite sphere 400 to 500 feet in diameter was taken, but it is scarcely worth while to give the results here. They are of the same import as those of the next case. We ought not to overlook the fact that this is the only known case of such an infall in the history of the earth. This is an embar- rassment in postulating a rapid series of infalls. Nor is its negative bearing merely a surface matter. If such a crater had been formed and buried in a natural way in any geologic formation, however old, there would be a fair chance of its detection. There is there- fore a complete absence of geological warrant for supposing that infalls of this kind were ever anything but very sporadic affairs. Ii Meteor Crater was formed by the impact of the nucleus of a comet, theory would make its repetition an extremely rare event. The concept of an enormous meteorite, or close cluster of meteorites, other than cometic, has no observational basis. If the views respecting the origin of meteorites, later expressed in this article, have any cogency, the infall of such bodies would be governed by the same order of chances as those of comets. From no point of view, therefore, does Meteor Crater offer substantial ground for supposing that the earth was once molten because of the impacts of meteoritic bodies. 690 T. C. CHAMBERLIN Case II. The questionable intimations of the craters of the moon.— The impact theory of the craters of the moon affords a concrete basis for the study of infalls of a still larger order. To fit this case, bodies of the order of five miles in diameter, more or less, seem to be required, and for working convenience these may be given the specific gravity of the moon, 3.34. The assumed size in this case has about the same ratio to the larger order of the moon’s craters that the assumed 400 or 500 foot meteoritic body had to the size of Meteor Crater, but the mass is made relatively less to be in better accord with the moon’s mass. The size is about the lower limit assigned to planetoids. No atmosphere can be supposed to have broken the effects of infall in this case or to have checked the free dispersal of the débris. In the previous case there was surprisingly little evidence of liquefaction. What is the evidence here? ‘The steep walls of the deep craters are quite incompatible with a liquid state, so far as this outermost part is concerned and this is the part subject to direct impact. There was strength enough in the crust to support the lunar Alps and Apennines, some of whose peaks tower to heights of 20,000 feet and more above the adjacent surface, i.e., 5,000 feet higher than their terrestrial prototypes. No less than ten mountain ranges have been recognized on the moon, which implies general crustal strength. The great relief of such eleva- tions towering above such depressions is uncontrovertible evidence of strength and stability. The significance of this is emphasized, if the supposed impacts are made a part of the formative process of the moon, for then they are very old and have stood in this strong relief in spite of all the creep of the geologic ages. A search for direct evidences of molten matter gives meager results under the most favorable interpretation that is tenable. Such of the craters as have level bottoms have been thought to imply a partial filling of lava, supposed to have risen from below after the craters had been formed. These bottoms may, however, be interpreted as level beds of clastic débris, like those that form the level bottom of Meteor Crater. So, also, the seemingly smooth, but really quite accidented, plains of the ‘‘maria’’ have been inter- preted as great lava flows, but these may likewise be merely débris DIASTROPHISM AND THE FORMATIVE PROCESSES 691 plains. In the best photographs they are seen to have considerable relief and to be crisscrossed in different directions by lines of débris obviously shot from neighboring craters. They are thus at least surficially covered with clastic débris. But granting that every- thing which appears at this distance like lava really is lava, the whole does not imply a liquefaction of the moon of any other order than that signified by the great lava flows on the earth whose essen- tial solidity is now beyond question. But let us look at the question of rapid infall quantitatively and numerically. Let us assume that at the beginning of the accretion process, one-third of the mass of the moon was already in its core, while the remaining two-thirds had been gathered into bolides five miles in diameter which were yet to fallin. The mass of the moon is about 73210" tons. There would then have been 244 X10%7 tons in the moon-core and 488 X10” tons in the bolides yet to fallin. The mass of each of these bolides would have been about 997 X 10° tons, and their total number about 49 X10°. Their individual volumes would have been a little over sixty-five cubic miles, while the volume of the moon-core would have been about 14 10° cubic miles, and the radius of the core 708 miles. As the radius of the full-grown moon is 1,080 miles, the core would have had to grow radially 372 miles. Now the surface area of the moon-core would have been 6 X 10° square miles, while the disk of the five-mile bolides was a trifle less than twenty square miles in area, so that there would have been over 300,000 disk-areas on the surface of the moon-core. It would thus have required less than two hundred bolides to each disk- area to complete the full growth of the moon. The liquid-forming impact theory now takes a critical form. We have seen that the surface of the moon shows that the last craters were not attended by general liquefaction or even a viscous state of their immediate walls. The last falls, however, were accelerated by nearly the full mass of the present moon, while the first falls were accelerated by only one-third the mass of the moon. The individual effects of the last infalls should, therefore, have been greater than any that preceded. They should also have inherited whatever benefits were transmissible from previous 692 T. C. CHAMBERLIN infalls, in proportion to the time between falls. As these last impacts left no conclusive evidence of molten residue, it follows that no previous infall, in itself, can be consistently supposed to have left any greater molten residue and if their inheritance was greater it could apparently only come from a closer succession of infalls. Apparently, then, the only way in which a general molten condition can reasonably be supposed to have arisen was from the cumulative effects of such inherited residues of heat from the earlier infalls in excess of those of the later infalls. How tenable is this? ‘There were by computation less than two hundred infalls of the specified kind to each disk-area during the whole accretion period of the moon. If that accretion period were essentially the same as that of the earth, as it should theoretically be, and if we compute the rate of infall by using the minimum accretion period assigned the earth based on mechanical and biological evidences, to make the rate as high as consistent, the mean interval between impacts would be about 3,000,000 years. If the mean accretion period had been used, the mean interval between infalls would have been more than twice this time. Very little inheritance of heat from a surficial bump can be postulated over an interval of this order. But we are not left wholly to computations on estimated require- ments. There is the direct evidence of the craters themselves. Some are fresh and their débris lines lie straight across older pits and older features of all sorts. Some pits and rims are worn or buried to the very limit of recognition, and there are all grades between. These features offer no warrant for the hypothesis that there was a closely crowded infall. They distinctly imply that the formation of the visible craters stretched over a long period. This evidence is the more cogent when the limited means of denudation, owing to the absence of an atmosphere and hydrosphere on the moon, are considered. Now let us turn to the theory itself. If it be supposed that the five-mile bolides are planetesimals, the supposition itself hides under its cloak a quasi-assertion of the rate of their infall, for, as we have seen, all planetesimals started as very minute bodies controlled by a system of dynamics that imposed upon them slow growth as a DIASTROPHISM AND THE FORMATIVE PROCESSES 693 necessity of the limited amount of planetesimal matter, the large amount of space through which it was distributed, and the mutual relations of the planetesimal orbits, as already brought out. There were besides obstacles to growth beyond quite small sizes. Even if these obstacles be supposed to have been ineffectual, time for growth from the minute sizes to five-mile bolides must have inter- vened before the latter could function as crater-formers. They could thus have come into function only at a late stage. But accretion could not have been suspended in the meantime. They could therefore have come into function only as a partial source of lunar accretion. Growth from the smaller planetesimals must have gone forward during all the intervening period. Accretion simply by such giant planetesimals is thus incompatible with the funda- mental conditions postulated by the basal hypothesis on which it FESts. The hypothesis is not much more promising if planetoids are substituted for the supposed giant planetesimals, for, by the mechanics of the case, the planetoids were given courses less favor- able to aggregation than were the planetesimals, and hence greater intervals between their infalls must in consistency be assumed. This is in harmony with the observed fact that at least eight hun- dred planetoids are still following their own individual paths in a relatively limited tract and yet no collision or even dangerous approach to one another has been noted during the whole period of astronomical observation. In addition to this, we have found reasons for doubting whether planetoids could organize as nuclei of the planetary type under the differential stresses of the solar attractions that prevail in the region of the earth and in regions still nearer the sun. The hypothesis that the pits of the moon were formed by the impacts of great meteorites offers no presumption that one infall would be followed by another in the same spot within any short period. Asa cause of general melting, this is even more unpromis- ing than the preceding. The discussion thus far has proceeded on the assumption that the pits of the moon are the scars left by the impacts of great bolides of one sort or another... Before turning to the next topic, 694 T. C. CHAMBERLIN it may be well to forestall misapprehension by making clear our view that such an origin of the craters of the moon is in itself improbable, for bodies moving in orbits under the control of the sun should plunge into the moon, if they strike it at all, at various angles to the vertical. In many cases the stroke should be quite oblique to the surface and should leave elongated pits, unsymmetri- cal rims, unequal dispersions of débris, and other tell-tale features; all the more so because the moon had no atmosphere to retard and turn downward the path of the bolides. Apparently the only escape from these grave objections lies in supposing that the explo- sive reaction was so great that it completely overwhelmed the effects of the direct stroke. If this assumption were tenable, it would seem to imply that the explosive dispersion was so great that it must have scattered all mobile matter, and especially all liquid matter so effectually as to insure its cooling while in flight. THE SIGNIFICANCE OF THE EXPLOSIVE PHENOMENA OF THE MOON The moon seems to have been a paradise of Krakatoas and Katmais. Interpreted as the product of gaseous explosion, the abundance and the greatness of the craters of the moon carry special significance. They have commonly been thought to imply a once molten state of the moon. I think the argument lies in precisely the opposite direction. If the moon, in its formative stage had been a molten globe, its high temperature should have set free all gases susceptible of being freed by any temperature that ever arose afterward. Its liquid state and its convective cir- culation would have brought these gases to the surface and given them opportunities of escape never equaled later, for the high temperature of the surface would have forced unsurpassed molec- ular activity and have insured their escape from the control of the moon. Even the cold full-grown moon cannot hold the vol- canic gases. After all such gases had been boiled out of the moon and had escaped, and the gas-free lava had cooled, the moon should have been devoid of the means of explosive action. On the other hand, if the moon were built up of minute clastic particles which carried such amounts of occluded and combined gases as meteorites do, and as they naturally would from their DIASTROPHISM AND THE FORMATIVE PROCESSES 695 long flights in the ultra-atmospheric field of the sun, and if the porous surface of the moon received and held by adsorption, chemi- cal combination, or otherwise, molecular planetesimals of the gas- forming order, such as would inevitably plunge into it from the interplanetary field, there would be entrapped in the body of the moon, as it grew, a supply of disseminated gas-producing material sufficient to actuate great explosions whenever concentrated later by conditions favorable for such action. As the moon grew, its self-compression and the strains developed within it by neighboring bodies should have forced this potentially gaseous material toward the surface and developed the conditions of eruption. The moon should also have inherited its quota of radioactive substances and these should have played their part in the lunar vulcanicity. The fragmental constitution of the outer part of the moon, postulated as an inevitable feature of an accretional origin, should have rendered it specially susceptible to explosive effects. More or less local lava-production, as well as the quiet type of vulcanism, are entirely consistent with this view of the exceptionally gaseous eruptions, and they are postulated, but the evidence of the moon’s surface seems to give this more quiet action a place quite subordi- nate to the gaseo-explosive phase. THE SIGNIFICANCE OF TERRESTRIAL VULCANISM The inferences that seem so imperative in the case of the moon apply also to the gaseous phases of vulcanism on the earth. The argument is a little less imperative because the earth is able to hold an atmosphere and would probably do so to some less extent in a molten state, and so, if it were once in that state, volcanic gases could have been retained in its liquid mass sufficient to balance the partial pressures of the like gases in such lessened atmosphere as the earth then held. The amount of gases so held in equilibrium could not have been large, and such as existed would not have served as explosive agencies because of the very fact that they were held in balance by opposing pressure. They were there merely because there was an outside pressure holding them there and that outside pressure was never removed under nor- mal conditions. It seems merely declaiming the obvious to say that 696 T. C. CHAMBERLIN such a gas content is incompetent to produce explosive eruptions and that such eruptions occur on the earth only when there are special developments or accumulations of gas within or beneath the exploded matter. In a molten earth, stirred during its long cool- ing stages by effective convection," the gases set free by the various stages of heat of that period should have been so far brought to the surface and dissipated by molecular activity—except the limited equilibrium amount—that the earth when cooled and solidified should have been as deficient in explosive material as lavas are now found by experiment to be when melted in the open | air at the surface and, after a long stage of boiling, cooled to the solid state. In essence, therefore, the case of the earth is the same as that of the moon. The studies of terrestrial volcanoes of recent years have brought forth accumulating evidence that volcanoes are actuated by inborn rather than outside gases, and that they are essentially independent of one another, though of course not independent of common conditions. Their explosive- ness seems thus clearly due to their own individual resources and has no obvious dependence on any molten zone, sheet, pool, or other remnant of a once pervasive liquid state. THE TESTIMONY OF THE ABERRANT BODIES OF THE SOLAR SYSTEM We have now considered at some length the bearings of various lines of evidence drawn from the normal elements of the solar system. Let us turn for a moment to such suggestions as may be derived from the aberrant members of the system, the meteors, meteorites, and comets. If these are merely aliens that have been introduced incidentally from foreign sources, as some of them may be, there is little reason to expect them to teach much relative to the domestic organization; but if they were born in the system and are products of its dynamics, they may be quite as instructive as the normal members. To discuss them with any definiteness, however, it is necessary to postulate the modes by which they came into being. These should reveal why they are aberrant, though products of the same t See pp. 481-87 of previous article, this Journal, Vol. XXVIII (1920). DIASTROPHISM AND THE FORMATIVE PROCESSES 697 dynamics as the normal elements. I venture, therefore, to offer three hypothetical, but mutually consistent, ways in which meteors, meteorites, and comets may have arisen naturally and inevitably out of the dynamical system that gave rise to the planets as its normal product. The problem of the meteors, meteorites, and comets is regarded as essentially one. Though complete demonstration has perhaps not yet been reached, it is assumed that meteors, meteorites, and comets are not only close of kin dynamically, but in some sense mutual derivatives. The spectacular phenomena which seem to put comets in a class by themselves are here supposed to be mainly the effects of the strongly contrasted conditions to which they are subjected at the extremes of their very elongated orbits. It is a suggestive fact that those comets which are supposed to have been reduced from extremely elongated orbits to shorter ones by the action of the great planets, show a notable tendency to lose their spectacular features and finally to pass by disintegration into meteor swarms. In the case of typical comets of extremely elon- gated orbits, a small loosely organized head—apparently a cluster of still smaller bodies held together by rather feeble gravitative control—swings from a relatively hot perihelion close to the sun to a very cold aphelion far out in space. During its long outer journey, all the constituents must become intensely cold to great depths and be liable to be deeply riven by shrinkage cracks, which, besides leading to coarse fragmentation, should facilitate the adsorption of molecules belonging to the sun’s ultra-atmosphere. The action is supposed to be the same as that which gives to meteorites their occluded or combined gases. The outward swing of the comet occupies many years and often centuries, and there is time for even a very attenuated source of supply to furnish the requisite amount of gas-producing material. When later the comet head, thus charged, approaches the sun, the gases are supposed to be set free by the solar heat on the sun- ward side, and to be dr:ven forth toward the sun. At the same time, this differential heating is supposed to give rise to rapid and rather violent exfoliation, hurling the dissevered chips to 698 T. C. CHAMBERLIN considerable distances against the feeble gravity of the head. By collisions in the course of their flights, these develop a quasi- gaseous meteoritic swarm whose triturative action should give rise to products of dustlike fineness. Both of these processes would doubtless be attended by much electrical dissociation in which the negative electrons would escape and the positive remain attached, so that electric repulsion would tend to drive both gases and dust sunward until repellant action from the sun reversed the movement and drove the whole backward in the form of the comet’s tail. This sketch is very inadequate, but it may serve to suggest ways in which the distinctive features of comets may arise. The nuclei of the comets’ heads may be merely clouds or clustered groups of meteorite-like masses loosely assembled by their own - feeble attractions, and so subject to easy deployment and reassem- blage as conditions require. If the spectacular features of comets may thus be reduced to the incidental effects of extremely elongated orbits, the way is cleared for explaining how the material of the meteors, meteorites, and comet-heads may have originated, how their highly elliptical orbits were given them, and why these orbits lie in all azimuths and the bodies in them revolve indifferently in forward and retro- grade directions in contrast to the systematic, orderly, and con- current habits of the planetary bodies. 1. The first hypothesis assumes that, previous to the genesis of the present planetary system, the sun had a system of second- aries of the type which it could generate without the co-operation of any outside body. The assigned principles of such generation are those rigorous deductions from the kinetic theory of gases on which orbital ultra-atmospheres are postulated.t This class of second- aries would be, in the nature of the case, of a very much smaller order than our present planets. The orbits of such bodies would be likely to be thrown into erratic courses by the near approach of the massive body to which the origin of our present planetary system is assigned. Such of these small bodies as were thrown into very long elliptical orbits were made to suffer great extremes of heat and cold and might thus, it is postulated, have taken on t “Celestial Kinships,” The Origin of the Earth (1916), pp. 101-2. DIASTROPHISM AND THE FORMATIVE PROCESSES 699 cometic features, for a time, and later suffered dispersion into meteors and meteorites. If any of these ancestral secondaries had attained a notable size and happened to be disturbed so as to be drawn through the Roche limit of the sun, it might be disrupted and become a clustered group suited to serve as the nucleus of a comet. This, however, would not be likely to occur in many cases, and so appeal is made chiefly to the riving action of cold in the aphelion journey as the dependable cause for the disrupted character of the comets’ heads, and the fragmental features which meteorites commonly show. 2. The second hypothesis assumes that forces of the kinds dis- closed by the observations of Pettit on the solar prominences of May 29 and July 15, 1919," projected solar gases and precipitates into the outer regions of the sun’s sphere of control, where its attraction was feeble, and where the attractions of neighboring stars and star groups were relatively strong. During these outer flights, the pull of some star, group of stars, or other outside source of attraction drew the ejected masses aside from their normal paths sufficiently to cause them to swing by the sun on their return and thus be forced to take highly elliptical orbits. The planes of these orbits and the direction of revolution thus generated would be determined by the various deviating attractions, so that a system formed by a large number of such deviations would be very hetero- geneous orbitally. High ellipticity would be a common charac- teristic. The principles that control aggregation, as previously sketched in this series of papers, would apply to the projected matter in all such cases. In so far as this matter retained self- control, it would assemble by the precipitate-aggregate method into clouds of aggregates, and these would usually be still more closely assembled into loosely organized bodies well suited to function as the nuclei of comet-heads. These would be subject to all the vicissitudes of temperature and of alternate absorption and evolution of gaseous material, sketched above, and so display for a time the spectacular features of comets, and ultimately be dis- integrated into meteorites. In so far as the projected solar matter was too highly dispersed for mutual control, it should have passed t Astrophys. Jour. (Oct., 1919), pp. 206-19. 7OO T. C. CHAMBERLIN directly into meteoritic matter of the minutest type. At present, meteoritic particles, assumed to be of this type, abound in inter- planetary space in such prodigious numbers that many millions are picked up daily by the earth. The formation of precipitate aggregates, in the methods previously sketched, seems to furnish an apt explanation of the origin of chondrules and of the other minute integers that so largely make up meteorites. The collisions of these little bodies as they were entering into the formation of larger bodies, seem well fitted to account for the intimate breccia- tion, the minute specks of glass, suddenly cooled liquid drops, as well as the strange mixtures of stony and metallic matter, and other distinctive features of meteorites. _ 3. The third hypothesis is dependent on the pre-existence of the present planetary system. It supposes that the ejected solar matter passed so near some one of the more massive planets that it was thrown into an elliptical orbit in a way similar to the pre- ceding and with similar results. A certain portion of the particles so diverted would take orbits of the planetary type, so far as their planes are concerned, but only a part of these would have a planet- oidal degree of circularity. Other portions would have orbits whose eccentricities, orbital planes, and directions of revolution were as various as are those of meteorites and comets. Certain comets are known to have orbits definitely related to the giant planets. This relation is commonly interpreted as the result of reduction from larger and more eccentric orbits by the planet’s influence. Without questioning the validity of this interpreta- tion, it is not inconsistent to hold that in a part of such cases, the comets arose de novo from planetary action in the way here sug- gested. Most comets developed in this way would probably belong to the feebly developed evanescent type. These three hypotheses are entirely consistent with one another and may all be true. They have the merit of being made to rest on the same dynamic basis as the planetary system itself. These hypotheses for the aberrant factors, when added to the planetesimal hypothesis for the normal factors, give a theoretical unity to the whole solar system. DIASTROPHISM AND THE FORMATIVE PROCESSES 701 Now, if these are the true lines of interpretation, the masses of meteors and meteorites, and their methods of infall, throw a flood of light on the sizes and the modes of infall of the planetesimals, for, by this interpretation, they are bodies of like origin and like general conditions. On a conservative estimate, there are 100,000,000 or more minute meteorites, so small as to be wholly dissipated in the upper air, for every one that is massive enough to remain a visible body until it reaches the earth. Of the latter, none are known to exceed a dozen feet in mean diameter. No meteorite has ever been seen to produce melted soil or rock by its impact. When collisions with bodies that have no atmosphere take place, local melting probably results. The glassy bodies common in meteorites may very likely be such products. But the retention of such heterogeneous structures as are common in meteorites implies that there has been no general liquefaction. In so far, therefore, as the testimony of the aberrant factors bears on the size, rate of infall, and liquefying power of their dynamic relatives, the planetesimals, it supports the view that these are small, and in other respects it is in close accord with the deductions hereinbefore drawn from dynamical considerations. It is in intimate harmony with the testimony of the normal factors of the system. Professor F. R. Moulton and Dr. W. D. MacMillan have been kind enough to read the manuscript of this and the three previous articles (X, XI, and XII), and to make valuable suggestions and criticisms. They are not responsible, however, for the computa- tions. These have been verified by Miss Daisy W. Heath. A PLEISTOCENE PENEPLAIN IN THE COASTAL PLAIN HERDMAN F. CLELAND Williams College, Williamstown, Massachusetts The Black Belt of Alabama is famous throughout the state, and in the surrounding states, for its great fertility, its production of cotton and corn, the levelness of its plantations, the large pro- portion of negroes to whites, and its numerous ante-bellum man- sions—the visible manifestations of its former wealth. As one rides over the gently undulating surface of the region, with its deep black soil, and crosses the steep-sided gullies and the bluff-bordered rivers, he is impressed with the aspect of topographic youth. However, a more careful study in the field and of the geological literature forces one to the conclusion that the region is not in the youthful stage of a first cycle of erosion, nor in a mature stage of erosion, but that the surface is a recently raised plain, so flat as almost to make the term peneplain—almost a plain— inappropriate. The following excellent description will assist one in visualizing the region: The surface of the country, underlaid by the Rotten Limestone, is but little diversified; it is, however, occasionally broken into rounded bald knolls, as may be seen between Arcola and Demopolis, and between Livingston and Sumterville. The summits of these hillocks are sometimes ornamented with cedars, but more frequently they are quite bare, or covered with but a scanty vegetation; even where the surface is but slightly undulating, bald spots occur where the naked rock has come up. But the most remarkable feature of this region is the extensive tracts of land covered with a deep, black soil of great depth and extraordinary fertility, which may be seen in various parts of Sumter, Greene, Marengo, Perry, and Dallas, but more particularly in the “‘cane brake.’ The surface of these remarkable tracts has barely sufficient inclination to admit of easy drainage, without giving the water force enough to remove the soil, so that, instead of excavating a channel at the bottom of the trough-like depressions where this sort of land occurs, it is absorbed by the soil, or spreads over a considerable space, where it loses all transporting power. 702 A PLEISTOCENE PENEPLAIN IN THE COASTAL PLAIN 703 The unbroken surface of this region is due to the homogeneous character of the limestone, which suffers waste equally on this account, over considerable areas; and hence the entire absence of ravines, and other abrupt irregularities. .... In the uncleared parts of the cane brake, .... one can scarcely satisfy himself that he is not standing on the low grounds of a river; the deep, alluvial-looking soil beneath his feet, the moisture-loving long moss (Tillandsia usneoides) above his head, together with an undergrowth of Sabals, Palmettoes, and other natives of damp soils, strengthen the illusion. Professor Eugene A. Smith’s accurate and suggestive description is as follows: The Selma chalk underlies a belt entering the State from Mississippi and extending eastward with an average width of 20 to 25 miles, to a short dis- tance beyond Montgomery, where its distinctive characters are lost or merged into those of the ‘‘blue-marl region.” . . . . The somewhat uniform composi- tion of the Selma chalk has caused it to be more deeply and evenly wasted by erosion and solution than the more sandy formations north and south of it. As a consequence, its outcrop is in the shape of a trough, with a gently undu- lating, almost unbroken surface except where remnants of the once continuous Lafayette mantle have protected the underlying limestone from erosion and have thus formed knobs and ridges capped with its loams and pebbles. In this belt, more than in any other of the Coastal Plain, the soils show their residuary character. They are, as a rule, highly calcareous clays and, where much mixed with organic matters, of black color. Throughout this section are areas originally destitute of trees and hence known as “prairies.” From the agricultural point of view, the Selma chalk or black belt is the most highly favored part of the State and, apart from the cities, holds the densest population.? R. M. Harper’ characterizes the topography as “gently undu- lating in a manner difficult to describe, though probably due almost wholly to normal erosion processes,” and points out that “some of the region, mostly remote from the rivers, is so level that the railroads have built straight tangents (i.e., straight tracks) a dozen or more miles in length.’”’ He also points out the rarity of swamps. The region is traversed by rivers that are, in most places, bordered by steep, bare bluffs—in some places 60 feet = Tuomey’s Second Biennial Report, pp. 134-37, 1848, quoted by Eugene A. Smith in his report on the Geology of the Coastal Plain of Alabama (1894), pp. 282-84. 2 Underground Water Resources of Alabama (1907), p. 13. 3 Roland M. Harper, ‘‘Economic Botany of Alabama,” Geographical Report on Forests, Monograph 8, Part 1, 1913. 704 HERDMAN F. CLELAND high—of chalky limestone, and the tributary streams have all the characteristics of youth. The sides of the Black Belt trough are bounded on the north and south by ridges, formed of the more resistant strata of the Coastal Plain, which rise 200 to. 300 feet above the general level of the surface. The pronounced cuesta which forms the southern border of the trough is composed of the sandy, more resistant Ripley (Cretaceous) sediments. The Black Belt, Black Prairie, Cotton Belt, or Cane Brake, as it has been variously called, can be briefly described as a belt of rich, black soil with an average width of 20 to 25 miles, and an area in Alabama of about 4,300 square miles. It extends in an east-west direction in south central Alabama and conforms exactly with an easily decomposed, impure, chalky limestone of rather uniform composition (Selma chalk) which has a thickness of about 1,000 feet in the western part of the state and thins out and disappears in the east near Montgomery. This formation dips to the south at the rate of 30 to 4o feet to the mile while the surface slopes at a much less rapid rate in the same direction. It is the weathering of the beveled edges of this limestone that determines the width and position of the Black Belt. The soil formed from this rock is a clay of exceptional fertility but some- what difficult to cultivate because it bakes in summer and becomes tenacious mud in winter. After the deposition of the Coastal Plain sediments a deposit of red sandy loam, called the Lafayette formation, was laid down on them, either during the early Pleistocene or near the close of the Pliocene, and formed a veritable mantle covering many hun- dreds of square miles. The depth of this formation is, in places, as much as fifty feet, but little of it has a thickness of more than 25 feet. The origin of the Lafayette has given rise to much dis- cussion,’ but as the underlying formations in Alabama contain little quartz from which pebbles could be made, the abundant water- worn quartz pebbles show that in this state, at least, it must have been transported long distances. On the sides of the Black Belt trough some knobs and ridges are capped by this deposit, proving W. W. Shaw, U.S. Prof. Paper 108 H. A PLEISTOCENE PENEPLAIN IN THE COASTAL PLAIN 705 that the Black Belt was once covered withit. The almost complete - absence of the Lafayette over the area underlain by the Selma chalk and its presence on other parts of the Coastal Plain north and south is attributable to the greater ease with which the chalk is weathered and eroded. Because of its solubility and lack of strength, the streams that flow through the limestone quickly cut their beds to grade. In other parts of the Coastal Plain which are underlain by limestone, it is also found that very little remains of the once widespread cover of Lafayette. The features which lead to the belief that the Black Belt of Alabama is in the youthful stage of a first cycle of erosion was based upon the facts (1) that its surface is so level in certain areas as to give it an appearance of topographic youth; (2) that the rivers are bordered by steep banks or bluffs and are in a youthful stage of an erosion cycle. The evidences which indicate that the region was peneplained and has been elevated in comparatively recent times are: (1) that it occupies a troughlike depression 200 to 300 feet lower than the bordering lands to the north and south; (2) that, although the soil is a clay, and is consequently very favorable for the retention of water, swamps are nevertheless uncommon except in river bottoms, showing that the drainage had been thoroughly established; (3) that the Lafayette, which once covered the Black Belt, has been almost entirely removed from it; (4) that the thick, residual soils of the region were probably formed chiefly after the land was reduced to a peneplain (at the present time they are being rapidly eroded away); (5) that the present youthful appearance of the region is due to a comparatively recent elevation of the peneplain 60 or more feet, which permitted the rivers to sink their beds; (6) that the peneplanation must have taken place during the Pleistocene, as is shown by the fact that the region was reduced to a nearly level surface and that a thick residual soil was formed after the removal of the Lafayette, a formation that was deposited not earlier than late Pliocene and, more prob- ably, during the Pleistocene. Estimates of the length of geological time are so uncertain that little dependence can be placed on them, but it is, nevertheless, 706 HERDMAN F. CLELAND interesting to speculate upon the time required for the removal of the Lafayette loams, sands, and gravels from the Black Belt and . for the reduction of the surface during part of the Pleistocene. Penck’s estimate of 500,000 to 1,000,000 years for the duration of the Pleistocene, based upon the rate of advance and retreat of the Pleistocene Ice Sheets, is to be contrasted with Barrell’s' minimum estimate of 1,500,000 years based upon a study of radioactivity. A few years ago Barrell’s estimate would have seemed extravagant, but when one considers that a region, such as the one under discussion, has been denuded of a thick deposit of gravel and loam, has been reduced to a peneplain, has been weathered so long as to form a thick residual soil, has been raised, and, finally, has been so dissected by streams as to make a topog- raphy of youthful aspect, the larger estimate does not seem impossible. In 1906, Chamberlin and Salisbury? presented figures as to. the duration of time since the Kansan glacial epoch, giving 300,000 as a likely minimum, and 1,020,000 as a likely maximum. Had the statement covered the time since the beginning of the Pleisto- cene, these figures would have been considerably larger. The physiographic history of the Coastal Plain of the Gulf of Mexico has not as yet been carefully worked out, and it is probable that a thorough study will show that this surface instead of being the youthful topography of a first cycle of erosion, is, for the most part, the incised surface of a peneplain or a plain of marine abra- sion, in which are subordinate peneplains such as that of the Black Belt. The unconsolidated sediments and broad intervales give the impression of youth but the beveled edges of the forma- tions which underlie the Coastal Plain and the level, outstanding cuesta ridges are suggestive of peneplanation. The writer hopes to be able to make a further study of the physiographic history of our Gulf Coastal Plain and with it a study of the Atlantic Coastal Plain. t J. Barrell, ‘Measurements of Geological Time,’ Geological Society of America Bulletin, Vol. XXVIII (1917), p. 802. 2 Earth History, Vol. II, p..420. THE MECHANICAL INTERPRETATION OF JOINTS? WALTER H. BUCHER University of Cincinnati PART I OUTLINE THE Joints ON MINE ForK HARTMANN’S LAW HARTMANN’S LAW APPLIED TO EXPERIMENTAL AND FIELD OBSERVATIONS Compressive stress vertical, tensile stress horizontal Experimental observations The joints on Mine Fork, Magoffin County, Kentucky. Both, compressive and tensile stresses horizontal Daubrée’s experiments on Torsion The joints on Crooked Creek, Adams County, Ohio The joints of Lake Cayuga, New York The joints of the Wisconsin shore of Lake Superior The areal study of joint systems Greatest compression horizontal, least compression vertical Small “symmetrical faults” of Lake Cayuga, New York Experimental observations I. THE JOINTS OF MINE FORK In the course of field work in eastern Kentucky, in 1917, the writer observed a case of local jointing which gave him the clue to the following investigation. On Mine Fork, a few hundred yards above the mouth of Lacy Creek, in Magoffin County, close to the Morgan and Johnson county lines, in a nearly vertical cliff of a strongly cross-bedded, coarse-grained sandstone forming the top of the Lee Group of the Pottsville series, the system of intersecting joints shown in the accompanying sketch (Fig. 1) is exposed along the roadside. Unfortunately the commercial work in which the writer was engaged at that time did not permit him to spend any more time in that vicinity than was necessary for a hasty survey of this exposure. * Part I of this paper was presented, in essence, at the last meeting of Section “KE” of the American Association for Advancement of Science, at St. Louis, December, 1919. 797 708 WALTER H. BUCHER It was found that (a) The jointing is confined to the upper third (or even less) of the massive sandstone which is here about roo feet thick. It is entirely lacking below. (b) It marks the crest of a minor anticline on the downthrow side of a conspicuous fault. (c) The average hade of the joint-traces on the practically vertical exposure which trends about ina NNE-SSW direction, is: set I: 27°—NNE; set II: 35°—SSW; inclosed angle: 62°. (d) The average strike of the joint-traces, measured on the horizontal surface of projecting ledges, is: set I: N 78 W; set II: N 27 W; inclosed angle: 51°. (e) The joints of set I are much better developed, Fic. 1.—Jointing in vertical cliff of massive, cross-bedded sandstone on Mine Fork, Magoffin County, Kentucky. longer, more continuous and more regular in their course than those of set II, both in the vertical and in the horizontal planes. For two reasons the occurrence of this system of joints at this locality seemed surprising. There could be little doubt that these joints represented planes of shearing. The writer had, however, always associated the fracturing of hard materials by shearing with compressive stresses or, at best, with compound stresses resulting in torsion. But here he was dealing with a clear case of simple tension along the crest of an anticline, causing a hard sandstone to fail along typical planes of shearing. He had also been accustomed to ascribe to the planes of maxi- mum shear a general tendency to intersect at right angles. No THE MECHANICAL INTERPRETATION OF JOINTS 709 such tendency could be inferred from these joints, which intersect uniformly at an angle close to 60°, with the obtuse angle facing in the direction of the tensile stress. II. HARTMANN’S LAW Following the clue given by these observations, the author became acquainted with a book published in 1896 in Paris by L. Hartmann under the title Distribution des déformations dans les métaux soumts a des efforts,. containing a wealth of experimental data and a fascinating discussion of the lines forming on the sur- faces of metals when strained beyond the elastic limit, known as Liiders’ lines.” When a highly polished plate of metal is subjected to a very gradually increasing simple tensile stress, the first permanent deformation is accompanied by the sudden appearance of one or several delicate straight lines or bands cutting in an oblique direc- tion across its surface. Suitable illumination shows them to be depressions. When the stress is further increased, the existing lines widen and new ones appear, forming two conjugate systems of oblique lines, symmetrical to the direction of maximum stress and intersecting at a constant angle which in most metals (and rocks) is greater than go0°3 This angle remains unchanged with growing tension and is thus independent of the intensity of the stress. The final rupturing may entirely or partly follow these lines or cut across them at right angles to the tension. Under compression, similar systems of lines form, but now the angle of intersection bisected by the direction of the compressive stress, for most rocks and metals, is smaller than go°, and for the same material is the supplement of the one obtained under tension. t Berger-Levrault, Paris, 1896. 2 Called after Liiders of Magdeburg who first described them fully in 1860. “Uber die Ausserung der Elastizitat an stahlartigen Eisenstében und iiber eine beim Biegen solcher Stabe beobachtete Molekularbewegung,”’ Dingler’s Polytech. Jour., Vol. CLV (1860), p. 18 (not seen). 3Ten good illustrations of strips of low steel showing yield lines developed under tensile stress, are given in H. Marten’s Handbook of Testing Materials (translated by Gus. C. Henning), John Wiley & Sons, N.Y., 1899, Vol. I, Pl. 1, Figs. 3, 5, 12, 14-20. 710 WALTER H. BUCHER The lines in this case are depressions only on one side, with the corresponding lines on the reverse side forming delicate ridges. Final rupturing, under compression, always follows these lines. The great importance of these lines of Liiders for our purposes lies in the fact that they represent the outcrops of internal planes of yielding, differing largely in scale and degree of deformation, not in origin, from the planes of shearing observed on a large scale in nature. In fact, in the small test piece as on a gigantic scale in nature, we see that the stress acts not uniformly on every unit of the mass undergoing deformation, but that it reaches a maximum along these geometrically distributed surfaces, while maintaining lower values in the volume between. We seem to be dealing here with a sharply defined application of the principle of least work.* At every point along every imaginary line of stress within a body undergoing elastic deformation there exists the tendency to shear in any one of an infinite number of directions all inclined to the direction of stress at the same angle, the sum of which forms two infinitely small cones joined by their apices.” Out of the infinite number of surfaces which may be obtained by connect- ing any two of such adjoining possible directions of shearing, those only will form which involve the expenditure of a minimum of energy. When the lines of stress are not parallel, owing to the unequal distribution of stresses, the resulting surfaces of yielding may be very complex and the pattern of lines formed by their traces on the surface may be far from regular (Fig. 2A—C*). In such cases Liiders’ lines can be used to reconstruct the lines of maximum stress on any given test piece, by drawing the lines bisecting the angle of shear at every point of intersection of the shearing planes. Figure 2C represents the lines of stress derived in that way from Liiders’ lines as obtained in the experiment illustrated in Figure 2A and B. 1H. von Helmholtz, “Uber die physikalische Bedeutung des Princips der klein- sten Wirkung,” Wissenschaftliche Abhandlungen, Vol. III, pp. 209-10. 21. Hartmann, op. cit., pp. 18-19. 3 Hartman, Joc. cit., Figs. 48-50. THE MECHANICAL INTERPRETATION OF JOINTS 711 Ss SR “LAN | Fic. 2—A. Arrangement used in one of Hartmann’s experiments in which the test piece (in the center) was subjected to uniform compression over its whole base, while the upper surface suffered compression in the center only. (L. Hartmann, 1896.) B. Liiders’ lines produced in the experiment illustrated in Fig. 2A. (L-. Hartmann, 1896.) C. The theoretical lines of stress (bisecting the angles of Liiders’ lines) reconstructed on the test piece illustrated in Fig. 2B. (L. Hartmann, 1896.) as Sek 712 WALTER H. BUCHER The most irregular pattern of Liiders’ lines results when the lines of stress are not parallel to the axis of the test piece, but inter- sect with it at varying angles. In that case, the angle formed by planes of yielding may be cut by the surface in all possible directions and the apparent angle of intersection of Liiders’ lines as seen on the surface varies from point to point and must not be mistaken for the true constant angle bisected by the line of maximum stress of which it is only the oblique outcrop. In 1900, O. Mohr published a mathematical study which led him to views practically identical with those of Hartmann. They may be summarized as follows" a) In all hard materials (except the most brittle ones), under tensional as well as compressional stresses, deformation by shearing takes place in two systems of intersecting planes of shearing. b) Adjoining planes of one system are parallel. c) The angle at which the two systems intersect is constant for any given material, that is, it is independent of the nature or intensity of the stresses involved. d) For the same kind of material, this angle differs the more from go° the harder and the more brittle the material is (e.g., hard or soft steel). e) If we consider tension as negative compression, the law governing the arrangement of the yield planes with reference to the principal axes of stress which will be referred to as Hartmann’s Law, can be expressed as follows: In brittle materials, the acute angle formed by the shearing planes ts bisected by the axis of maximum compression, and the obtuse angle by the axts of minimum compression which is generally negative, representing tension. f{) If the position of the principal axes changes from point to point, the shearing surfaces are warped. ‘The less this is the case, i.e., the more nearly homogeneous a material is, the more regular are the shearing planes. g) The shearing planes do not originate simultaneously, and are not uniformly distributed. t F, Rinne, ‘‘ Vergleichende Untersuchungen iiber die Methoden zur Bestimmung der Druckfestigkeit von Gesteinen,” N. Jahrb. f. Min., etc., Vol. I (1907), p. 45. THE MECHANICAL INTERPRETATION OF JOINTS Tite III. HARTMANN’S LAW APPLIED TO EXPERIMENTAL AND FIELD OBSERVATIONS Hartmann’s law enables the geologist as well as the mechanical engineer to reconstruct the position of the principal axes of stress in any given body subject to mechanical deformation—hbe it a test specimen in the laboratory or the exposed portion of a fractured rock-mass—by analyzing from point to point the position of the planes of shearing. The direction bisecting the acute angle formed by the planes of shearing corresponds to that of the greatest principal axis of compressive stress, while the bisectrix of the obtuse angle gives the direction of the least stress, which in most cases represents active tensile stress. The direction of the inter- mediate principal stress coincides with the line of intersection of the two planes of shearing. It is essential, however, to realize at the start the limitations of this law. a) It applies only to brittle substances. b) Not all lines of fracture are lines of shearing. Brittle materials, such as cast iron or hard steel, and most rocks under simple tension habitually fail along planes of fracture at right angles to the direction of maximum tensile stress.’ Soft steel, on the other hand, fails along inclined planes of shearing under tension as well as under compression. c) The position of the planes must be studied in space, not in any accidental plane of exposure. d) The principal stresses inferred from them need not be identical with any real stresses, but may be only the resultants of the combined action of several stresses (‘‘equivalent”’ stresses). _ We may now proceed to test the usefulness of Hartmann’s law by applying it to a few selected experimental data and geo- logical field observations. 1. Compressive stress vertical, tensile stress horizontal—a) When a cylindrical test piece is subjected to compression beyond the elastic limit, Liiders’ lines make their appearance on its surface, forming a characteristic pattern of symmetrical intersecting spiral 1 See, for instance, A. L. Jenkins, ‘Combined Stresses,” Jour. Amer. Soc. Mech. Engineers (1917), p. 696. 714 WALTER H. BUCHER lines, with the acute angles formed by their intersection facing the direction of pressure. On specimens of Carrara marble used by Rinne’ this angle measured 60°, on those used by Karman? it measured 54°, while red sandstone (Buntsandstein) gave a value as low as 38°. When the pressure is increased until rupture occurs, the plane of fracture forms a symmetrical cone with an apical angle equaling the angle of shear characteristic of the material. In this case, the least principal stress equals the intermediate stress. ‘Thereby its position is made indefinite with reference to the infinite number of directions in the plane common to the two lesser stresses, normal to the greatest principal stress. The peculiar conical fracture is the result.3 As soon, however, as any one of the infinite number of possible directions in the plane normal to the greatest stress offers a mini- mum of resistance, rupture occurs‘ along two well-defined planes, as indicated in Figure 3. Daubrée’s classical experiments on blocks made of a mixture of plaster of Paris and beeswax® correspond directly to this case. t F. Rinne, “‘ Vergleichende Untersuchungen iiber die Methoden zur Bestimmung der Druckfestigkeit von Gesteinen,”’ Neues Jahrb. f. Miner., etc., Vol. I (1907), p.45- 2Th. von Karman, “Festigkeitsversuche unter allseitigem Druck,” Zedtschr. d. Vereins deutscher Ingenieure, Vol. LV (1911), pp. 1748-57. 3The remarkable fracturing in the form of parallel and interpenetrating cones observed in the brittle white limestones of the Upper Jurassic along the intensely shattered margin of the crypto-volcanic basin of Steinheim seems to be due to this con- dition. W. Branco u. E. Fraas, “Das kryptovulkanische Becken von Steinheim,” Phys. Abhandl. d. K. Preuss. Akad. d. Wissensch. (Berlin, 1905), pp. 36-38. 4In a cube where four directions offer an identical minimum of resistance, the planes of fracture form a pyramid as may be seen in any ordinary crushing test. 5A. Daubrée, “Etudes synthétiques de géologie expérimentale” (Paris, 1879), pp. 315 ff. and Figs. 93 and 94. Fora copy of Fig. 93 see, e.g., Van Hise, “ Principles of North American Pre-Cambrian Geology,” Sixteenth Ann. Rep. U.S.G.S. (1895), Pt. I, p. 644, Fig. 126. Note in this figure the difference between Liiders’ lines and the final plane of shearing. The former, marked “R,”’ do not, at first, correspond to continuous internal surfaces. They represent purely local effects along individual lines of stress. The establishment of large planes of shearing (marked “‘F”) isa later development. The difference between the two is shown strikingly on the right side of the block, where the main fracture cuts diagonally across Liiders’ lines. This contrast between Liiders’ lines and the final planes of rupture is met with in all experiments. It seems to indicate that at first the greatest tension exists parallel to the surface of the test specimen, due to the stretching of the horizontal dimensions accompanying the vertical shortening. Rupture, on the other hand, gives dominance to the direction of easiest movement in a radial direction. THE MECHANICAL INTERPRETATION OF JOINTS 715 b) It is easy to see that the joints on Mine Fork, Kentucky, described in the introduction to this paper, correspond to this type. Here, however, the active stress was the horizontal tension existing at the top of the anticline, while the weight of the overlying rock-masses, giving the compressive stress, was merely passive. The analysis of joints can, however, be carried farther and _ may often yield information of decisive value to the field geologist. Fic. 3.—Diagram illustrating the position of the planes of shearing in a brittle body subjected simultaneously to vertical compression and horitonzal tension. A detailed analysis of the joints on Mine Fork will be given here to illustrate the method of analysis used by the writer. The exposure on Mine Fork is such as not to give the true dip of either of the joint planes. The joints themselves are filled with mineral matter and their surface is nowhere exposed. But their apparent hade was measured on the vertical face of the exposure trending essentially NNE—SSW and their strike was determined on the level top of the cliff. | Set I Set II Apparent hade: 27° northward 35° southward Strike: N 78 W N 27 W 716 WALTER H. BUCHER A complete analysis from these data involves the following steps. Find 1. The actual position of the two planes in space. 2. The direction, in space, of the line of intersection of the two planes which corresponds to the position of the intermediate princi- pal stress. 3. The position of the plane normal to this line. 4. The location in this plane of the other two principal stresses bisecting the acute and obtuse angles respectively. The stereographic projection is admirably adapted to the demands of problems of this kind. By its use, the position of the principal stresses in space can be obtained in the field from any given set of joints within a few minutes. In the following brief description of the construction of Figure 4, a working knowledge of the stereographic projection is assumed.* 1. Draw the line Ex-Ex, trending N 23 E, to represent the vertical plane of the exposure. On it, mark the point c, 27° south- ward from O, and c’, 35° northward from O. The planes acb and a’c’b’ represent the two joint planes and can now be drawn. 2. Since the two points O and d are common to both planes, Od is the line of intersection of the two planes, that is, the direction of the intermediate stress. 3. On the great circle acb mark point e, and similarly point e’ on a’c’b’, both 90° from d. Through e and e’ draw the great circle fee'g, representing the plane normal to Od. On it we can read directly the true value of the acute angle of the shearing planes, which in this case is 72°. «For a detailed discussion of the stereographic projection see A. Johannsen, Manual of Petrographic Methods, p. 17. McGraw-Hill Book Co., 1914. For most purposes a protractor giving great circles and vertical small circles 1o° apart, such as is given (after Penfield) in A. F. Rogers, Introduction to the Study of Minerals (McGraw-Hill Book Co., N.Y., 1912, pp. 82-86), is perfectly sufficient. It can readily be copied and carried in the notebook for use in the field. Greater accuracy can, of course, be obtained by the use of Wulff’s net, a large copy of which is con- tained in E. E. Wright, “‘The Methods of Petrographic-Microscopic Research,” Carnegie Inst. Pub. 158, Pl. IIT. The reader who has had little practice in the use of the stereographic projection will find it easy to visualize Fig. 4 by remembering that the great circles must be imagined to be drawn on the surface of a hemisphere resting on the circle NESW with O at its center. A line such as dO, therefore, represents a radius extending from the surface of the hemisphere, at d, downward to the center O. ——_; —- THE MECHANICAL INTERPRETATION OF JOINTS 717 4. Find the point 2, located halfway between e and e’. The line 2O, lying in a plane normal to the intermediate stress dO, and in the plane /dj bisecting the acute angle of the shearing Fic. 4.—Stereographic projection of the joints on Mine Fork, Kentucky. Ex= Trend of exposure; ab=set I of joints; a’b’=set II of joints; od=line of intersection of joint planes=position of intermediate principal stress; hidj=plane bisecting the acute angle of the joint planes; oi=located in this plane, normal to od=position of the greatest (compressive) principal stress; md/=plane bisecting the obtuse angle of ' the joint kid planes; og=located in the plane md/, normal to od=least (tensile) prin- cipal stress; mdl, gif=principal planes. planes, represents the direction in space of the compressive stress. The line /7 gives the trend of this stress in a horizontal plane. 5. go° from 7, on the great circle gif, mark the point k, which in this case practically coincides with f. The line kO, lying in 718 WALTER H. BUCHER the plane normal to Od and go° from Oz, represents the direction in space of the tensile stress, and the line ml, in the vertical plane lkdm, gives the horizontal trend of this stress. This analysis leads to the following conclusions: The direction of Ok, of the tensile stress, differs only 3° from the horizontal, as would be expected at the crest of an anticlinal bulge. The pull was slightly inclined downward in the direction N 33 E. The very crest of the anticline, therefore, must be sought on the left side of the exposure, a short distance to the southwest. The differential movement which develops when strata slip past each other in the process of folding was here directed toward the crest and favored the development of the joints of set I which are more numerous, more regular, and stronger than those of set II. ' The direction N 33 E of the greatest tension suggests in a general way the dip. and therewith also the strike, of the strongly cross-bedded sandstone. 2. Both, compressive and tensile stresses horizontal.—a) When Daubrée subjected to torsion narrow strips of glass, measuring about a yard in length, and produced on them the well-known system of intersecting fractures, he gave the science of geology one of its most impressive laboratory experiments and one of its most popular textbook illustrations on the subject of joints. Careful analysis, however, reveals the fact that the conjugate systems of fractures which he produced, do not correspond directly to similar joint systems in nature. Figure 5 is a sketch of the fractures forming two prominent ‘“‘fans” on one of Daubrée’s plates.* The tendency to form such “‘fans” is obvious in all torsion experiments made with glass. Duparc and LeRoyer found that it is the more pronounced, the thicker the glass plate is which is used for the experiment.” « The one in the center of the plate reproduced on Plate XII of Haug, Géologie, Vol. I (Paris, 1911) (opp. p. 228). 2. Duparc and A. LeRoyer, Contributions 4 l’étude expérimentale des diaclases produites par torsion,” Archives des Sciences phys. et nat., 3me sér. XXII (Genéve, 1889), p. 307. Daubrée used plates 7mm. thick; on plates of 2 mm. thickness or less, “‘fans”? may not form at all. THE MECHANICAL INTERPRETATION OF JOINTS 719 Each “‘fan”’ consists of a gently curved ‘“‘master-joint,” marked é and ¢’, from which start, at a very acute angle, a number of minor joints, s and s’, which unmistakably tend to be straight and parallel to each other. The clue to this peculiar fan-structure of fractures we find in Hartmann’s experiments with rectangular strips of soft steel.t Under torsion, two systems of Liiders’ lines appeared on them, each parallel to one of the sides of the test piece, intersecting practically at right angles. This indicates that the direc- tion of greatest tension traverses the surface obliquely, forming an angle near 45° with the axis of torsion.2, When the deformation was carried farther additional lines of defor- mation appeared in the vicinity of the longer edges, bisecting the angles formed by the first set of lines. When a plate is subjected to simple torsion, each element of the upper surface suffers simultaneously tension in one direction and compression at right angles to it. The same is true of the lower surface, but with the directions of tension and compression reversed.3 At any point on either surface, therefore, the position of the shearing planes is sharply defined through the combined action of tension and compression as shown diagrammatically in Figure 6. In case tension fractures are \ Fic. 5.—The fractures forming two characteristic ‘“‘fans”’ on one of the glass plates used in Daubrée’s experiments on fractures produced by torsion. formed in addition, they bisect the acute angle of the shearing planes. This is what happened in Hartmann’s experiment with t Hartmann, loc. cil., p. 175 and Fig. 173. 2 This can be verified readily by drawing a circle on the flat side of a rubber eraser and twisting it. G. F. Becker, ‘The Torsional Theory of Joints,” Trans. Amer. Inst. Mech. Engineers, Vol. XXIV (1895), p. 136. 3 For the purposes of the following discussion it is important to remember that essentially horizontal tensional stresses arise in surfaces made convex, and similar compressive stresses in surfaces made concave through the process of bending. 720 WALTER H. BUCHER strips of soft steel. The second set of fractures, formed after shearing was well under way along Liiders’ lines, consisted of tension fractures, one started from the lower, the other from the upper surface. Glass, on the other hand, being a highly brittle substance, in contrast to soft steel, will fail along tension fractures rather than along planes of shearing. The fractures marked ¢ in Figure 5 are the only ones that form when the glass plate used in the experi- ment is very thin. They must, therefore, be tension cracks, one set produced on the under side, the other, symmetrical to it, on: Fic. 6.—Diagram illustrating the position of the planes of shearing in a brittle body subjected simultaneously to compression and tension, both in a horizontal direction. the upper surface, and both finally extended to both surfaces, owing to the thinness of the plate. The gentle curving of these cracks is quite in harmony with this interpretation. The other set of fractures, marked s in Figure 5, intersects with the tension cracks at angles varying from 15° to 25°. This, however, is one-half of the angle of shearing characteristic of glass.* The same angle for soft steel is approximately 45°. It is evident, therefore, that these fractures represent shearing planes produced by the compressive stress acting in the direction normal to the tensile stress. In the experiments made with glass, however, in contrast to those with mild steel, only one set of the shearing planes forms in connection with a tension crack, that only which t To verify this, it is sufficient to compress small pieces of thick plate glass in a strong vise. The resulting angle of shearing can be measured conveniently by means of Penfield’s contact goniometer. ee a a THE MECHANICAL INTERPRETATION OF JOINTS or les favorable to the differential movement resulting from the process of torsion. Here, again, the fractures produced on the upper surface extend down to the lower surface and vice versa. A Fault Anticlinal bulge fi Gemtibees | Drag ok Fault Set, General 3° ° SEE eastward dip Fic. 7—A. Map sketch, showing relation of stations on Crooked Creek, Adams County, Ohio, at which joints were measured, to the fault and to other structural features. B. Diagram showing the position of the joints observed at the stations 1, 2, 3,4, 5- Sz S2=shearing joints; ‘=tension joints. Since the intensity of the compressive stress increases with the thickness of the plate, it is evident that the fan-shaped groups of cracks will form the more freely the thicker the plate is. The most striking feature of Daubrée’s famous experiment, therefore, namely the formation of two systems of fractures inter- secting approximately at right angles, is an accidental result of the exceptional brittleness of glass and the thinness of the plates W202 WALTER H. BUCHER used, permitting the fractures produced on the upper and lower surfaces respectively to interpenetrate. Most rock materials, on the other hand, are less brittle than glass and therefore more inclined to yield along shearing planes when subjected to torsion. Moreover, at least in the case of the larger joints observed in nature, the thickness of the formations undergoing deformation through torsion is sufficient to keep the fractures formed on the upper and lower surfaces separate. In general, therefore, joints produced by torsion in the course of larger earth movements should occupy the position indicated in Figure 6 with the direction of both, compressive and tensile stress, not differing much from the horizontal. In addition to these, tension fractures, bisecting the angle of the other joints, may occur and even dominate. These, together with an unequal develop- ment of the two sets of shearing joints, with possibly one even missing completely, may give considerable variation to the appear- ance of the same joint system from point to point. We may now turn to a discussion of three selected cases of joint systems. a) In Figure 7A the general structural relations are given for five points along Crooked Creek, Adams County, Ohio, at which the position of joints was determined by the writer. The joints here cut in a nearly vertical position through the rather thin and even beds of the fine-grained dolomite of the Bisher formation. Figure 7B shows the strike of these joints as contained in the following field notes. Station Set sz Set sa Set ¢ Teepe N25 E N 40 W strongly developed sharp and persistent |..............2+-00- Clase orate N 20 E N 70 W (average) well developed strong but variable |....... “oe 2s attain iC oner epee N 20 E N 70 W (average) sharp and regular; closely| few and far apart spaced (2-10 in. apart)| (several feet); irregular.|.................... (ite ee N 22 E N 70 W N 62E sharp and regular; (1-2] very few, only three sharp and regular, 1-2 feet apart) good joints seen feet apart Ree aeieion. N 20 E N65 E sharp and:regular wae sii ledis eecreranacueucrtcvsrercrn dominant system, closely spaced THE MECHANICAL INTERPRETATION OF JOINTS 7723 The very regular systems of joints observed at station 1 obviously owe their existence chiefly to the compressive stress caused by the upward buckling of the strata farther north along the fault. The change of the angle of shearing from 65° at station 1, to go° at stations 2 and 3, probably is due to factors which will be discussed in the second part of this paper. It is brought about by a shifting of the trend of the system s, which at the same time becomes more and more irregular and scattered. At station 4 set s, is only represented by a few widely separated cracks, while a new system of fractures makes its appearance. They are parallel to the fault and become increasingly prominent and closely spaced as the fault is approached. They must be tension cracks. The joints at station 5 offer an exact analogy to the fan-shaped cracks formed on the upper surface of a glass plate under torsion, with shearing planes developed only on one side of a tension crack. The great practical value of such an interpretation of joints, during the progress of field work, is obvious. In this case, for instance, the joints at stations 1 to 3 would lead the field geologist at once to look for an uplift either north or south of these points, or both. The sudden appearance and increasing importance of an additional system such as the joints of set ¢ would suggest the neighborhood of a flexure or a fault not far to the northwest striking N 65 E. Information such as this will certainly pay for the time spent on detailed intelligent observation. b) The remarkable jointing exposed along the shores of Lake Cayuga and Lake Seneca. “resembling the gigantic ruins of Cyclo- pean architecture,’ has been made classic through the series of woodcuts published by Hall in 1843.* Miss Pearl Sheldon? has given us a large number of careful measurements of these joints in the vicinity of Cayuga Lake. Two systems of joints stand out from all others. They are gen- erally stronger, more regular, and remarkably constant in their 1 Geology of New York, Pt. LV (Albany, 1843), pp. 303-6. For good modern illustrations see, e.g., Watkins Glen-Catatonk Folio, No. 169 (1909), Pl. I, Figs. 15 and 16; and Jour. Geol., Vol. XX (1912), p. 78. 2 Pearl Sheldon, ‘‘Some Observations and Experiments on Joint Planes,” Jour. Geol., Vol. XX (1912), pp- 53-79; 164-83. 724 WALTER H. BUCHER trend. They are practically vertical. The strike of the one set ranges from N 70 E to N 80 E with the majority lying between N 75 E to N 78E. The other set has a strike ranging from N 20 W to NioE. Frequently joints of the two extremes, near N 20 W and N 10 E, are present at the same locality. No detailed data are given for the large number of minor joints of this region. Their trend seems to be highly variable, in general and from point to point, and ranges through all points of the compass. They are often curved and irregular, and as a rule small. They generally show a large hade, ranging as high as 60°. The remarkably uniform position of the two major joint systems" stands in strong contrast to the highly variable dip of the limbs of the low anticlines and synclines formed by the rocks of the region, as shown on the geological map.? This contrast is especially striking, as Miss Sheldon points out herself, where the dip and strike of the rocks changes rapidly from point to point along the pitching end of an anticline (for instance, the Shurger Point anticline)’, while the position of the joint planes remains unchanged. It is evident, therefore, that the aacon of these joint sys- tems was independent of the folding and followed it. Here, as on Crooked Creek, one system is quite constant in its trend, while the other varies in such a way as to form angles ranging from about 65° to go° with it. Weare, therefore, justified in the assumption that they represent planes of shearing produced by compression in a NE-SW direction under general conditions of torsion. To test this interpretation, we turn to the geological maps of Watkins Glen and Catatonk quadrangles. West. of Cayuga Lake, along the axis of the Watkins anticline, the contact of the Portage and Chemung formations is nearly level, varying between 1,480 and 1,560 feet above sea-level for a distance of over 18 miles. As it approaches the valley of Cayuga Inlet, it rises above 1,600 feet. East of Ithaca, in the same general t See especially Figs. 6 and 7 of Miss Sheldon’s paper. - 2 See Watkins Glen-Catatonk Folio, No. ae 3 Loc. cit., p. 67. THE MECHANICAL INTERPRETATION OF JOINTS 725 direction, the contact rises rapidly to over 2,100 feet in a similar distance. A closer inspection of the northeast corner of Catatonk Quadrangle reveals the presence of considerable doming in the general direction suggested by the position of the joint planes. This broad anticlinal bulge does not seem to be mentioned in the text of the folio. The fact that the writer’s attention was called to it through the analysis of the joints of Cayuga Lake, serves well to illustrate the practical possibilities of the method employed. The presence of an uplift to the northeast accounts for the existence of a compressive stress in that direction. The horizontal tensile stress implied by the position of the joint planes can be accounted for equally well. Crossing the shores of Cayuga Lake in a southeasterly direction (suggested by the obtuse angle of the joints), we find that, on the crests of the Fir Tree Point, Watkins. and Alpine anticlines, the contact of the Portage and Chemung formations remains essentially at an elevation between 1,600 and 1,700 feet above sea-level. Beyond the Alpine anti- cline, however, in the same southeasterly direction, within a similar distance, the same contact drops to near 1,000 feet in the vicinity of Jenksville in Newark Valley Township. The existence of this depression in the direction suggested by the position of the joint planes, leaves little doubt that this relatively pronounced flexure gave rise to the tensile stress involved in the formation of the joints. c) For a last example we turn to Thwaites’s paper on the “‘Sandstones of the Wisconsin Coast of Lake Superior.’ When we plot the strike of the joints of this region as recorded in the table on page 96, it appears that the peninsula north of Washburn, including the Apostle Islands, in contrast to the regions to the west and south, is traversed by two dominant and persistent systems of major joints. One of the two strikes on the average E-W, the other about 10° east of north. Most probably they represent planes of shearing. The position of the acute angle points to the action of a compressive stress in a NE-SW direction, with a tensile stress acting in a NW-SE direction. 1F, T. Thwaites, ““Sandstones of the Wisconsin Coast of Lake Superior,’’ Wis. Geol. and Nat. Hist. Sur., Bull. 25 (1912). 726 WALTER H. BUCHER The map accompanying Thwaites’s paper shows that the jointed area lies in the continuation of the northeastern end of the great Douglas thrust fault. If this fault had any horizontal component in the northeasterly direction, it could have supplied the com- pressive stress responsible for the position of the joints. Unfortunately, the fault contact of the Middle Keweenawan traps with the underlying much younger Orienta sandstone was found exposed only at four localities. At three of these, the exposures were not even found sufficient to measure the hade of the thrust plane.? In the vicinity of the falls of the Amnicon River, however, the fault-plane proper was found exposed at two separate localities, about 500 feet apart. Here, at both points, two systems of grooves were observed on slickensided surfaces in the immediate vicinity of the fault, on surfaces of conglomeratic beds of sandstone which represent most probably shreds of lower beds dragged up along the thrust-plane.2 One of the sets of grooves ‘‘is parallel to the dip, the other is inclined at an angle of about 30° in a NE-SW direction.’’3 Although the grooves are not part of the fault-planes proper, but occur on what seem to be irregular fragments of sandstone wedged in front of the fault, their constancy on seemingly different planes at points 500 feet apart can hardly be looked on as due to purely local movements. They are more likely the direct result of the last movements along the major thrust-plane and essentially parallel to them. If the joints in the vicinity of the Apostle Islands really owe their origin to the action of this pressure directed upward at an angle of about 30° toward the northeast, we should find evidence of it in the position of the joint planes themselves. According to the table on page 96 of Thwaites’s paper the joints have the tendency to be vertical. From the text we learn, however, in addition that “‘many of the E-W joints are inclined, usually at a steep angle to the north.’’4 tF. T. Thwaites, op. cit., pp. 66, 76, 80, 81. 2 [bid., p. 83. 3 [bid., p. 78. 4 [bid., p. 94. THE MECHANICAL INTERPRETATION OF JOINTS 727 When we plot the attitude of these joints, using the stereographic projection as explained on page to of this paper, we find that the greatest (compressive) principal stress should have been directed upward from the southwest at an angle of 15° if we assume the E-W joints to dip 70° north, and of 30°, if they dip 50° north. This correspondence is quite striking and leaves little doubt as to the correctness of this interpretation. The northwest-southeast tension, indicated by the position of the obtuse angle of the joints in the Apostle Islands, is parallel to the gentle dip of the rocks of the Bayfield Group. Both dip and tension probably resulted from settling along the axis of the Lake Superior syncline simultaneous with the last movements along the thrust-plane. ‘d) The last two examples serve well to show how important it is that each observation of jointing be studied in its geographical -and structural relations to all others. To assemble the joints observed over a large area in a single diagram means to veil their true relationships. A diagram, published by Hobbs in 1905," shows the strike of 1,004 joints measured by Mr. C. G. Brown in the vicinity of Cayuga and Seneca lakes, New York. It clearly indicates the existence of two nearly orthogonal double systems of conjugate joints. A comparison with Miss Sheldon’s diagram? shows that the earlier diagram represents a composite picture of the joint systems of two different localities, since the vicinity of Cayuga Lake exhibits only one of the two double systems, as described above. A fine example of carefully recorded data giving definite measurements of strike and dip, and accurate geographical location of every joint measured, may be seen, for instance, in R. S. Tarr’s field observations embodied in Shaler’s “‘Geology of Cape Ann.’ 3. Greatest compression horizontal, least compression vertical.— The position of the principal stresses and of the resulting planes tW. H. Hobbs, “Examples of Joint-controlled Drainage from Wisconsin and New York,” Jour. Geol., Vol. XIII (1905), p. 370. 2 Loc. cit., p. 66. 3N. S. Shaler, “The Geology of Cape Ann, Massachusetts,” Ninth Ann. Rep., U.S.G.S. (1889), pp. 597-602 and Pls. 72-74. 728 WALTER H. BUCHER of shearing for this grouping of stresses is represented diagram- matically in Figure 8; Figure 9 illustrates the occurrence of this type in nature on a small scale." It shows “symmetrical faults” in a hard encrinal layer a foot or two in thickness in the Hamilton shales, exposed along the shores of Cayuga Lake. ‘““The exposures of this layer along the lake show faults every few feet.” “The strike of the majority is from 20-25° north of west.” “Their inclination is sometimes south and sometimes north and the angles are nearly Fic. 8.—Diagram illustrating the position of the planes of shearing in a brittle body subjected to compression in a horizontal direction with the direction of easiest relief (least principal stress) vertical. the same in the two cases, making the faults symmetrical about a nearly horizontal plane.” ‘The hade varies from 45° to 75°, but most are near the average, which is 62°.” The fault surfaces are slickensided and covered with strong, even striations. “The vertical displacement along these faults is from a fraction of an inch to three inches.” The faults “usually continue for a few feet in the adjacent shale, but instead of continuing with the same bade, they flatten out and become nearly horizontal in the shales where no hard layer is present.” ™ Pearl Sheldon, ‘‘Some Observations and Experiments on Joint Planes,” Jour. Geol., Vol. XX (1912), Fig. 3, p. 61. 2 Ibid., pp. 60-62. THE MECHANICAL INTERPRETATION OF JOINTS 729 Fic. 9.—Planes of shearing due to stress relations similar to those illustrated in Fig: 9. Bed of hard encrinal limestone in Hamilton shales, exposed along shore of Cayuga Lake. (P. Sheldon, 1912.) ofl WALTER H. BUCHER Horizontal faults ‘“‘occur by the score in the shale beds,” usually offsetting the vertical joints for distances often measuring several inches. These “‘faults’’ are unquestionably shearing planes forming an acute angle of approximately 60° facing the direction of the hori- zontal compressive stress which is clearly manifested in the differential movement between adjoining layers of shale referred to above as ‘‘horizontal faults.” Identical results have been obtained in the laboratory, when- ever sufficiently brittle materials were subjected to horizontal — compression in the course of experiments on overthrusting,? and before our eye loom up the sections of the mountain ranges which these experiments were to help explain, with thrust faults in astonishing numbers and on gigantic scales. Here the subject of our discussion assumes different proportions. We realize that the grouping of stresses resulting in the formation of any of the fracture systems discussed before remains the same whether the fracturing finally results in the formation of vast lines of displacement generally referred to as major faults, which may bound mountain ranges or even continents, or ends with the production of minute cracks which, filled with white calcite, form a delicate network on the dark rock and, found on the surface of flat pebbles on the wet beach, are the delight of children. Before we can extend the application of Hartmann’s law to the larger scale of the great overthrusts of folded mountains, we must first answer a question which now assumes fundamental importance: Is the angle of shear, in a given substance, sufficiently constant under widely different conditions of pressure and tem- perature so as to exclude the possibility of grave errors in its use ? We may approach this question best by turning to the ingenious mathematical theory which Mohr has given to account for the results of Hartmann’s and others’ experiments. t See, for instance, H. M. Cadell, ‘Experimental Researches in Mountain Build- ing,” Trans. Roy. Soc. Edinburgh, Vol. XXXV (1890), pp. 337-57; and R. T. Cham- berlin and W. Z. Miller, “Low-Angle Faulting,” Jour. Geol., Vol. XXVI-(4918), Pp. 1-44, especially Fig. 9. Note, however, that the position of the strain ellipsoids in the rigid layers in Fig. 10 does not correspond to the writer’s interpretation. [To be continued] —_ : GEOLOGIC RECONNAISSANCE OF THE SOUTHERN PART OF THE TAOS RANGE, NEW MEXICO JOHN W. GRUNER University of Minnesota INTRODUCTION The Taos Range is a part of the Sangre de Cristo Range. Not more than about two-fifths of the Sangre de Cristo Range extends into New Mexico. The larger portion, the Culebra Range and the Sierra Blanca, lies in southern Colorado. Where the Culebra Range crosses the boundary line into New Mexico it: splits into two great uplifts, the Taos Range and the Cimarron Range, which find their continuation in the Mora Range and Las Vegas Range respectively. The Taos Range proper is about 30 to 35 miles long and has an average width of 15 miles. Its northern limit is Costilla. Creek, its southern Ferdinand Creek. The region described in this report is situated in north central New Mexico near the Colorado line. (See Fig. 1.) The southern part of the Taos Range, as seen from the Rio Grande Valley to the west of it, offers an imposing view. From an altitude of about 7,000 feet, the elevation of the valley above sea-level, a number of peaks rise to snowy heights within a distance of a few miles. The mountain front is deeply incised by Pueblo Creek, Lucero Creek, and Rio Hondo, which flow westward to, the Rio Grande. (See topographic map, Plate XIII.)' The Red River (called ‘‘Colorado Creek’? by Stevenson’), of which only a few miles are in the area mapped, has a northerly direction before it turns westward and breaks through the main chain of the ~ mountains below Red River City. «The attached topographic map was made by the writer who used as basis a. map compiled by the United States Land Office and the United States Forest Service. The peak which is called Taos Peak by Stevenson is generally referred to as Wheeler’ Peak now. 2 John J. Stevenson, “Report upon Geological Examinations in Southern Colorado and Northern New Mexico, 1878-1879,” Report U.S. Geog. Sur. West of the tooth: Meridian, Vol. 3, Supplement, 1881. 731 JournaL or GEoLocy, VoL. XXVIII, No. 8 Pirate XIII 4 FAN RISE RECONNAISSANCE MAP OF RIFE A PART OF TAOS RANGE : a NEW IMEXICO. SORA CRIME t WiLEF Contaur Interval 500 Ft. Carborilireus Gadimenta. Trait Pre-Cambrian % Prospects Guarti/tan. rian Fehiats, aaa Granites. mea One 48 ZONE p <8 CLS etteaives| BY Whe VW & Basse O/hea. A 44 ‘ ¥) g ‘ = SNVW a Nos N= aii _ : J TA ASS SS SS eS N= —— ws ES 7 26. 732 JOHN W. GRUNER On the east of the Taos Range the broad Moreno Valley separates the Taos Range from the Cimarron Range, which is nearly as high. On this eastern slope of the range watercourses are rather scarce. 106° los? 37° Boundar Tne Colorado 37? eS Gastilla | ae Boundary f/Line New \Mextco ranco 7 aa ae N e Penasco M ORA oss. Llano | é site Selae a) ag Oman Pe va e 362 2 4 356° 106° | Scale 1£:1,000,000 Fic. 1.—North central New Mexico showing location of area mapped. Copied from Plate I of Professional Paper No. 68. DESCRIPTIVE GEOLOGY The Taos Range is built up of three great rock systems. The pre-Cambrian gneisses, schists, and granites constitute the basement and greater part of the core of the uplift. Upper Carboniferous strata of great thickness are turned up on the east and south sides of the range. These older systems are intruded by stocks and dikes of granite and rhyolite porphyry respectively. Along the western slope of the mountains large alluvial fans spread over the plains toward the Rio Grande for a number of miles, GEOLOGIC RECONNAISSANCE OF TAOS RANGE 733 where they are finally encroached upon by the basalt flows of the Rio Grande Valley. In Stevenson’s report" the mountains north and west of Pueblo Creek are assigned to the ‘‘Taos axis,” those south and east of that creek to the ‘‘ Mora axis.” It was believed that the Taos axis does not continue beyond Pueblo Creek, but that a new one, the Mora axis, begins at the head of the Red River and runs southward, parallel to the Taos axis on the west, until the latter vanishes. No such structural division could be noticed by the writer between the Taos Range and Mora uplift south of it. The misconception was probably due to the belief that the Pennsylvanian strata exist also on the west side of the range. The sedimentary outliers on the main range, to be described later, when viewed from the distance, easily give such an impression. No new axis begins in this district, but the Taos axis pitches steeply toward the south and the pre-Cambrian rocks disappear at the junction of Pueblo Creek and Indian Creek beneath the Pennsylvanian strata, which form here an uninterrupted anticline across the range. In the region mapped this anticlinal structure is absent. No Pennsylvanian sediments were found on the western slope north of Pueblo Creek. The mountains present a bold fault scarp facing west. Whether sedimentary rocks of Pennsylvanian age underlie the thick débris fans and basalt flows or not is unknown at the present time. But farther north, in Colorado, Siebenthal mentions their occurrence on the west side of Culebra Peak and the anticlinal structure of the Sangre de Cristo Range at that latitude. PRE-CAMBRIAN CRYSTALLINE ROCKS Ancient gneisses and schists—The most ancient rocks are amphibolite and chlorite schists and gneisses that grade into green- stone in places. They cover the larger portion of the northwestern half of the area mapped, form an almost continuous outcrop along the western scarp of the range, and cap all of the high peaks with the exception of Old Mike. Lack of space will not permit to t Op. cit., pp. 41-42. 2 Op. cit., p. 42. 3 C. E. Siebenthal, ‘“‘Geology and Water Resources of the San Luis Valley, Colo.,”’ U.S. Geol. Survey Water Supply Paper 240 (1907), p. 34. 734 JOHN W. GRUNER describe these occurrences in detail, but as a rule sheeting in these rocks has a steep westerly to northwesterly dip. Nowhere in the ancient rocks were any close folds or signs of distortion and twisting seen, as might be expected in schists, and as were actually observed in the metamorphosed sedimentary rocks described below. Granitic gneisses occupy a rather obscure position with respect to the more basic varieties, and may, in some cases, be of the same age as the batholithic granite intruding the older schists. In one instance, on Old Mike, this relationship was proved. Here the gneiss could be traced to its parent rock, the granite beneath. The following outcrop deserves special attention in the opinion of the writer. The plateau south of Lucero Peak, between the Salazar and Lucero Canyon, is covered with a dark-green, fine- grained hornblendite and greenstone which resemble a flow. A ““sheet”’ at least 200 to 300 feet thick, of dark, indistinctly schistose rock, overlies the granitic batholith. Ramifying apophyses from the granite beneath can be seen in the greenstone. Just north of Lew Wallace Peak lies a relatively small mass of greenish-gray, very fine-grained altered diabase. No cleavage or regular sheeting is visible in it. Whether this rock bears any genetic relation to the sheet on the opposite side of Lucero Canyon, just described, or not could not be determined. Metamorphosed sedimentary rocks—The metamorphosed pre- Cambrian sediments occupy belts of greatly varying width between the Pennsylvanian series on the southeast and the granite batholith on the northwest. The assumption that the batholithic granites are younger than the metamorphosed sediments is based upon the attitude of these formations, which flank and abut against the gneisses and granites. (See Fig. 2.) The formations are chiefly composed of quartzites and quartz and chlorite schists, and are undoubtedly of sedimentary origin. The area of quartzite as outlined on the map claims accuracy only along the western margin. The eastern limit could only be estimated; therefore the outcrop may be somewhat narrower. The dip of the quartzite is steeply eastward, varying from 45° to go°. Jointing at right angles to the beds is the rule. Figure 3 shows a nearly perpendicular exposure of quartzite, 300 feet high, ——— Pre-Camb. GEOLOGIC RECONNAISSANCE OF TAOS RANGE 735 northeast of Ben Hur Lake. No trace of the bedding of the original sandstone was detected, perhaps due to the possible identity of the original bedding and the sheeting. The attitude and character of the quartzite strongly suggest this possibility to the writer. The color of the formation as a whole is yellow, but the southern end of it becomes reddish and purplish gray. Southeast of Sacred Lake the very steeply dipping quartzite overlies a very much folded and twisted, thinly laminated chlorite schist. The foliation of the latter, as a whole, is parallel to that of quartzite. What appears to be a continuation of the quartzite x : : tr ~ ay ¢ Ye v 8 Xs x ° XO P 4 Qu oq M4 8 i.) N ie) cs iS SS —--—- — — x < 3 = $ 8 ~1 /e N 9 ACU er a FESO Bie : A Som A ‘i if Wiii{{\ Ancient Schists and Gneisses \Z\}] Intrusive Granite, Pre-Camb. RE Chlorite Schist = Pennsylvanian Sediments ANNAN Pueblo Quartzite 7 al Rhyolite Dike, Post-Paleoz. ° 3 I 13 2 I i ! i | mi. Horizontal and vertical scale Fic. 2.—Cross-section from Salazar Canyon to Pueblo Creek, along line A-B on map. and schist is seen half a mile southwest of this locality, just below Larkspur Point. Here steeply tilted chlorite epidote schist forms a cliff of conspicuous gray color. Strike and dip of the sheeting are very similar to that of the quartzite. Southwest of Pueblo Peak, adjacent to and north of the Pennsylvanian sediments, steeply inclined quartz-schist forms sharp craggy outcrops and cliffs. The formation flanks Pueblo Peak parallel to the fault line for an unknown distance toward the west. Intrusive granites—The distribution and composition of the granites suggest a close genetic connection between the individual areas. It is highly probable that all belong to one great batholith 736 | JOHN W. GRUNER which arched up the overlying formations and metamorphosed the sediments. One of the largest exposures of this batholith is along the Rio Hondo, where a light-gray, very coarse biotite granite outcrops. It weathers easily and forms curiously shaped pinnacles in some places. Also along Lucero Creek a pinkish, more or less gneissoid, granite is found. The latter is also the predominating rock in the Salazar Canyon, and from here a broad belt of it passes beneath Lucero Peak to Old Mike. The rocks from Old Mike and Red Dome show every gradation from a typical red granitic gneiss to a medium-grained biotite granite. Medium-grained biotite granite vary- ing in color from pink to greenish gray also covers a large part of Red River Canyon, especially on the west side. A detailed examination was impossible in this part of the district. Half a mile west of Larkspur Point pink medium to coarse-grained granite outcrops on the steep slope above Fic. 3.—Pueblo quartzite on : Ben Hur Lake. Looking northe Indian Creek. South of Larkspur east. Unconformity on upper left. Point the continuation of this outcrop Pennsylvanian beds nearly at right P i is found in contact with remnants of angles to sheeting of quartzite. ancient schist that caps a part of Larkspur Point. Farther toward the southeast the gneissoid granite is exposed along the northwestern tributary of Pueblo Creek for a distance of two and a half to three miles. Basic dikes —A number of basic dikes are intrusive into the granite of the batholith. Their age is probably pre-Cambrian. The most prominent one occurs just east of the highest point of Pueblo Peak and has a width of 100 to 150 feet. Its trend cor- responds to that of the others, which have a northwest to southeast direction. In composition and texture it approaches a gabbro. GEOLOGIC RECONNAISSANCE OF TAOS RANGE 737 Two pre-Cambrian inliers that cannot be classified outcrop on Pueblo Creek. Quartz and mica schists and some amphibolite are the only rocks in these inliers seen by the writer. CARBONIFEROUS SEDIMENTARY ROCKS In the district mapped all sedimentary rocks belong to the Pennsylvanian series. Stevenson’ described them as Carbonif- erous, attempting no further divisions then. Later writers, espe- cially W. T. Lee,? who examined parts of this series farther east and north, recognized them as belonging to the Pennsylvanian series only. A number of fossils, collected by the present writer near the base of the sedimentary series, belong to the Penn- sylvanian fauna. Six species were identified: Lopophyllum profundum, Siminula subtilita, Spirifer cameratus, Spirifer rocky- monianus, Productus cora, Productus semireticulatus. No generalizations concerning the thicknesses and divisions of the series can be given in this paper with the exception of the state- ment that by far the largest portion of the beds is composed of clastic material. The lowest member is usually a basal con- glomerate grading into a sandstone, but in some localities limestone overlies directly the pre-Cambrian, and the sequence is reversed. It is very common to see a limestone in sharp contact with a coarse clastic in some places. On the other hand, formations hundreds of feet thick are found in which the transitions from one member into another are very gradual. Lithologically very similar beds of clastic material occur at many different horizons of the series and some of them are of such thickness that even considerable displacements by faults may be easily overlooked. The fact that all the faults seen by the writer in the sediments are normal and that evidence of folding due to lateral compression is absent seems to indicate that the Taos Range, if not the whole Sangre de Cristo Range, was formed solely by intrusive activity, probably in early Tertiary time. 1 Op. cit. 2W. T. Lee, “Geology and Paleontology of Raton Mesa and Other Regions in Colorado and New Mexico,” Prof. Paper ror (1918), pp. 41-42. 738 JOHN W. GRUNER Distribution of the sediments.—The sedimentary rocks cover portions of the Pueblo Creek and Red River drainage basins, and extend far beyond the southern and eastern margins of the area mapped. While the thickness of the formations may be estimated at several thousand feet, at least 2,500 feet in the southeast corner of the district, erosion has reduced the thickness of the sediments toward the northwest to less than 300 feet above Sacred Lake. The contact of the sediments with the pre-Cambrian rocks follows approximately a line from the southwest corner of the quadrangle to a point on Starvation Creek, about 1 mile south of Pueblo Peak. A normal fault of unknown displacement, probably relatively small, has sharply upturned the sandstones and limestones against quartz- chlorite schists west of Starvation Creek. Here at the bottom of the creek 1oo+ feet of dense gray non-fossiliferous limestone are seen the base of which is not exposed. Onit rests a dense, brownish- gray, arkosic sandstone that is brownish red when weathered. ‘This rock caps most of the ridges and is of great thickness. Farther east, on the high divide between Pueblo Creek and Lucero Creek, the foregoing limestone is overlain by a greenish- gray calcareous and arkosic grit. This grit is of wide extent and great thickness, not only in this district, but beyond its limits. “At certain horizons this rock is replete with fossil fragments, especially crinoid stems. It becomes gradually coarser toward the top of the formation and changes to a conglomerate which contains subangular pebbles of quartz, gneiss, granite, and schist. They do not exceed a diameter of 1 inch, but attain greater dimensions on Burned Ridge. About 13 miles south of Larkspur Point, where the strata rest on pink granite, the beds dip steeply soutwestward. The basal conglomerate clearly derived from the pre-Cambrian rocks beneath grades into sandstones, grits, and conglomerates. ‘They are of great thickness and constitute no definite, sharply separated mem- bers. The attitude of the beds in connection with the dip of the same formation in the opposite direction on the southwest slope of Burned Ridge, near a point where a fault crosses Meyer’s Creek, indicates a synclinal structure whose axis runs at right angles to Burned Ridge and pitches southeast. GEOLOGIC RECONNAISSANCE OF TAOS RANGE 739 At the head of Pueblo Creek the Pennsylvanian rests on pre-Cambrian quartzite. A north-south fault crosses the divide between Red River and Pueblo Creek, bringing the sandstones and grits into juxtaposition to the granite on the west. A little farther north two minor step faults cross the ridge from northwest to southeast. North of this exposure the sedimentary rocks are confined almost entirely to the area east of Red River. The river has cut a deep canyon into the sedimentaries, exposing on the east slope a thickness of nearly 1,000 feet of the Pennsylvanian. About 200 feet above the creek bottom the following approximate section is exposed: Feet Red sandstone, very fine grained............... IOO-150 Arkosic, light-colored sandstone, at least......... 200 Gray, fossiliferous limestone . sie era 50+ Dark-gray shale, carbonaceous i in 1 places. sae, 50+ Light-gray quartz conglomerate. . Sara eee 40+ CGoncerledthbasen ee sek ek ce ee en eee A number of outliers of the Pennsylvanian are situated on the east slope of the main range and extend from Bull of the Woods to the head of Indian Creek. The most interesting outlier, a block faulted down on at least three of its four sides, lies east of Larkspur Point. (See Fig. 2.) A partial section of it measured just south of Sacred Lake is given below: Feet 15. Light-gray, very hard and massive arkosic sandstone... 15 meEbrowmsh-eray, Shaly limestone... .. 2... :2o: was - 30 13. Lime cemented conglomerate (description below).... 20 Tze uddinestone conglomerates. 2). 2.50-.--.5+-ee<- ie) EE Daricredauvery dense shale. s19", Uiec. ec. 22 <2 2 oles 20 ROM Light-ray ALKOSIC SANGStONE™ 4.2 — s/o Meee ce jas =: 12 go. Puddingstone conglomerates. '. 2+005 2.2... y.ccae. 2 25 8. Gray, massive, argillaceous limestone. ERT ae SOO 45 7. Light-brown, medium-grained calcareous grit........ 20 6. (eraicheercen Shale; non-laminated... a2... 0-6-6 r- 6 5. Puddingstone conglomerate (very coarse).......... 25 4. Brownish-gray, gritty limestone, fossiliferous....... 21 3. Red shale, somewhat arenaceous.................. 9 2. Puddingstone conglomerate (extremely coarse) ...... 32 1. Dark-red, very dense, massive shale............... 15+ 395 Concealed base about 50+ feet above schist 740 JOHN W. GRUNER Of special interest in the outlier are the five members listed as puddingstone conglomerates on account of their unusual texture and composition. Their color as a whole is dark red. Joints and bedding planes are few and far apart in the three lower members. The lowest one (No. 2) of the formation consists of very angular “pebbles” and platy fragments of green chlorite and gray quartz- schist and gray slate, varying in size from mere sand grains to great bowlders 3 to 4 feet in diameter. They are imbedded in a red argillaceous and arenaceous cement, which makes up 80 to 85 per cent of the volume of the conglomerate. The highest member of the conglomerates (No. 13) which contains pebbles not exceeding 1 inch in diameter has only calcite as cement, which contains abundant fossil fragments. { POST-PALEOZOIC IGNEOUS ROCKS Though outcropping in areas 5 to 7 miles apart, the Red River rhyolite flow, the intrusive Opal Peak por- phyry, and the numerous rhyolitic dikes are chemically and mineralogi- cally much alike and probably of the same age. They are certainly post- Fig. ao Opal Fee f ockine Carboniferous, for one of the dikes northeast. Center of porphyry ‘ stock with nearly vertical sheeting. cuts the Pennsylvanian beds on the divide east of Red River. Only a part of the thick rhyolite porphyry flow on Red River lies in the area mapped. The rock is light gray in color. The white porphyry of the extremely rugged Opal Peak and Cuchilla de Media lying between the darker gneisses and granites offers a con- spicuous color contrast. The porphyry in spite of its prominent sheeting (Fig. 4) is very soft, and no dark minerals were seen in it. The scattered white rhyolite porphyry dikes, intrusive into the pre-Cambrian rocks, as a rule have a northwesterly trend and steep or vertical dip. GEOLOGIC RECONNAISSANCE OF TAOS RANGE 741 GEOLOGICAL HISTORY While the pre-Cambrian history of the Taos Range must neces- sarily remain rather obscure until further investigation and correla- tion with other regions, some of the events may be enumerated with more or less accuracy. Nothing is known about the origin of the ancient gneisses and schists. During the long erosion interval that exposed them and probably reduced the ancient mountains to base level, thick clastic deposits accumulated along the eastern and southern margins of the area now occupied by the granite batholith. Upon this time of great erosion a period of intense orogenic movement followed, probably accompanied or closely succeeded by the intrusion of enormous volumes of granitic magma into the overlying schists, gneisses, and sediments. Later a number of basic dikes pushed their way into this batholith. No record of the geologic events that followed is preserved until Pennsylvanian time. At the beginning of this period the present site of the range most likely formed the eastern shore of a considerable land mass west and northwest of it. Siebenthal, in his study of the San Luis Valley, has come to the same conclusion.‘ The very coarse and angular basal conglomerates of the Pennsylvanian leave no doubt as to the near-shore conditions that existed during their formation. The deposition of the puddingstone conglomerates and breccia and such bowlder beds (some bowlders with a diameter of 25 to 50 feet) as S. F. Emmons mentions farther north, on the east side of the Sangre de Cristo Range,’ can have been brought about only by talus and wash from a precipitous coast directly into deep or quiet water. The fact that the pebbles of all conglomerates consist of pre-Cambrian schists, gneisses, quartzites, and granites suggests a land surface composed chiefly of these rocks. The enormous thick- ness of the strata leads also to the conclusion that a gradual sinking of the coast and progressive submergence from the east to the west took place during this period. C. E. Siebenthal, “‘Geology and Water Resources of the San Luis Valley, Colo.,” U.S. Geol. Survey Water Supply Paper 240 (1907), pp. 50-51. 2S. F. Emmons, “Orographic Movements in the Rocky Mountains,” Geol. Soc. Amer. Bull., Vol. I, pp. 245-86. 742 JOHN W. GRUNER A more difficult problem arises from the question when deposi- tion of sediments ceased. Stevenson speaks of the Jura Trias “Red Beds” that occur farther east and south as resting conform- ably upon the Carboniferous.* Lee, on the other hand, would rather assign them, at least partly, to the Pennsylvanian system.” He also favors the assumption that during Cretaceous time the sea covered practically all of the territory now occupied by the southern Rocky Mountains. Until further evidence is found to prove that this view is correct, the present writer is inclined to believe that the site of the Taos Range proper during the Cretaceous was not, or only for a short epoch, an area of deposition for the reason that no Cretaceous sediments have been discovered on the west side of the Culebra and Mora ranges as far as can be learned from the available literature. Probably during early Tertiary, deep-seated intrusive ice resulted in the uplift of the Sangre de Cristo Range. Since that time erosion has been at work continually. Glaciation in recent time has been an especially powerful agent in the process of destruction of the mountains. ECONOMIC GEOLOGY It is not likely that this district will ever attain great importance on account of its mineral resources. Three deserted camps on the Rio Hondo tell of an attempt to extract gold at South Fork and Almozzet, and copper at Twining. Lindgren has described these occurrences.* A number of short prospect tunnels are situated in and close to some of the rhyolite dikes near Fairview Mountain and Lucero Peak where pyritization has altered the schist. Another claim is at the head of Elm Creek, near the base of the Pennsylvanian, where a narrow vein of barite and galena outcrops in the sediments. 1 Op. cit., p. 85. 2 Op. cit., p. 39. 3 W. T. Lee, ‘‘Relation of the Cretaceous Formations to the Rocky Mountains in Colorado and New Mexico,” U.S. Geol. Survey Prof. Paper 96 (1916), p. 40. 4 W. Lindgren and L. C. Graton, ‘“‘The Ore Deposits of New Mexico,” U.S. Geol. Survey Prof. Paper 68 (1910), p. 83. SUMMARIES OF PRE-CAMBRIAN LITERATURE OF NORTH AMERICA EDWARD STEIDTMANN University of Wisconsin I. QUEBEC THE EASTERN PART OF CANADA, NEWFOUNDLAND, AND GREENLAND Recent studies of the pre-Cambrian in Nova Scotia and New- foundland have been of a reconnaissance nature and do not modify earlier stratigraphic studies in any important way. In northern Quebec, Cooke, Wilson, and Tanton have extended reconnaissance mapping to new areas. Moore has studied the Belcher Islands of James Bay and finds a series of slates, graywackes, quartzites, limestones, and sandstones similar to the Nastapoka and Richmond groups described by Leith and Low. In northern Quebec, the succession from the base upward includes mainly (1) basic lavas, ferruginous dolomites, iron formations, rhyolites, other volcanics of the Keewatin type, etc.; (2) crystalline limestones, etc., of the Grenville type; (3) intrusives of granite and granite gneiss. In western Quebec, Timiskaming County, rocks of the preceding type are unconformably overlain by conglomerates and other poorly sorted clastics. ‘The youngest rocks are basic intru- sives. The sections made do not all agree as to the relative position of the Grenville and Keewatin types. Buddington'’ finds that the Algonkian rocks of southeastern Newfoundland include 16,000 feet of sediments intruded by granite, syenite, and gabbro, basic and acid dikes and flows. The sediments consist of green and purple cherty slate, volcanic con- glomerate, and upper series 6,000 to 7,000 feet thick of red and green sandstones, conglomerates, and shales, containing fresh feldspars, cross-bedding, and intra-formational conglomerates. The upper sediments show evidence of continental origin. tA. F. Buddington, “Reconnaissance of the Algonkian Rocks of Southeast Newfoundland” (Abstract), Bull. Geol. Soc. Am., Vol. XXV, No. 1 (1914), p. 40. 743 744 EDWARD STEIDTMANN Buddington’ studies the petrography and origin of the pre- Cambrian rocks of Newfoundland. Cooke? says that the underlying rocks of the region at the head waters of the Broadback River in northwestern Quebec include a complex of basic schists, which is overlain unconformably by sediments including mica quartz schists, quartzites, arkose, and conglomerate. The youngest rocks are intrusive granites. Cooke’ states that the pre-Cambrian rocks of the northwestern part of Quebec are probably all pre-Huronian, using the latter term in the sense of the International Committee. The succession is as follows: Mattagami series—In scattered patches Unconformity Nemenjish series—Seems to correspond to Grenville series farther south Abitibi series—Basic lavas probably the equivalent of the Keewatin Dresser? reports on an area along the south shore of Lake St. John about 120 miles north of the city of Quebec. Here are notable outliers of Paleozoic rocks preserved by faulting. The pre-Cambrian rocks of the area include granites and anorthosite, the latter containing important titaniferous magnetic ores. Faribault’ reports that the pre-Cambrian rocks of the Pleasant River Barrens, Lunenberg County, Nova Scotia, comprise the Goldenville quartzite which is overlain by the Halifax slates. According to Faribault’ the pre-Cambrian rocks underlying the Port Mouton map area comprise the Goldenville quartzite tA. F, Buddington, ‘“‘Pre-Cambrian Rocks of Southeast Newfoundland,” Jour. Geol., Vol. XXVII (1919), pp. 449-79. 2H. C. Cooke, ‘An Exploration of the Headwaters of the Broadback or Little Nottaway River, Northwestern Quebec,”’ Canada Geol. Surv. Summ. Rept. 1912 (1914), Pp- 337-41, map. 3 H. C. Cooke, “‘Some Stratigraphic and Structural Features of the Pre-Cambrian of Northern Quebec,” Jour. Geol., Vol. XXVII (1919), pp. 65-78, 180-203, 263-75. 4 John A. Dresser, ‘Part of the District of Lake St. John, Quebec,” Canada Geol. Surv. Mem. No. 92 (1916), 88 pp., 5 pls., 2 figs., map. 5E. R. Faribault, “Geology of the Gold District of Pleasant River Barrens, Lunenberg County, Nova Scotia,” Canada Geol. Surv. Summ. Rept. 1913 (1914), Pp. 259-63, map. 6. R. Faribault, EGeslon: of the Port Mouton Map Area, Queens County, Nova Scotia,” Canada Geol. Surv. Summ. Rept. 1913 (1914), pp- 251-58. a PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 745 18,348 feet in thickness and the Halifax slates 11,700 feet thick which overlie it. Hovey’ states that Archean gneisses appear on Parker Snow Bay, Greenland. They are overlain by Huronian quartzites, quartz schists, etc. Malcolm? reports that the gold fields of Nova Scotia occupy the eastern half of the province bordering the coast. The oldest rocks are either Cambrian or pre-Cambrian sediments consisting of the Goldenville quartzites 16,000 feet thick conformably over- lain by the Halifax slate 14,500 feet thick. Unconformably over- lying them are Devonian or Carboniferous sediments. The -quartzites and slates are thrown into folds having an east to west trend. Locally they are altered into gneisses and schists by a granite intrusion. Moore’ finds over 9,000 feet of pre-Cambrian sediments on Belcher Islands about seventy miles from the south coast of Hudson Bay. These sediments resemble the Nastapoka and Richmond groups described by Leith and Low. ‘The sediments include iron formation, concretionary limestone, and dolomite, various slates, some of which show marked banding, quartzites, graywackes, and sandstones. The iron formation consists of jaspilite, chert, cherty-iron carbonate, green granules probably iron silicate, hematite, magnetite, and shale. Diabase sills and basalt flows of uncertain age are associated with the sediments. Moore con- cludes from his study of the concretionary structures of the lime- stones and the granular structures of the iron formations that they were formed in part by algae and other lowly organism. The chief source of the iron solutions, he believes, was _ lateritic weathering. 1E. O. Hovey, “Notes on Geology of the Region of Parker Snow Bay,” Bull. Geol. Soc. Am., Vol. XXTX (1918), p. 98. 2 Wyatt Malcolm, “Gold Fields of Nova Scotia,”” Canada Geol. Surv. Mem. No. 20 (1912), 331 pp., 42 pls., 24 figs., 2 maps. 3E. S. Moore, “The Iron Formation on Belcher Islands, Hudson Bay with Special Reference to Its Origin and Its Associated Algal Limestones,”’ Jour. Geol., Vol. XXVI (1918), pp. 412-38, 18 figs. 746 EDWARD STEIDTMANN Tanton* maps and describes an area northeast of Lake Abitibi. His succession of pre-Cambrian rocks follows. : Post batholithic intrusives Olivine diabase Keweenawan Quartz diabase, minette Batholithic intrusives Granite and granite gneiss Laurentian Igneous contact Harricanaw series Arkose, conglomerate, graywacke Unconformity Abitibi group Ferruginous dolomites and iron formation rhyolites, basalt, other volcanics, etc. The Kewagama Lake area described by Morley E. Wilson? includes about eighty square miles bordering on the Province of Ontario. Cobalt is about thirty miles south of the southwest cor- ner of the area. The region is a peneplain whose elevation above sea-level varies from goo to 1,100 feet. The divide between the James _ basin and the St. Lawrence system crosses it along a sinuous east ~ and west line. Many of the low hills and streams of the region are parallel with the rock structure, most of which trends north of east. Some of the streams and lakes, however, have a strikingly linear north and south direction. Wilson believes that they follow preglacial depressions. The bed rocks are all pre-Cambrian. In many places they are covered by stratified and unstratified glacial deposits and by postglacial, finely stratified lake clays and sands. Wilson classifies the pre-Cambrian rocks into two main divisions, but refrains from correlating them with any of the units recognized by the Inter- national Committee. The oldest division consists of highly metamorphosed and folded rocks intruded by batholiths of granite and granite gneiss. *T. L. Tanton, “‘The Harricanaw Turgeon Basin, Northern Quebec,” Canada Geol. Surv. Mem. No. 109 (1919), 84 pp., 1 map, g pls., 2 figs. 2 Morley E. Wilson, “‘Kewagama Lake Map Area, Quebec,” Canada Geol. Surv. Mem. No. 39 (1913), pp- 39-122, 24 pls., g figs., map in pocket. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 747 The intrusive granites and granite gneisses resemble the Laurentian. They intrude the Abitibi group. The latter include a volcanic complex, consisting of amphibolites and schists, chloritic rocks, slate, and ferruginous dolomite; and the Pontiac series of fine- grained mica schists and gneiss, hornblende schist, amphibolite, arkose, graywacke, and conglomerate. This old complex is beveled by a pre-Cambrian peneplain above which lies the Cobalt series. The contact is sharp in places; in others it is gradational. The Cobalt series consists of two tillite-like conglomerates separated by even-bedded graywacke, argillite, quartzite, and arkose. The conglomerates are believed by Wilson to be glacial because of their heterogeneous character, their great extent, the size of some of the constituent bowlders, the distance of some of the bowlders from the parent ledge, the soled and striated nature of some of the bowlders, the improba- bility that the large bowlders could have been deposited from checked torrential streams, since they rest on a peneplained surface on which streams must have had a low gradient. The stratified deposits separating the conglomerates are believed by him to be of interglacial, lacustrine origin. The Cobalt series are intruded by a mass of syenite porphyry classed as doubtfully Keweenawan. Considered as doubtfully of the same age are certain diabase dikes which cut the old complex, but are not known to intrude the Cobalt series. They are called the Nipissing diabase because of their lithologic similarity to the Nipissing diabase of the Cobalt district. Wilson’ presents a map and report on Timiskaming County, Quebec. An outline of his classification of the pre-Cambrian rocks follows. Keweenawan—Basic intrusives Huronian—Cobalt series Conglomerate Arkose Graywacke and Argillite Unconformity t Morley E. Wilson, ‘‘ Timiskaming County, Quebec,”’ Canada Geol. Sur. Mem. No. 103 (1918), 196 pp., map. 16 pls., 6 figs. 748 EDWARD STEIDTMANN Basal complex Pre-Huronian—Batholithic intrusions, granites, etc. Abitibi group Pontiac series—Sedimentary schists, iron formation, etc. Igneous intrusives—Chiefly basic Extrusives—Chiefly basic Grenville series Crystalline limestone, etc. As in certain other recent papers by Wilson, he argues for a local nomenclature of the pre-Cambrian and against extensive correlations. Wilson’ reports that the succession of rocks underlying a part of Amherst Township of Quebec about 60 miles northeast of Ottawa, is as follows. Late pre-Cambrian—A single diabase dike Basal complex— Batholithic granite and syenite gneiss Buckingham series of intrusives—Gabbro, pyroxene, syenite, anorthosite Grenville series—Limestone, garnet, gneiss and quartzite Wilson? reports on a geological reconnaissance of a part of northwestern Quebec. The region includes a southern limestone belt, Grenville series, a northern sedimentary and volcanic belt (Abitibi group), and an intermediate belt of banded gneisses largely igneous intrusives into the Abitib’ group. The Abitibi group includes schists, iron formations, and conglomerates which have not been stratigraphically separated. Wilson? concludes that the banded Laurentian gneisses are mostly of igneous origin and owe their banding to differentiation under deformative conditions, the latter causing fractures in the crystallized portions which become filled with magma. ™ Morley E. Wilson, ‘‘Geology and Mineral Deposits of a Part of Amherst Town- ship, Quebec.” Canada Geol. Surv. Mem. No. 113 (1919), 54 pp., 1 map, 17 pls., 3 figs. 2 Morley E. Wilson, ‘‘A Geological Reconnaissance from Lake Kipawa via Grand Lake Victoria to Kawikawinika Island, Bell River, Quebec,” Canada Geol. Surv. Summ. Rept. 1912 (1914), pp. 315-36, fig. 3 Morley E. Wilson, ‘‘The Banded Gneisses of the Laurentian Highlands of Canada,” Am. Jour. Sci., 4th Ser., Vol. XXXVI (1914), pp. 109-22. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 749 Wright* says that the pre-Carboniferous rocks of the Clyburn Valley, Cape Breton, consist of a bedded series of volcanics invaded by quartz diorite and granite batholiths and by sills and dikes of basic materials. IV. MANITOBA, SASKATCHEWAN, AND NORTHWEST TERRITORIES Alcock? has made a reconnaissance of the Lower Churchill River region of Manitoba and reports that the pre-Cambrian rocks of the area include the following: Pre-Cambrian Granite and gneiss Biotite granite gneiss, hornblende granite gneiss, amphibolite, grano- diorite, porphyritic granite Churchill quartzite Dominantly a dark gray, fine-grained quartzite Keewatin Local areas of chloritic and sericitic schists Bruce? has mapped and described the Amisk-Athapapuskow Lake area on the western border of the Canadian Shield on the boundary line between Saskatchewan and Manitoba. His succession follows: Kaminis granite Granite gneiss Hybrid granitic rocks Intrusive contact Upper Missi series Arkose Conglomerate Unconformity Lower Missi series Slate Graywacke ~ Quartzite Conglomerate Unconformity Cliff Lake granite porphyry Intrusive contact Kisseynew gneisses Amisk series Sedimentary and igneous gneisses and schists, lavas, tuffs, agglomerates, and derived schists =W. J. Wright, “Geology of Clyburn Valley, Cape Breton Island (Nova Scotia),”’ Canada Geol. Surv. Summ. Rept. 1913 (1914), pp- 270-83, map, diagram. 2F. J. Alcock, “(Lower Churchill River Region, Manitoba,” Canada Geol. Suro. Summ. Rept. 1915 (1916), pp. 133-36. 3 E. S. Bruce, ‘“ Amisk-Athapapuskow Lake District,” Canada Geol. Surv. Mem. No. 105 (1918), 91 pp., map, 7 pls., 4 figs. 750 EDWARD STEIDTMANN Camsell reports on the results of a reconnaissance of a portion of the northwest territories between longitudes 108°30’ to 114°30’ and latitudes 58°30’ to 61°30’. His classification of the pre- Cambrian rocks of the area follows. Athabaska sandstone (sandstone and conglomerate) Unconformity Granite and gneiss Intrusive contact Tazin series (mica, chlorite, and quartz schists, slates and limestone) Camsell and Malcolm? present a reconnaissance map and report of the Mackensie River basin between longitude 100° to 135° and latitudes 55° to 68°. Pre-Cambrian rocks occur along the eastern border of the basin. Their classification of the succession follows. Late pre-Cambrian Sandstone, limestone, and basic flows and intrusives Unconformity Granite and gneisses Intrusive contact Early pre-Cambrian _Schists, slates, limestones, and quartzites McInnes? reports on a reconnaissance of 220,000 square miles lying between 91° to 106° longitude and 53° to 59° latitude. The area extends from Lake Winnipeg to Fort Churchill on Hudson Bay eastward to Prince Albert. The area is underlain chiefly by pre-Cambrian rocks excepting on the southwest corner, which is underlain by Cretaceous. Most of the pre-Cambrian rocks resemble the Laurentian of the Lake Superior region. In the eastern portion are found patches of the Keewatin and Grenville type. Sandstones like the Keweenawan of Lake Superior are abundant in the northwest part of the area. The classification of the pre-Cambrian by McInnes follows. Keweenawan? Athabasca sandstone—White and dull red, coarsely granular, siliceous sandstone and conglomerate in thick, horizontal beds 1 Charles Camsell, “An Exploration of the Tazin and Taltson Rivers, Northwest Territories,’’ Canada Geol. Surv. Mem. No. 84 (1916), 124 pp., 18 pls., 1 map. 2 Charles Camsell and Wyatt Malcolm, ‘“‘The Mackensie River Basin,” Canada Geol. Surv. Mem. No. 108, 154 pp-, I map, 14 pls. 3 William McInnes, ‘‘The Basins of Nelson and Churchill Rivers,” Canada Geol. Surv. Mem. No. 30 (1913), 146 pp., 19 pls., r map. PRE-CAMBRIAN LITERATURE OF NORTH AMERICA 751 Laurentian Biotite granite gneiss, hornblende gneiss, amphibolite, grano- diorite, etc. Grenville ? (Lac La Ronge Quartz diorites, pyroxenites, amphibolites, gneisses, and series) schists, and crystalline limestones Keewatin Chloritic and hornblende schists, diorites, hornblendites, ser- pentines, etc. Igneous Granites, pegmatites, diorite, dikes, younger than Laurentian The pre-Cambrian’ rocks east of the south end of Lake Winnipeg are provisionally classified as: - Post Lower Huronian Manigolagan granite, Pegmatite and gneiss Huronian Wanigow series; conglomerate containing quartz, rhyolite granite, felsite, greenstone, jasper, and chert, maximum diameter of pebbles 1 foot Keewatin Arkose, graywacke, chert, jasper, gray gneiss, and schist Rice Lake series; greenstone, quartz porphyry, thyobite, trachyte felsite, green and gray schist tE. S. Moore, “‘Region East of the South End of Lake Winnipeg (Manitoba), ” Canada Geol. Surv. Summ. Rept. 1912 (1914), pp. 262-70, map. [To be continued] RECENT PUBLICATIONS —ANDERSON, C. B. The Artesian Waters of Northeastern Illinois. [Illinois Department of Registration and Education, Division of -the State Geo- logical Survey, Bulletin 34. Urbana, 1o19.] —ATTI DELLA REALE AccaDEMIA Der Lince1, ANNo CCCXVII._ 1020. [Serie Quinta. Rendiconti. Classe de scienze fisiche, mathematiche © enaturali. Vol. XXIX.- Fascicolo 4. Fascicolo 5. Roma, 1920.] —Atwoop, W. W. Relation of Landslides and Glacial Deposits to Reservoir Sites in the San Juan Mountains, Colorado. [U.S. Geological Survey, Bulletin 685. Washington, 1919.] —Bartrum, J. A. Additional Facts concerning the Distribution of Igneous Rocks in New Zealand: No. 2. [Transactions of the New Zealand Insti- tute, Vol. LII, pp. 416-22. Wellington, 1920.] The Conglomerate at Albany, Lucas Creek, Waitemata Harbour. [Transactions of the New Zealand Institute, Vol. LII, pp. 422-30. Wellington, 1920.] —Bastin, E. S., and LAney, F. B. The Genesis of the Ores at Tonopah, Nevada. 10918. [U.S. Geological Survey, Professional Paper 104. Washington, ro19.] —Bowen, C. F. Anticlines in a Part of the Musselshell Valley, Musselshell, Meagher, and Sweetgrass Counties, Montana. [U.S. Geological Survey, Bulletin 691-F. Washington, ro19.] Structure and Oil and Gas Resources of the Osage Reservation, Oklahoma: T. 28 N., Rs. 9 and 10 E.; T. 29 N., R. 10 E. [U.S. Geo- logical Survey, Bulletin 686-F.] Tps. 24, 25, and 26 N., Rs. 6 and 7 E.; -Tps. 25 and 20 N., R.5 E.; T. 26N.,R.4E. [U.S. Geological Survey, Bulletin 686-L. Washington, 1919.] i —BuRcHARD, E. F. Cement: with a Section on Concrete Ships, by R. W. Lesley. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part II: 24. Washington, ro19.| —Butter, B.S. etal. The Ore Deposits of Utah. [U.S. Geological Survey, Professional Paper 111. Washington, 1920.] —Canada Department of Mines. Summary Report of Geological Survey, ene, Parts B (No. 1805), D (No. 1806), and G (No. 1804). [Ottawa, 1920. —CaNu, F., AND BASSLER, R. S. North American Early Tertiary Bryozoa. [U.S. National Museum, Bulletin 106. Washington, 1920.] —CHANEY, R. W. The Flora of the Eagle Creek Formation. [Contributions from Walker Museum, Vol. II, No. 5. University of Chicago Press, 1920. (Price $1.10 postpaid.)] 752 RECENT PUBLICATIONS 753 —Cuapin, THEODORE. The Nelchina-Susitna Region, Alaska. [U.S. Geo- logical Survey, Bulletin 668. Washington, roro.| —Ciark, F. R. Geology of the Lost Creek Coal Field, Morgan County, Utah. [U.S. Geological Survey, Bulletin 691-L. Washington, ro19.] Structure and Oil and Gas Resources of the Osage Reservation, Oklahoma. T. 26 N., Rs. 9, 10, and 11 E. [U.S. Geological Survey, Bulletin 686-I. Washington, roro9.| —Crark, Marrua B. (compiler). Mineral Resources of the United States in 1919 (Preliminary Summary). . With Introduction by G. F. Loughlin. [U.S. Geological Survey. Washington, 1920.| —Co1tier, A. J. Coal South of Mancos, Montezuma County, Colorado. [U.S. Geological Survey, Bulletin 691-K. Washington, r919.] Geology of Northeastern Montana. [U.S. Geological Survey, Professional Paper 120-B. Washington, 1g109.| —Cooxe, H. C. Geology of the Matachewan District, Northern Ontario. [Canada Department of Mines, Geological Survey. No. 97, Geological Series, Memoir 115, No. 1783. Ottawa, 1919.| —CourTIN, T. Casing Troubles and Fishing Methods in Oil Wells. [U.S. Bureau of Mines, Bulletin 182. Petroleum Technology 57. Washington, 1920.] —CusHMAN, J. A. Some Pliocene and Miocene Foraminifera of the Coastal Plain of the United States. [U.S. Geological Survey, Bulletin 676. Washington, ror19.] The American Species of Orthophragmina and Lepidocyclina. [U.S. Geological Survey, Professional Paper 125-D. Washington, 1920.] —DeEWotr, F. W. Illinois Geological Survey, Year Book for 1916. Admin- istrative Report and Economic and Geological Papers. [Illinois Geological Survey, Bulletin 36. Urbana, 1920.] —Duvuntiop, J. P. Secondary Metals. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part I: 15. Washington, 1919.] —Eme_erson, F. V. Agricultural Geology. [New York: John Wiley & Sons, 1920. (Price $3.00.)] —GateE, H.S. Potash Deposits in Spain. [U.S. Geological Survey, Bulletin 715-A. Washington, 1920.] The Potash Deposits of Alsace. [U.S. Geological Survey, Bulletin 715-B. Washington, 1920.] —GAaLeE, H.S., anp Hicxs, W.B. Potash. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Washington, 1919.] —Gasxitt, A. Annual Report of the Department of Conservation and Development of New Jersey, for the Year Ending June 30, 1919. Depart- ment of Conservation and Development of New Jersey. [Trenton, r919.] —GEE, L. C. E. (compiler). Mining Review for the Half-Year Ended Decem- ber 31, 1919. No. 31. South Australia Department of Mines. [Ade- laide, 1920.] 754 RECENT PUBLICATIONS —GraBau, A. W. Geology of the Non-Metallic Mineral Deposits Other Than Silicates. Vol. I. Principles of Salt Deposition. [New York: McGraw- Hill Book Co., 1920. (Price $5.00.)] —GREENLY, EpwArD. The Geology of Anglesey. Vols. I and IJ. Memoirs of the-Geological Survey of Great Britain. [London: James Truscott and Son, Ltd., Cannon Street E.C. 4, 1919.] —GroveER, N. C., Chief Hydraulic Engineer; BALpwin, G. C., and HENsHAw, F. F., District Engineers. Surface Water Supply of the United States, tors. Part XII. North Pacific Drainage Basins. B, Snake River Basin. [U.S. Geological Survey, Water-Supply Paper 413. Washington, 1919] —Grover, N. C., McGrasuan, H. D., and Hensuaw, F. F. Surface Water Supply of the United States, 1915. Part XI. Pacific Slope Basins in California. [U.S. Geological Survey, Water-Supply Paper 411. Wash- ington, 1919.] —Grover, N. C., Porter, E. A., McGrasnan, H. D., Hensuaw, F. F., and BaLpwin, G. C. Surface Water Supply of the United States, rors. Part X. The Great Basin. [U.S. Geological Survey, Water-Supply Paper 410. Washington, r919.] —Grover, N. C., STEVENS, G. C., and Hatt, W. E. Surface Water Supply of the United States, 1917. Part II. South Atlantic Slope and Eastern Gulf of Mexico Basins. [U.S. Geological Survey, Water-Supply Paper 452. Washington, 1920.] —Hawaiian Volcano Observatory, Monthly Bulletin of the. Vol. VIII. [Honolulu, 1920.] —Heratp, K. C. Structure and Oil and Gas Resources of the Osage Reserva- tion, Oklahoma: T. 27 N., R. 7 E. [U.S. Geological Survey, Bulletin 686-K. Washington, 1919.] Structure and Oil and Gas Resources of the Osage Reservation, Oklahoma: T. 27 N., R. 8 E. [U.S. Geological Survey, Bulletin 686-Q. Washington, 1919.] —Heatp, K. C., and Marner, K. F. Structure and Oil and Gas Resources of the Osage Reservation, Oklahoma: Tps. 24 and 25 N., R. 8 E. [U.S. Geological Survey, Bulletin 686-M. Washington, 1919.] —He1kes, V. C. Gold, Silver, Copper, Lead, and Zinc in Montana. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part I: 16. Washington, 1919.] Gold, Silver, Copper, Lead and Zinc in Nevada. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part I: 14. Washington, 1919.] Gold, Silver, Copper, Lead and Zinc in Utah. [U.S. Geological Survey, Mineral Resources of the United States, 1917. Part Te ee Washington, 1919.] —--- UmDEex To. f-OLUME AAS 111 Adams, S. F. A Replacement of Wood by Dolomite Baker, Frank Collins. Pleistocene Molusca from Indiana and Ohio Berry, E. W. Upper Cretaceous Floras of the Eastern Gulf Region in Tennessee, Mississippi, Alabama, and Georgia. Review by R. W. C. Bituminous Coals, Compilation and Composition of. By Reinhardt Thiessen : Bowen, N. L. Deseeaton a Gaceiieras idler Branner, J. C. Geologic Map of Brazil. Review by Bailey Willis Brazil, Geologic Map of. By J. C. Branner. Review by Bailey Willis Bretz, J Harlen. The Juan de Fuca Lobe of Cordilleran Ice Sheet Bucher, Walter H. The Mechanical Intrepretation of Joints. I Carlson, Charles Gordon. A Test of the Feldspar Method for the Determination of the Origin of Metamorphic Rocks Case, E.C. The Environment of Vertebrate Life in the Late Balenenie of North America; A Paleogeographic Study. Review by M. G. M. Preliminary Description of a New Suborder of Phytosaurian Reptiles with a Description of a New Species of Phytosaurus Chamberlin, T. C. Diastrophism and the Formative Processes. X. The Order of Magnitude of the Shrinkage of the Earth Deduced from Mars, Venus, and the Moon pi yeth: Fg rr Diastrophism and the Formative Ee raresces XI. Selective Segration of Material in the Formation of the Earth and Its Neighbors i : : ; Diastrophism only cs Booman Eros: EL. «Phe Physical Phases of the Planetary Nuclei during Their Formative Stages ; : : Dyiastrophism and “the: Romane Prorees XIII. The Bearings of the Size and Rate of Infall of Planetesimals on the Molten or Solid State of the Earth Chester Series in Illinois, The. By Stuart Weller . . 281, Clarke, John M. The Great Glass-Sponge Colonies of the Devonian Their Origin, Rise, and Disappearance ao Cleland, H. F. A Pleistocene Peneplain in the Constal Plain Coastal Plain, A Pleistocene Peneplain in the. By H. F. Cleland Colorado Plateau, The Pre-Moenkopi (Pre-Permian ?) Unconformity of the. By C. L. Dake : 755 PAGE 126 473 756 INDEX TO VOLUME XXVIII Columnar Structure in Lavas, Factors Producing, and Its Occurrence near Melbourne, Australia. By Albert V. G. James Conemaugh Formation, “Slides” in the. By Earl R. Scheffel - Cooke, H. C. A Correlation of the Pre-Cambrian Formations of Northern Ontario and Quebec 5 yay eke Crystallizing Magma, Deformation of. By N. L. Bowen Crystallizing Magma, Movements in. By Frank F. Grout Dake, C. L. The Pre-Moenkopi gat ae ?) Unconformity of the Colorado Plateau Soh ae Deformation of Crystallizing Varina ‘By N. iv Bower : Diastrophics of the Northern Mexican Tableland, ealeaceiel By Charles Keyes . sont § . ; Diastrophism and the F cere Processes ew Line Order of Magn: tude of the Shrinkage of the Earth Deduced from Mars, Venus, and the Moon. By T. C. Chamberlin Ue wa eae S en Diastrophism and the Formative Processes. XI. Selective Segrega- tion of Material in the Formation of the Earth and Its Neighbors. By T. C. Chamberlin oie. oe lane Diastrophism and the Formative Piocetes’ XII. The Physical Phases of the Planetary Nuclei during Their Formative Stages. By T. C. Chamberlin ee aiial «Sage ee Diastrophism and the Formative Pipeete XIII. The Bearings of the Size and Rate of Infall Planetesimals on the Molten or Solid State of the Earth. By T. C. Chamberlin 3 Discussion of “Notes on Principles of Oil Accumulation” by Ae W. McCoy, A. By Chester W. W. Washburne : Dolomite, A Replacement of Wood by. By S. F. Adams Eastern Gulf Region in Tennessee, Mississippi, Alabama, and Georgia, Upper Cretaceous Floras of the. By E. W. Berry. Review by RWere hadwrig genase ar ee ok Uae er Elastico-Viscous Flow, The hare ale II. By A. A. Michelson Experiments in Water Control in the Flat Rock Pool, Crawford County. By F. B. Tough, S. H. Williston, and T. E. Savage. Review by ReSAGa |e oe Tae ne SSO eae die oe ae ah Feldspar Method for the Determination of the Origin of Metamorphic Rocks, A Test of the. By Charles Gordon Carlson Fenner, Clarence N. The Katmai Region, Alaska, and the Great Eruption of 1912 Flat Rock Pool, Crawford Gomes ixpéements in water Contralli in the. By F. B. Tough, S. H. Williston, and T. E. Savage. Review bya aecnele SOR Se ee PAGE 458 340 304 255 126 473 665 366 356 180 18 659 632 569 650 INDEX TO VOLUME XXVIII Geologic Map of Brazil. By J.C. Branner. Review by Bailey Willis Geologic Reconnaissance of the Southern Part of the Taos Range, New Mexico. By John W. Gruner Glass-Sponge Colonies of the Devonian, The Great Their Origin, Ree and Disappearance. By John M. Clarke . : Ly es Grout, Frank F. Movements in Crystallizing Magma . Gruner, John W. Geologic Reconnaissance of the Southern Part a fhe Taos Range, New Mexico ; Gumbotil, The Origin of. By George F. hey, and I Newton eee Z Heart Mountain Overthrust, Wyoming, The. By D. F. Hewett . Hewett, D. F. The Heart Mountain Overthrust, Wyoming . ‘ Hills, T. M. Some Estimates of the Thickness of the hirer aed Rocks of Ohio . + ee Ice, On Some Physical Properties of. By Motonori Matsuyama Igneous Rocks, A Quantitative Mineralogical Classification of— Revised. By Albert Johannsen .. 4 sia an5s. Illinois Geological Survey. Bulletin 49: Oil es Salinas: in r917 and 1918 P James, Albert V. G. Factors Producing Columnar Structure in Lavas and Its Occurrence near Melbourne, Australia . : Johannsen, Albert. A Quantitative Mineralogical Giieinestien 5 Igneous Rocks—Revised : 5 : Beh (opt aye Joints, The Mechanical Interpretation of. I. By Walter H. Bucher . Juan de Fuca Lobe of Cordilleran Ice Sheet, The. By J Harlen Bretz Kaolin of Indiana. By W.N. Logan. Review byS. S. V. : Katmai Region, Alaska, and the Great Eruption of 1912, The. By Clarence N. Fenner : : ; . : : Kay, George F., and J. Newton Bence: The Origin of Gumbotil Keyes, Charles. Geological Setting of New Mexico : Paleozoic Diastrophics of the Northern Mexican Tableland ; Logan, W. N. Kaolin of Indiana. Review by S. S. V. Matsuyama, Motonori. On Some Physical Properties of Ice McCoy, A.W. ‘Notes on Principles of Oil Accumulation,” A Discus- sion of. By Chester W. Washburne Reply to Discussion by C. W. W Fehiene on “Notes on Principles of Oil Accumulation” ; Mead, Warren J. Notes on the Mechanics of Gooloeie Sptcnines Mechanical Interpretation of Joints, The. By Walter H. Bucher. I Mechanics of Geologic Structures, Notes on the. By Warren J. Mead Metamorphic Rocks, A Test of the Feldspar Method for the Determina- tion of the Origin of. By Charles Gordon Carlson . bp! 758 INDEX TO VOLUME XXVIII Michelson, A. A. The Laws of Elastico-Viscous Flow. II SeDeL eh Mollusca, Pleistocene, from Indiana and Ohio. By Frank Collins Baker Mt RT Movements in Ga eeleme Nene By Frank F. Grout New Mexico, Geological Setting of. By Charles Keyes . gaat North America, Summaries of Pre-Cambrian Literature of. By Edward Steidtmann ibd hie Or BS ees Omen “Oil Accumulation, Notes on Principles of,” by A. W. McCoy, A Dis- cussion of. By Chester W. Washburne Oil Investigations in 1917 and 1918. Illinois Geological Sucvey, Balle 49. Review by R. A. J. : Osbon, C. C. Peat in 1918. Review by E. Ss B. Paleozoic Diastrophics of the Northern Mexican Tableland. By Charles Keyes . Pearce, J.. Newton, George F. an, ant: The Oren of Gambon Peat in 1918. By C. C. Osbon. Review by E. S. B. Phytosaurian Reptiles, Preliminary Description of a New Suborder of, with a Description of a New Species of Phytosaurus. By E. C. Case Planetary Nuclei, The Bayeieal phases of the. Acting Their Formate Stages. By T. C. Chamberlin ; Pleistocene Mollusca from Indiana and Ohio. By mle Callies Baker Pleistocene Peneplain in the Coastal Plain, A. By H. F. Cleland Pre-Cambrian Formations of Northern Ontario and Quebec, A Correla- tion of the. By H. C. Cooke : Pre-Cambrian Literature of North Agneriéa, Sumeiakes a By Edward Steidtmann : , ; Ramensisto)-, (Osc Pre-Moenkopi (Pre-Permian ?) Uacontonnity 6 the Colorado Plateau, hes, by Cala Dake Quicksilver in 1918. By F. L. Ransome. Review by E. S. B. Quirke, Terrence T. Concerning the Process of Thrust Faulting Ransome, F. L. Quicksilver in 1918. Review by E. S. B. Recent Publications ; ees Bas, Bay ty) zs, 66a) Reply to Discussion Dy C. W. ‘Weawliounane on nets on Principles of Oil Accumulation.” By A. W. McCoy Reviews . ; : ; : ; s : 5 ; ‘ val Hee. ae Savage, T. E., Tough, F. B., Williston, S. H., and. Experiments in Water Control in the Flat Rock Pool, Crawford County. Review by R.A. J. ro deh Sin cs RACINE HR ing ie a ee Scheffel, Earl R. “Slides” in the Conemaugh Formation 18 439 255 233 743 366 6590 277 75 89 277 524 473 439 702 304 743 61 276 417 276 752 371 659 659 340 INDEX TO VOLUME XXVIII 759 Sedimentary Rocks of Chio, Some Estimates of the Thickness of the. By T. M. Hills : ; 84 Selective Segregation of Miferal in the Baumation of the Earth aad Its Neighbors. By T. C. Chamberlin : 126 Shrinkage of the Earth, The Order of Magnitude of aie, ipedneedi aie Mars, Venus, and the Moon. By T.C. Chamberlin... I Size and Rate of Infall of Planetesimals, Bearings of the, on the Molten or Solid State of the Earth. By T. C. Chamberlin SiS ae PROS “Slides” in the Conemaugh Formation. By Earl R. Scheffel ct een Steidtmann, Edward. Summaries of Pre-Cambrian Literature of North PRIM CHICH mame ie RC one al eM wl eI OAR MAR Taos Range, New Mexico, Geological Reconnaissance of the Southern Part of the. By John W. Gruner , 730 Thickness of the Sedimentary Rocks of Ohio, Saint Metimaees of fe. By T. M. Hills Til DS ey ees Mina AEM Pea a aL ca TE 84 Thiessen, Reinhardt. Compilation and Composition of Bituminous Coals . Saae e : ; : - vias Thrust Faulting, Conese? the proeess a By Terrence T. Quirke . 417 Tough, F. B., Williston, S. H., and Savage, T. E. Experiments in . Water Control in the Flat Rock Pool, Crawford County. Review ayer Aton ae Mckee rds kosher ee ewe ae We at O86 Upper Cretaceous Floras of the Eastern Gulf Region in Tennessee, Mississippi, Alabama, and Georgia. By E. W. Berry. Review by Raw. €. ; 3 ihe: aie edt eee LOO Upper Cretaceous of Tennessee, ae Studies a the. By Bruce Wade 377 Vertebrate Life in the Late Paleozoic of North America; The Environ- ment of; A Paleogeographic Study. By E. C. Case. Review by GM: 5 , : ‘ : : : ; : Sac Deen Wade, Bruce. Recent Studies of the Upper Cretaceous of Tennessee. 377 Washburne, Chester W. A Discussion of ‘‘ Notes on Principles of Oil Accumulation” by A. W. McCoy : 366 Water Control in the Flat Rock Pool, Crawford Couey) Beperimenien in. By F. B. Tough, S. H. Williston, and T. E. Savage. Review by Rear oan | ere P98 ca ae SE te Fn hcp) ahd go ae ate One Weller, Stuart. The Chester Seriesin Illinois . . . . . 281, 305 Willis, Bailey. Review of Geologic Map of Brazil, by J. C. Branner . 268 Williston, S. H., Tough, F. B., and Savage, T. E. Experiments in Water Control in the Flat Rock Pool, Crawford County. Review | Shiela eal fe : ee : : +, 050 Wood, A Replacement of by Dolaete. By S.F. Adams . : yh etic pp ats pot ay “S559 aM aes ae 7 alleer Wikecan, Vol. We cNes 5s -- The Flora _ of the Eagle Creek Formation By RALPH W. CHANEY : 9 The Eagle Creek formation is . | exposed along the bottom of the Columbia River gorge on the Ore- gon side. It is the oldest forma- tion in the region. ~ § The monograph contains a num- ber of drawings, tables, and nearly K: a hundred halftones. Paper, $1.00, postpaid $1.10 ‘THE UNIVERSITY ° Ns CHICAGO PRESS CHICAGO - - ILLINOIS POSITIONS OF ALL KINDS Never was the demand so great for qualified teachers and specialists. For ten years we have given our time and energy to this work. Write for our free literature. State qualifications briefly. Co-operative Instructors’ Association Marion | - Indiana WATER REPTILES OF THE PAST AND PRESENT — By SAMUEL WENDELL WILLISTON Late Professor of Paleontology in the University of Chicago . Professor Williston, who is widely known as a student of extinct reptiles and as the author of American Permian Vertebrates, which has now become a standard work, presents in this new - volume a summary, divested as far as possible of unnecessary scientific details, of our present knowledge concerning the reptiles of the 'seas, lakes, and rivers of past and present times. The numerous illustrations, in large part from the pen or brush of the author himself, include not only living types and twenty-four restorations of extinct forms, but also many figures elucidat- ing the structures and habits of the animals. —— ee eee ee ee vitit252 pages, royal 8vo, cloth; $3.00, postpaid $3.20 The University of Chicago Press Chicago - - - - Illinois American Permian Vertebrates By SAMUEL W. WILLISTON > HIS work comprises a series of monographic studies, with briefer notes and descriptions of new or little-known amphibians and reptiles from the Permian deposits of Texas and New Mexico. The material upon which these studies are based was for the - most part collected during recent years by field parties from the University of Chicago. The book is offered as a contribution to knowledge on the | subject of ancient reptiles and amphibians, with such summaries and SI a ae a eee ar na ee eX definitions based chiefly on American forms as our present knowledge per- mits. The work is illustrated by the author. vit146 pages and 38 plates, Svo, cloth; $2.50, postpaid $2.70 ‘THE UNIVERSITY OF CHICAGO PRESS van - ILLINOIS ‘REMINGTON FIRST— NOW AND ALWAYS Today, as always, the typewriter user who. wishes to reach the lowest cost level of typing must go to the Rem- ington. prawint Inks Eternal Writing Ink | The Self-Starting Reatinepati for cor- : Pa Engrossing 1 respondence, the Key-Set Remington for form work, and the Remington eat Accounting Machine (Wahl Mechan- Vegetable Glue, Et ism) are the latest achievements in ae the Bangst au Best Inks and Adhesives Emancipate yourself from the use of corrosiveand _—‘f}_ clerical labor saving. ill-smelling inks and adhesives and adopt— { Higgins Inks and Adhesives. They willbea revelation to you, they are so sweet, Cie an well put up, and withal so efficient. ‘ At Dealers Generally © Remington Typewriter Company (Incerporated) 374 Broadway — New York CHAS. M. HIGGINS & CO., Mites. Branches: Chicago, London — “ 271 Ninth Street Brooklyn, N. Y. ‘The Geology of Vancouver and Vicinity By EDWARD M. J. BURWASH 14 pelene®: 10 line drawings, and 2 colored maps Kae al / A thorough and authoritative survey of the region, made =—_ ||. more interesting by the remarkable photographs, which — include a profile of the Grouse Mountain spur, the Lions — a | from Mount Brunswick, a view of Red Mountain from Black ; Tusk Mountain, Mount Garibaldi from the south, a- nce 4 section from a cutting on Keith Road, and others equally inter- esting. One colored map shows the topography of Vancouver range and the other is a geological map of Vancouver and — vicinity. Rat a t . 112 pages, paper covers; $1. 50, pestpaid $1.60 2 Ohtae : THE UNIVERSITY ae CHICAGO PRESS CHICAGO : cgi ILLINOIS, ° 4 ae ‘ ie SMITHSONIAN INSTITU ON LIBRARIES 6 iin, 3 9088 01367 0179 l |